You are on page 1of 213
Gare Sade in Linge: ‘Samuel Jy Keyser general editor 4 dro the Sari Groans 3. Sta eter ‘Spon prettin Coon Grama Ry Saket Tre Sint of te fro Laue Su Kans Spcoh Sut nt Fare Coa Fo tne O Re of ag Grama a Thee Peto, ‘Bouse Peal Preach Spe: The Psonic Richa. Kaye ‘os are Farner Pn Kip DS aor Seman Cogito Ry no ey Ste A Sty of mead i, A Kan 10, Phong Spree: he Reto ener Sel on Src, Babi oS ‘ie Grammatical Be of Ling Referee Lrg Us lei hon Roben € Berk od Amy 8 Westar 12 bare othe Tho of Grammar Henk var Roma Ea Mian 18, Wont a Soene Pry Seat, te Lie eo Paste i 15. The Raptr ef lene, Ee Resa Aloe. Bt Maen tos [anaes on Stes, Montel od Sn Roger rg mage an! Paton of Koowle The Mana Lecire, Noe Glomty 1A Cour G8 Syea:Lecre om adn and gt Ceres, Homa ‘as ant far Urge 18. Seman sans Ra cent 19. von eh Soman Engh & Sid Semi Smanic. Tete Parone 20. Prine on Para n Conger Grammar, Rober Fei eter 121 Randa of Goma Syne Raber Fein 2 Mowe «Connon on Pe Apteanon and Opt, Howard Lasik end Maren Sao 23 Pata Bes, Bary Sebin 24 The Tie rom Ding 20° Eye inguin m Hem of Sybae ‘Brrr Kent a Sere uy Reo ees 24, Grad Phony, Diss Arsene end Dogs te 24, The Map of a CenmonLangog: Jaton. Maths. Teale, a n & ‘he Prog Lng Cre, teh Toman ore Smet Hagricer on Casa, Desi Petey ‘eimai Progra, Nos Ch) ‘The Minimalist Program Noam Chomaky The MIT Press Cambridge, Masaacactts ondoe, England {© 1985 Meche Inia of Tehooey Al hs serve No psf ths bok ey te erode in ay Frm by any eter: or mahaneal mers eons peecpyine ore lore ‘nslorapand ene wll person wt he pth, “Ts tok was sen Tine Ronan by Aico Trae Typing Ld Hoag Keng rd esr oud nth Une Sirf Are. uta of Cong Castopne so Pubison ta ‘Chamsty Noses “The minal program Noam Chay em (Corot sts ings 28) Anson itgrapia eens on oN O2«2090595 (ah pape) SBN O20 1263 (peat pepe) Minster ings). Tie Sere Canetti ere Piseincen ins fea st Gentenis Cage The They of Pes ad Siow Lak 13 hoger Same ote Een of Deaton ‘ma Regecinion TS hope 8 ‘A Mia Progra or Le Tey 17 meee Cours od erste Reteness 35 Indes 4 Inirodiction “The chapters that follow are based in arg parton regular Ise seminars at MI from 1986 though 1994. These have beep conning ‘ow for over 30 year, wth bread participation by ster cy, a ‘ters, from various insintions and disciplines. Io thes ntodtory temaris I wil oullne some ofthe background forthe material that Tetons "This work is motivated by to reste questions (2) ware he gen- fal conditions thatthe human tage fculty should be ented 10 sathf? and @) te what extent ie the Ingunge Fcuy determined by ‘hese conttions, witbout spas seurtre that le beyond then? The frst question in ara has (Wo aspects: what condone ate posed on the langue faclty by vite of (A) ts place wai the array of og ‘ve systems ofthe mindtrain, and (B) general considerations of eon cept suturales tha have some independet aust, nel. sing, economy, symmetry, rowredundstey, athe ike? ‘Question (8) rao pres, but not without Conte aterton Wo these mates ean provide guidelines here, aim rational ei generally ‘ofa as sch considerations canbe cad and rendered Hlausibe we ‘an ask sister a paicular stem saises them in one or another Fo. Question (A. in conta, as an exact ans, though ony pas ‘oF it ean be surmised in the Light of cute undesnnding about lat ‘age and related copie systems. ‘To the exent thatthe answer to queton (2) is positive, language is something ike “perfect system," mctng external eoeatrsins 96 well scan be done, in one of the eaonable ways. The Misra rogram ‘or inguisti theory seks to explore thew pees. ‘Any progress cmard this goal wil daepn a proble fo he Wslogical sciences that 8 tea fr fom nk: how cata ster sch as rm language arse in the mind/bean, oF for thet mtr i the onic orl, unieh one ses ot wo find anything tke te bese popes of human language? That proble: has scmatines ben posed ee a forthe copitve sciences. The ecers ate erpropit, butter lacs ‘s isplaced: they are primanly a pete fo cogy snd the bev se fs, whch as eurenly understood, do nt prowde ony bas for wha Appear to be fait well exablshedconcusons about lreuaye” Much ‘ofthe brosdr interest of the detale and technical sty of fnguage lies ight ere, n my opinion, ‘Tne Minimalist Program shares several underyng fetal assomp- ions with i predesesors tack te the early 1950, though thas have taken somevtat diferent forms as iquiry his proceeded One tet ther is «component of he han midbrain dedicated ¥o language the langue faculty interacting wih other ystems. Though rot se ‘oust correc, this assniption sess eesonably wel-esabiahed, and wil cootius to tae it for grated hex, long withthe further riper {hess thatthe language fc bas a least wo components cogeine ‘tem that stores information, and performance tems that acer tht ‘information and use it in various ways. Iie the copmtive ste tat ‘rims concerns tere, Performance systems are presumably at eas in part langugespeciti, ‘cose comporenis of the laeeunge faculty, But they ave ener sk ‘ned not to be specif wo partir langage: ty do not vary in the 'anner ofthe gm system as Hinge evionvnemt ty, TRs the smpist assumption, and snot known to be ae, though a may ‘ell be, Knowing of to iter ideas will Kept suming anguage ‘ation to he etic f the copie sate "ao bortow fom ear werk the sstmpton thatthe cognitive sys tem tracts wth the performance systems by meaas of eves ing ‘te representation. inthe echo vense of ths noion? A more spec sumption i thatthe ceive syst interacts wih jun two auch enteral” systems: the artaatory-erceptil ayer AP and the ‘onecivalinetional system C-1. Acton. there are two ierface ‘evel, Phone Form (PF) at he AP ntetare and Logeal Fort (LP) the Ct trace. This “double inerface™ propery is one way t ek ‘ress the rational description of language sound with ¢ mesing, Inacable at et ck to Aree, ‘Though comrenly adopt, a les tails, thee assumptions abut the imal wcitecire of the language Tacully and place song ‘ther systems ofthe mindibrain ae nt at ll ebous. Even within te genera ramework, th idea that arteation snd perception involve the same interface representation is comtroveria, and arpealy incre in some fendamental way Problens relating othe © antec at tl more obscure and poorly understood. Tilt keep to thee fay cone entinal assumptions. nly noting here that i they tim out to be ‘ore, even in part, that Would be 3 surprising and hen intersting iscovery “Te lending qustions that guide the Minimalt Progr came into focus asthe peinspler an parameters (PAP) model took Hage shut fiten years ag A look at recent itory ray be help in placing ese question in content, Needs 104. thete rma ae schematic ad aleve, ad beet rom hid arty generative grammar fice two indie problems: 10 finda ‘say to acount forthe phenomena of particu languages (decipive dcquaey") and to explain how Koowedge of thes fats aie In the tind of the speaker hearer ("explanatory wdeyuary”) Though it was scarcely recognized atthe tine, hie rseatch program revived the cow ‘ims ofa rch tation, of which erhape telat jor mpresetatie was Ouo Jesperen* Hpertenreognied that the xrucres of lat guage “come ino exisnee inthe mind of «speaker by abstraction fiom experience with uteranes, ying “notion of ther strcre™| {hat “deine enough to guide hm in Framing stnces of his om oc “Tew expressions that are typically raw wo speaker end heart We ca take these properties of language (0st the priary onl of "puis theory: spell ou clearly this “notion of stucare” and the Procedure by which i yields “ree expressions." and 1 explain how it aie in the mind of the speaker -tbe problems of descriptive nd Plaatory adequacy, respectively. To atin deserptive neque for arcu language L. the tory of Lis grammar) rst characteie {esate stained bythe language facuky, or at eas ome of aspects “To ain explanatory adequacy. theory of langue must characte ‘hei sate of he language fact and sv how tas xpi 0 the state tained Jepsen bed frther that iw oly “with regard to ‘tax that we expect "iat there must Be soeting in common tall aman speech”; there canbe a “univers er peer) grammar” hence ‘perhaps larseaching account of te ntl state of ie langue aly in this domain, though "no one ever dreams of «universal orp gy" That ies too ak eran vonage eet work the modern period thew adtonal concems wove dispiced. in pat by bhavivsetrents in prt by various structurale appa, ‘ich raicalymarowed the domain of inquiry while greatly expanding ‘he tse for some fate inguiry that mish retun t he aio ‘ad surely valldconoers, To aes tern equ attr under standing ofthe fact that language eves fii se of este meu.” in one uric formulation. Advances in the fornia sciences provided thal Uundestanding, making it feasble to deal with the problems construe. ively. Generative grammar can be regarded as Kind of confuses of long-erporten concems of he study of lrguage and mind, and new understanding provided bythe formal iene. “The est forts te approach these problems quickly reveled tha ta- Aitionalpranmatics| and lesa! dee 60 not hep to devrte, et lone explin, the mest elementary fis about ven the bests lan ‘guages. Rather, they provide hints that can be used by the rider who vey har tai knowledge of language, and of particular langues the eral Lope of equry was, substantial measure. spy ignored. Sine the requisite tact Leow is easly aened without ree ‘lon tadienalgramars wad daionarenappenr to have very breed coverage of lnguisic dat. That aniston, ower a We qu) {scour when we uy to spel out whet i taken or granted the nature cf the enguge uly an its sae paniculr cases “Tiss harly a Staton unique to the sty of language. Typical, wher quesicas are mae sharply formulate, i lewaed that ven ce smenily phenomene had exaped notice, ad that itive acount tha scat spl and persuasive are emily inadagute ve are mid ‘hat ape fl to he ground because thats atoral plac, there ‘wl be no srioae science of mechanics The sue i tue i oe st {sfed with taditioal rus for forming quesions, or with the leical otis it the most elaborate dictionaries, none of which come cose to esting imple properties of hese igus objets. -Recoptitin of he unsuspected ices and compesity ofthe pheno ‘ns of langue created aversion atheen the goal of deserpive snd ‘rplnatory adequacy. I wa cst tht to achieve explanatory seus fu theory of the nal sate mst low only ited arom parc. ‘lr Iaeguages must be largely known in advance of exprine. The ‘pens pent in Univers! Grama (UG) ms be highly ese [Experience mist sfc to Bx thm one way oe smother ing a sate of the lamespe fact that dtemines the vaied and complex aay of expen. hit sod and meting’ ad even the most super ot ee hn hare tno eg ter the dat of exer But the gl ef explain nea) Fe cell utero he dtc pert syns we need ‘n pur of doeipieadeqsy in aceasta fore {nt engine Te problem ws cred ty he ge eof omen dicoveed ben steps were mee To ferme eal ‘steno Lng “Ths teson dine the reac program of ety generate rarer attest he dey within tha cence ee. From ely 1s is ema eine ws 0 aap in ey the ome le sens deed for articles ving als {silent in hr praton by thaw UG pce Sey ints don reese vanity of neugespestic pers th Contig to expansion adeacy. They sho tend to eld net fd more tral hors, ping grounder oan erental tals appreach Theres ey ht ths eth ci cul tm outta “ope rier and more cme veraon of UG recs permsile vay. thar comune to he pny cla Explanatory adequacy. Te ral, howe, theo ener hve proven ob mut refracted ey se One usraton cones tefondent peice mth evelapeg empl coverage: Rept, thas foun tha hoa eng Sted en mast be eed by mnredidan os. The Cateye Dee So repr tha the nd cae eugene hes bso ‘ching prin nai. Aes in eure oye a & ological ace “The rt mist inthe PAP mde Com 1h, or oe fsslaton). This cena ra ea fom he Ach a ton of tesa of yoo of tei gtin, far mere so taney cate ona, hh eal ens rl tonal com Ces an approves to theta een vy a te ee and $-Srutre; he Empty Category Pisce: Xbar theory sally; the operation Mowe the spit hypothess and other. All ae minted o subsanialy rized in secre chapters, prey ete. ‘The co ret ia picture of langage that difers considerably fom ‘vents ied peers: Whether hee sep arc onthe it rack ‘ener, couse. ony time il tl Notes . 1, Forse dscsion of tht ia eChmsky 1M. fering Ena 9m Eran ths the eri obese fatal for cone sore 2. Acai, cs, oet Chery 1S 3. The em ory oo nro i tht gs th he egane - ‘hyn pce, wah pc elt a rae Werk oe ps dear nan ngage ceri hie uso! saren wl enn 0 {he the ters, tet hoot any ipl soow prey of eu ye Wheering othe cue of sponge, 41 Fr seme dcuson ee Chak 177, ht 5. mupret teri of coun to dren theyre se Ta oom foals, egestas bythe haga ay 09 {pera st) oars of tet cing ages: omens nee Sos expen of othr gia, post nent emer mel (6 Ths, wit ell “Eaghs” “French” “Spun” oo en oder ‘elon owl in haraateus Spt commen el the Now tnen Congost proxy (0 Corman soc = Beg tish oter Fates tht ear Svs te reared ar poyeras oh age ey Pau te ova easing si har ona ta the pei he Tange Tc) et eject ft ar! wenda san ey ‘umpios retake for gale he yo crams erly Chapter t ‘The Theory of Principles and Parameters ‘vith Howard Lasnik’ 1 Introdecton Principlevand.zarameters (PAP) theory is not 9 precisely articulated theoretical stem, but ase a parcel apprcech fo cas problems ofthe sted of anges, guided ty certain fading ideas at had een taking shape since the origins of modem generative grammar some 40 years ago These iat ery ito a Gtncve approach 10 the topic by abou 1980. nthe years Se, many spe variants have Been eveloped and explored. The epic base ofthese ings bat bo _weaty expended as they have ended o languages of wie vain Iypes and have eugaged «eich broader range of evden concen Jangvnge and its ws, also pentating wo far greater depth, In his survey ‘we will not attempt to delineate the vanity of propos the fave en investigated o oases theremin success and nadequacis Rasher, we will pursue a yarcula path hough the ara a des and ins that have been develope, scmetimes ating oer directions that have been persue, bot without any ate 0 be compehensve: similarly, biiographic references ae fete comprehonsive, usualy indian only few sie of particaar questions. The de of 2 tcl path should be regarded only a an expository deve, sh cf fort oda the kinds of questions tat ae being addresed, some of the thinking tha pues mich earch, and it erie ciation, We “Tis chop, couthred wits Howat Lae, wa ly plied in Sy lax An ura! Honk of Cneguray Revers hey Johan _hect,Ari on Seow. Wollgen Smad. a'Then Van (hn ‘and New York Wal de Goer. 199). le oppess Rv wth Ton oe ‘y permiton ofthe pub. " caper o not man opty that shes prt choces ive en well eta Ssh in eontas to cthers, ony some of which we wl eal ven fo “The study of pererative grammar has bee ei by sever fund mena poblems ach witha waditoral aver. The bane concer f 0 termite and characterize the hnguistie pace of particular ini tah, We ar concerned then. with sae of he nigga. wich 1c undestand to be some arayofcomtive tate and apc pat ‘ular component of te human mind brain. The lena faculty ha 4am ini ste, genial detemied i the nogmsl course of devel ‘pret it uses trough a sees ates in early childhood, teaching 2 rately sable sady sta that undergoes lite subseques change, part from the lercon. To «good fst afprenimation he ital ate sppens to be uniform fort specie, ASAP tratinal fers Yo Speci segs we cl the theory of teste atid i roma a the theory of he iti state Une Grammar (UG). ‘Thetis ao reason to Beleve that tet ste is in cul respects «spiel characters of humans, ith prope thtaphear to be ur, ‘estan the Biologia work. IF tut tht a matter of broader ster, ‘ut one of no get reeranc to determining the properties ad nature ofthis facly ofthe rindi. ‘Two fundamental problems then. are to dstermin, fr each ini ual ay, Jones) the properties of the steady state that Jones language Faculy ati. and the properies of the nial state tha sa common human endowmen. We Gitingish tetneen Jones competence (new le and understanding) std fis performance (what he does with that rowledge ard understanding). The steady state contines Jones's ma ture guste competence ‘A salient propery ofthe seuy ste that it permits nlite we of frit means te borrow Wier von Humboksaphor Apert hole of hee ite mean sa por lnguage taking langue to bea way to speak and understand a tradtonal formulation Jones's ‘competent is consitue by the particu ste of ite means Be has equ ‘The nocon of “infinite we requites further ana. f the gh of| Insights of the fora scienes in the 2th enty, ee singh 40 seme ofthis notion, the st reatingo competence the sco to fe formance. fn the fst sense a language species am infinite range of sen Dole objsts, which we cll senna despre (SD), We may thik Pitino Pareto s ofthe language. then. a8 fitely specif penectve recede une= ton) that enumerates an infinite set of SDs. Each SD, in ur, species ‘he fall ray of phonetic, semantic, sn sybacie prope of parte ‘ie ingle exresion. This sense of “infite we rls to Jone linguistic competence: the perrativ procedure wit sft seope "Tne second sens of iit use™ has to do with Jones's performance a he makes use of his competence (0 express his though to ref to reduce signals, to interpret what he ear, sad 50 on. The langues acl is embedded in performance systems, which acces he generative procedure. I in this broader conte shat quero of eaten an tne of SDs ars, question of arculaion, intonation, tnd the ke: How does Jones say X? What Jones talking about? What ‘oes Sones ake Smut o be sping or intern to convey! And seo. We might think ofthe SD ar providing isructions vo the performance stems that ema Jones to cay out these ection. ‘Wen we sy the Jones asthe anuape Ly we cow mean tat Jones's lanpuage fut i in he sate L, which me iden with» proertve -roeedare embeded in performance sytes, To sings this cor ‘Spt of anguspe rom eters, et us fer to 8 Hogue, here Hs 0 ‘sugs “ier.” “individual” nd “itesiona.” The coop of an ‘age is intra in tha it deals with am nner sate of Jost ‘in, independent of other eles inthe woeld nil tha ‘deals with Jones, ad wit language commits only deviate) as ‘soup of people wth smitar Elanguaes I intensional nthe thai a) sense that the language sw finetionspetied in intension, not ex ‘selon ts eterson isthe st of SDs (vt we might cal the ere ofthe engunge) Teo dstnst Hanguages igh, in princi, have the fame structure, though asa matter of enpinca fact, human lngings ‘nay happen net perm thi option. Tha i tight arn et ha he range of Hanguages permite by UG isso nattow tht the theoretcl ‘ion is snpy ot relied, ka here are no distinct Manghages gen ‘ning the sane set of SDs. This seem in fat, no uly Bat i is ots lope necesity. When we we the term ange Below. we mean angus. “ln the cast work in generative grammar, i was sumed that Jone’ language penerates un SD foreach of the perme phonetic forms for han language 0 beeps! by UG. Thus, Jeness Janguage assign a particulr satus to such expuessions a (1) wher ¢ (ace) indicates the postion a which the quasion word tconsued (2. otis seeing rk hat Mary Bd + (anew: the cr) 4. what do you sonder wheter Mary Bed 1 answer theca) © how do you wonder whether Mary fixed the carr (ensne: with awench) © expression of Seah, Hungtian, fact. some of the mos insti recent work has bron concerned With he difleencs sete by (1c, heh in some sense “deviant,” but assign a fleet stats by Jone’ langage (tions 3% 141) and one might wll enn about the languages of Jones and Wang by sying thet reactions to slerancs of Sah “Anotber sion that appears commonly i the Hsrture He “ormal angus” in the technical sone set f well formed formals in afar tar vatety of formal arithmetic, "2 + 2)= 5" bat aot "2+ ~2)5¢~ (Cast eset a Elanguage where Es to sgget “eternal ad “ek ‘ersona” In the ceory of rma anaes, tbe E-angue is defied by aipulation, hence is unproblematic. But i i «question of emptied fact whether atarllnghgs has any eter 0 this oto thet wheter Jone’ Flineuge generates not ony set of SDs bul alo iatnihod language some subset of the ponte forms of UG. i ‘lating ome burro ao hese of 1). Apt from expiry passes the concer of Eanguageseareely appetite tradion of generate aroma that we are considering ere. As diet fem the notions di ‘sed err, i tat no Known statis i the stay of langue. One ‘nih deine Etanguage in one or another way, but # dows ot seen fo ‘uate how tis is done: ther so keown gp i ings thors, mo xplnatryFuneton that would be filed were rch cones presente ere, wl play o ole or dacaton, nthe ty of formal anguoges, we may distinguish weok generation of Elanguage from stone generation ofthe sircture of the langue (these of SDs). The weak generate capac of x theory of Uanapes ' the st of Flnguspes weakly seer, adits song gerenve poets isthe set of strates strongly genersted. Ia the sty ef natura language, the conceps of stroctire and stor generation ae ceil the umepts of Elanguape snd wotlk gesatiom best rare an et haps not empirically mearingfl ata. Not at f Eanes do ext, ‘hey ate att considerably further resve fom mechani si Beha ond Pacts ” age Thus, the cid is peste with specimens of = Ibvioe in particular ccemstanes aad acquires an Tanguage io some ‘manner to be determined. The angnage sate ofthe inden ‘hs a certain structure (ie, strongly penerates a et of SD). tm oF ‘may not allo weakly generate an Eangege, 8 Tigly abstae cet ‘emote rom meckansms ued behavior. Inthe terms rst outlined, we ean comider soe of the case prot sem of te study of tangs, @) 2 What docs Jones know when he has patel lige? 1. How id Jones acquit this knowles? Mow dees Jones pt this knowlege to use? 4. Hw did these properties ofthe idan coe in the spe? How are these properties realized in merase of the brain? Under 2). we want o account for 8 wide varity offic or example, that Jones knows that a Pinshyrnes with fn Eich expression of (1) basis specifi sae. © IPMay is too elever to expet anyone tea then we dont ‘expect anyone to catch Mary (bt noting eat bout whether Mary expets anyon oat ws) {41a stoo angry o ros the mesg, then iter May 0 angry that she cet run the meting. o he sso angry ha ‘ust rum the meting (empate:the coed sto gr tr he ‘meeting in contr, which meetings Mar o0 amet ra bk ‘only the Former ocndeviand interpretation ©. Mary pred the Rouse row, then i exterior (nak ‘neces isinter) x brown {AF May persuade ilo got college, he Bl caret intend to go twealleg (utile Mary mayo tay not have) The proposed answer to problem (28) would be tht Jones has a sage L yeneratng SDs that expres such fet as 3), Note tht Jones ‘as this owls whether oF not he rear of thee ft Abt Ht fe it may take sme effort to eit such awareness, andi might even te beyond Jone’ capacities. Ths i 4 question that fills within the ‘cade content of performance sytems. “The ansocr to problem (2) sn substan pan in UG, The covrect ‘teary ef th inl ate wl bee enough east forthe atment ‘of spf language onthe basis of be evden aval othe eid, bu oot so ras 1 ext atainabl Inganges We may proceed fo ‘ska wll how environmental factors and mattatons proces ni, 2x wit heist ate ested by US. Treble 20) cals for the development of performence theereh ‘mong than, thors of proton and inkpetation. Put Eenerally the problems are teyond rac: would be unreasonable to pose ie prot of ow ones dies to ay hat he dso how Deiter ‘what be ens in priculr circumstances, Bt highly seized aspects af the problem are simerable to study. A standard emprical hypottet ‘atone component of the min rain ia por, which assent 06 to signal (abstracting from otter crests relevant wo ln Dretaton), The parser presumably incorporates the language and nich ss, and the Rypotheis is that inlepretationinvolss such a sey, embeded i ther Tehas sets been argued that Engst theory mua met the em ‘scl coniion that account For theese sed pity of pring But baring does no, infact, have these properties Parsing maybe low ed “ic, or even imposible, and it may be “in ero” in the sence that the pereetasipned i any) Fito match the SD eseiaed wih the sional: many fair eases ve ben suid. Im ger ts ot the ‘case that language is eadlyasale or “designed for use.” The sebparts tat are wed are ube vl iologkal considerations He! Us to ‘expect no more tno tat. Sitar, etening to prcblem (2), there '8 1 «pron reson 0 exec that the languages permite by UG he ‘esenable-—tht atainuble ue normal cteumaonees. All de we ‘am expect that some of thr may te; te ots wil nt be ound ‘numa seis. If propesas within the PAP approach ae clse to the ‘mar then it wil fll tht neuapes are in ft eat, bu that sm empscl discovery and a fathe surprising one, Problems (ae) appear to be beyond serio ingly forthe time Be ing, along with many sna queionssboutcoeition general, Her apuin one must te wary of many pitts (Lewontin 190), We wil put thse mater ide A grammar for Joes is rif (or othe exe hat) the erp it estes isthe one Jones has. In that case the grammar wil accoun! for sich Fact 5 (3), by providing ensue tat genotes appropriate SDs A true gramiar i ai co weet the condition of deserting ade ‘ocr. A ieory of UG is tue i (oro Be ener thai corety 3 operand Prancrs » sere the ii state of he langage fey. that case wil presi descriptively adequate grammar foreach sine language. A tue {ory OF UG mers th condition of explanatory adequacy. The terme oly is intend to suggest a cerainplausbe pte of explnaton Giver an aay of facts Such as (), we ean give eh socount of thee tne fev! by providing gramme for Jones, ahd we can provide an ek [laoation fr them st a deeper lel by annering problem (2). that by showing how thse facts dene rom UG, piven he "bur cond tine” set by experince, Note that this pater of explanatcn, thoveh standard, makes cernn epic sesumpons shows the actual poset ‘af susion that are by 80 means obviouly tre for example ha the ‘process a i were istantancs. Sack atumptions te inivetty fupponed to the etert that the explanations see, ‘Any stows approach to complex phenomena snvolves innumerable ‘teaizations, and the one just sketch ts na exception We do not ex: et to find “pre istamiaions” ofthe iil sae ofthe langage fac Aly iene of UG). Rater. Jones wil have some jumble ef rats, tau en the peculiar pattern of fs experience. The enplaratry model cuted dea specifies with language acqulion ue the Weaized ‘xndtions of @ homogeneous sprch community. We acute Hat the ‘yslem described by UG i real eompunent of the mind in, pt to ‘Se in the compe ercumstances of ordinary We. The vabty of thie assumption is hardly in question. To eect would be 0 ase ther that comtomogercous(conictng) data are regi for ngage ‘ssouson, oF (2) thatthe mind/ratm does indeed have the system de Serbed by UG, bu i isnot wed in langue sequition. Neither se ‘mption ie remosly pluie. Resting therm, we acep the appreach {Nat ound asa reasonable apreach fo the tuth about herons and & iy prerequisite 10 any serious inguin ite the complex and cote Phenomena word Furthermore, een if a homagencousspech cominunity existe, we would ot expect its linguistic system to Bea “pure ce” Retr, all sorts of acedems of history woul hve contained the tem, a the properties of (ough) Romance versus German origi inthe kx 2m of Esgsh, The proper topic of ngury, then, shoul be a theory of then tate that abatacs from suc asin vil mater, For ‘othing purposes (and nothing more than ht} we may make 4 rough hd tentative distinction betwen the core of x language and its pera where the core consis of what we tently atu to te: pe = chap \nstuntitens of UG and te penphary consists of marked exceptions CGreguaeverts ee). Note hat the periphery wil als exhibit properties ‘oF UG (¢gsblut phenomena), tough es raneparsely A reasonable approach would be to focus atcaion on the core system, puting aside phenomena that esl rom histo cident, ire mitre, perso ‘tiosynorses and the lke. AS in any eter empiielinguy, theory internat considerations enter int the effort to pore this Oued We capet further dstnctions to be necessary (consider, for exarple,the phenomenon of dosnsrtion i Engith ain (le) ot on 8 pe with nega verbs, but not of the peneralty of fronting of question werd). “The preceding remark are argey concepts thous ot without ee pirical comsequtnes, We now pried along paritlar path in the ‘manne nfs eri, assuming fer emia kat cach pnt ‘We assune thatthe Lnguage (he genertne procedure the langue) has two components: a compuaicnal system ands Fexion, The fest _serates the frm of SDs the second caaceies the ea ems that ppeat in them. Many crcl questions aie am to how thee ste ‘ere We wil sume that ove aspeet of an SD i 4 system of ‘enreseration elled D.Smnerare. st which lea items ee laser Dire expresses excl properties in a form accesible othe omputtional tem. We assume rhea dtinetion beeen ecto aod diana! process of morphology, the ater intra tothe tren, the former Javalvng cmputaional operations ofa broader sytactic scope. Tete onputationl operations ight involve werd format or checking Conse for example the pst teme Form wale. Te leicon conta {he oot [ak with its iiosneratc properties of snd ening. and form sped: nd the iesionl fare fre coe va of ich i Inst. One ofthe computational re call t R,esocsts the two by combing them (cher ain [walk] to flee), or convey). We nigh interpret hs deseripve comment 80 way Ore possi) tha: fk is drawn from the texicon as sus then combines it wih (pes). second possiiy tat processes interea o the lencon (re stander) form the ord Bid withthe properties (valk ind Ie} akeady speed. The rte R thee combines the amalgam with [past checking and ensing i inns eta [pst Un thi ese the fescon is more srctred. I contass the element [wake before, along wit rules indicating that any verb ray’ ele ietiesicaly possess iran Poremtr, a such properties a gas purl. and the like. Similar gustons ais shout complex werd (ciustvs, noun corporation structures, comm ound nouns ete) AS these topics are puri with more peson, Shin more closely articulated theories, important an often sl em ical asus arse (Marantz 1988, Fabb 1984, Baker 198%, Di Scialo fd Wiliams ISH, Grimshaw 190), "The SD provides information (o be ire by perfrmance sys tems) sbout he properties of cach linge caprenon, including sound and its meznng. We sssune that he design of langue provides fs eriay of symbote sjstems (lr of mpreenroin Tallin thee fats, incoding the level of Phenetc Form (PF) athe evel of Lege ‘Farm (LE), spciving aspects of sound and meaning, repectvly, iv tofer as thy are Knit determined. Ancthe i the level of D- Sucre, which lates the eormpzational system andthe ion, ‘The kel PF rust say thre bass conditions of adequacy. Te mst be acral inthe roe thal an expresion of ny actual o pot human lngge is representable within itt cnust be en interface, in that its elements ave iterprettion in terms ofthe sensoramotor oy ‘ems Ad it must Be wor, in at Chis interpretation ufos for sil languages, 028 to capture ll and oly the ropes ofthe sys ‘flangeage 0 such ‘Tee sane thes conditions Nol for LF. To capa what the language faculy determines about te meating ofan expresion, it mist be Wn ‘vers in that any theupht expres in » oman language is pte ental in ita inefacein tha hese seresenations have an interne {ation in tems of other stems of he mind ban involved in hough ‘efering, planning. and so on; and nifrs, in jot the ses that the ‘Mhonete system is We wal put aide importans ques concerning the fsature of the LF itrfce does i iveve conceal sem (Icke dot 1983, 1990), a ae theory of meaning «clusal theory of ref ‘rece, ete? The conions are more butte than inthe case ofthe phonetic analogue, because the ystems atthe terface are much es ‘sll uidetood, but there none a wel of evidence Brn eno {allo substantive inquiry According to ths conception, then eich SD contin the interface lees the extra interac eves PF and LF, ad the intr itoce level of D'Siucue. The elements at these levels ae furleranjecd ‘mo festre:phoeoogia, selecina,esegria ands on Ingenta, 2 ayer! ‘ich symbol ofthe represerations ia Feature Sin espects to be fr ther specie, ‘A furter assumption, developed ia the Eten Standard Theory (EST) that these levels are vot elated cy rater, tei elton, are medated by an ntemedate level of Site. Adopting ts ew, ach SD is a Sequence (2 2.80), where and 4 are represen’ at the external incerfae feels PE ane LE. 8 is the etna strfce of ‘computational sjsew ard lexicon, a 6 i envative. The fist three eels mee epiialconitions imposed by tbe performance systems ithe leon, The level ofS Strctre rst late t thee te eee tthe manner specie in UG: we might hak of formally 9 he (peesunatiyunigue) “clon” to eset of condos. the subse eent dacusion we react ours largely fo the els D Stacie, SSirsture, ad LF and the relotions among then (ht sh mare Sense) We are thus concerned pamly wih he dation Ieee D> Structure to LF in o DSinture «—~ Lercon PF = --Ssinsture 1 Lr Sune questions arise ao ow the lations among these els ato be consrua: special, there an inberent "rections," 3 tt the relaions should be consuued as « mapping of one eel ence, ‘ris thre Simply a nonditetionl elation? Te Foeulte this sa el fepircal sue i not siple mater, and empirical evidence 10 As tingih such possi s nox easy fo come By. Bot mtresing (sod coufitng arguments bave bee presented, Discrimination rong tes erative tacos pel if we adope (as we wil) the ‘sandard EST assumption, from the early 1970, tha representations nay ich emp eteorcs (ECS ements (Fate set) ha are pe Fey substantive fom he point of view ofthe computations stem, but that donot happen o be assigned an interpretation by te mapping from Stuur 10 PF though they may have indir phoneti ees, ‘hs the contraction rules of English comert nent 10 te the phenol ‘cal word wana when there no trace terse (who do yew wa see ut rot whe dy wenn se fel (Consky and Lasik 197), Principe ond Parts » ‘We wil temavely proseed on the assumption tha the ations aes ‘in ft, cvectonal DStrctreis app to S Siac, which (nde pendent) mapped to PF and LF. “The caren modern work in petratve grammar borrowed standard teas of tditional grammar, whic recognized (1 that seen as 4 Iherarhy ef phrases (noun phrases, clases, et) ad that ts (oie ead) ene foto certain grammatical relations; ad (0) that sentences talons to various grametcalcomioctions wih systematic relations “mong then, some more “bas” thom others actives mone base thn pastes, declartves more tsi then interrogates, ete), Comeipond- Fags, the cars versions of UG provided wo kinds of re (0 phase race rules geerating SDs tha express the berry of phrase, nd {AD transformational rules thar for grammatical eonsruciene fro absretunderyng. forms, with mote transformations tevebed in fot ination of theless bese constructions (us, only abiatry transfor tation apply 0 fom active declarative (Reel sentence), Dut sme ‘ptonal ones are involved in formation of fsves, evogse, ce) “The pase wractere rules provide x "seometrial™ acount ef games al Faton, understood felatonaly nat, sujet not aye ‘epory tke noun phase or ver, but Undestood ax the rion Sbjetof holding of the pair Ue, ki) im Jon jt, and 30.00 (Chomsky 1951, 1965, 1975). These notions were dein in sich nay that he phe trite rules (1) generate D-Stscures deep sr tors), each a phase marker that represents eazchy ane ratios “Transformations conver thes object mo new psc markers the later EST version. at noted. DStrctre art mapped to $ Stuer by fh rotons an he er af mapped nen 10 PF and “Tie coor to phase stuctre rules nar ako sugested by other ‘esideration. The earkest work concerned on wht ow called ‘erative phonology. and inthis drain “rewcing rules” ofthe frm XY, wiereX ican expression “ewriten” ae in the cote of er ‘ation secms an appropiate device I thse rls are restricted to the foam KAY = XZ¥./A sane symbol end Z nonnhthen we have & ‘stem of oes that can form phase swuctre representations in 2 net Fal wy (comevefee rules, ¥ ae nl) Further movation derived rom she theory of foal sjstens. Grimmatiol transformations 3s erative dvi were spgested by work of Hari (195), wich ed ™ caper {formal riaions among expressions a devi to orm” txt for the analysis of dacoure 'As for UG, the eats verions ase that provided a format for rule syste anid an evaluation mee that sega a “ae to etch {creative procedure of the proper format. The eral empirical eon- ion on UG. then is hat the system provide ony fev highvalud [Hamas onsen withthe kinds of dta available tothe i per haps only ene. FUG is fase ia this sens, the furdamena pele (2b) canbe adresse (Chasey 196). “This approach recorded many achievements but feed a fundamental and resuren problem the lesion between deeipve and explanatory ‘adaquacy. To achiew deseiptive adequacy it seemed noes 0 each the format of permite stems, bein doing 30 we lee the property ‘of feast, so that problem (2b) stil wsesoled, The crf arses 2s t00n awe move fromthe inline hints and exroples of wition franmar to exp generate peoeduts. IW was quickly repre "har the prom is inerent in te hans of re systens tat were be ‘consider. The most plausible approach ro its to try to “factor ou” fvearcing principles that gover sue appscason seca. esiging ‘hem to UG: the actual rues of pranar ca dn be ven i hes it foe, with thee princes ensring hat they wil operate in tach 1 way a8 10 yiid the observed pienomens in their fll complex (Chomsky 1964, Ross 1967) The li that might be eeched i tht rules se eliminated ciel, the upparet cues” being deduced from genera niles of UG. inthe sna hatte interaction of the principles woud eld the phenomena thatthe cus hid been constructed to esrb. To ‘he extent that thie sult can be achieved, the ues postulated for patic- ‘lr languages willbe show ro te epiphenerena ‘Such ess were pursued with g00d deal of ster from the ealy 1940s, teding to the PAY approach which assumed tha the bt ean it fact be steed: the yp fe tht al pincipls ee signe to UG and that language vrtion eerie to crtain opis fo how the ‘anes api Iso, then rule jsems ae eliminable, at east fr the Coreof the tnguage. “Tost, comider agin (Le), repented hee (6, wha do you think that Mary aed 4. shat do you wonder whether Mary Raed ‘© ow do you wonder wheter Mr) fed the ear Pipe and Parent » ‘The gol ito show that the question words or fom the poston of # bya general pind that allows movement ite freely with the op- ‘ions, nterpreiatons, and varying satus determined y the interaction “ofthis price with oer. ‘What she sates of he rls (9 (phrase rr) and (1 (anor rational) under ths cosention? The warsformational rls sil exist, bt only principles of UG. rey pleat ro arbitrary expressions ‘Sich deweer appt o be unavcdsbe inane or another form, whether taken 10 be operations forming derivations or reions ebished on -mpresntons. AS for phrase srcture rust appears tht they ay be completely superfuous. That would not be too surprising. With the “advantage of Binds we can sce sha, wake transformational rele. they were «dubioor device opin with, seapitlting information tat mos be pres incimiobly, inthe Fxcon. Fr example, the fac that perma rakes a noun phrase (NP) ad lesa phrase (CP as om Pemex ay Isl property, equi tht thee be phrase trace fleseling V-NP-CP an istincaion of he phrase XP headed by the verb persuade: and completely general properties require further tat XP nat be VP (er psc, ay say, NP- The apparent cena ‘otphasestrocture rls become cer bythe te 1960s, ith the sepa tion of he lexicon fom the computational sem and the evelopment “Abr theory tion 1.3.2), ‘Theiss canbe sharpened by considering two properties that de- ‘evptivestements about language might tave or lack, Thy may or ‘ay not be language they may oF may nt e construction: rial. The snerens ef traditional gramme wpa both ‘us and consimetion-peicular, and the sme ise of the eux of ‘rly aeneane grammar, Consider the role analyzing VP as V-NP. the rules roming the question words Eiftret ways (ee). Splled ‘ut in fl ett these phrase tutes trnefrmatoel als are ‘esi 10 English and to thse constructions. Tere ae few exceptions tothis pase, ‘The PAP sppeaach sins to race descriptive statements (0 t4o snegores langunge-iarot, and lnguage-paricuar. The lngusse invariant statements a pocinesGnchdg the parameters, each on 3 ‘er with» prise of UG he lengoage-prtilae ones sr specfes- ons of patiular yah of parameters. The notion of consrton, it th tational sen, efeively disappears: 8 pera Uf for de- serpiv tone but hat no thereto tats This there ae no sch Py hope comsiructions as Verb Phrase, or interopaive and rave cause, or amie and raising constructions. Raber, thre ares peel prin lesa teract to form these dsciptive artifacts “The parametric pons avaiable peat 1 be quit Tested, An a smpion ha sents ot uealsuc that tee ise ene compute tinal ster hat forms desvations fom DStruure to LP: at some ot i the dentin (& Strutt, the proces branches to form PF by En independent phonclopial derination (es 0). Options would hed eves 10 two ess: (1) properties of the lescon, oF 2) the point tn the deviation (3) from O-Stractore to LF a which sisures ae ‘mapped o PF (Stuer) (Stowell 1986), In the ctegory (pat fom Soasueanariraries an some lan ovary i the choice of substantive elements we Neve options to how nowuteantive (tinal) elements ae veazed (Borer 1984 Fokui 1986, Speas 1986) ane variations i slbel peepee of heads (6 do verbs pra oF Fllow ter complements) Trai 1984). In the eatery (2) e fn, for example, nage with overt 0x rent of questo plzae (Eas, tan, cc) and languages without ‘vat rmovenent (Chinese, Jape, ee). In thse ms languages tri the question phrse inthe poston that woul! be ocepid by a {race in languages wih over movement, there good evidence th simay movement operations ake place, bat ony inthe mapping rem SSiructure to LF, with no inition ithe psa foe isle: the ‘ranch point a which PF is ford rom S-Sirute precedes tse Op catens in te denvaton @) from D-Strctre to LE (Hang 1982, Lani and Saito 1988, 1992), Sin, we Br Iarauges wih over ‘matifesttion of gmmaical case (Gree German, Japanese te.) and ‘ers with vitally no such marsfotaion (Engh, Chinese, ec). But ‘sn, there ocd reson to bee that the ease ten ate bsicaly ‘le eres ngutialy and tat the erence pany in thei onesie ealzation (he mapping to PF). “The gen expectation fr al consractins, & tha langues wil be sey similar a the DStrcture and LF lvls, 36 the example as ‘dacs. us tha there or parame that act efor of EF representation or the compaatina! procs from §Strcture 10 LF: tie evidence is avila 10 the tanguage Kerner bearing on thee ‘ater and thre would be no way fr vals to be determined with ny ‘alibi. According, any varios atthe LF level must eres of DSwrctue parameter settings. or of variations i the mapping from Ponca Parancure ” DDSirscture to SStcture fo the extent thats proper are dtr ined rom iepscon of PF rornsD-Snustore, nt. rect exit proprts these 100 pear be nie in varie insofar as hey alec ‘De computational system. At tbe PF level. properis ofthe language an be eaily beret ed variation is pombe thin the Bred eeper tere of plone propertie athe insartan pricier of iver pb ‘sin SSirntures dre hot eontrsned by interface conics and ‘ary win the ange peste bythe variation of he strat el ‘he banc point o the PF mapping, and any independent contin tot may bold of SSirscure “The principles tat have ben iniesiated al into wo general este ‘ores: prices that ate applied v0 construct derivations (wansforns opal options and eonions onthe way they operate. amd pings ‘Oat apply to repeentations(iceng conditions). The warsormatona ‘operations are movement (adjunction, subsitition, deletion, aed per bape inoerion: we may think of thee ae ratancr ofthe general oper ‘ion Afect 2, arktrar (Lasmik and Saito 1988). Condon of tealty ‘nd cher constrain the aplication aed fetonng of these operations Lens contons atthe extem intro: eels PF and LF estabish therelation of language o other fis ofthe mind/brai, DStrcre ‘sontons specify the mone in wtich Knel propertes are expesed in grinmatiealsrctures. That tere should be S Suucturecondions les obvious, but it sens that they may ost Ge secion 13.3. "The principles have frtbersrctre, These eer groupings into modes oF tnguig:(intng tory, Bxbeory. Cae theory. ce). Cetin nipingeonepes eer ite many ral medoles: cenitons of fealty. somal” properties defined on phrase markers. and 50 en. There re also certain general idest that appear 1 have wide appcaiiy, ‘upong then. pencils of eensomy stating tha thee ca Be no sper ‘Maou symbots in repesnitions (he principle of Full Interpretation, For supefuoes sips derivations (Chomsky 186b, chapter 2-4 of this book). As these pins are giver an explicit Formlstion, they become empirical rypothess with spc mport and range “The prnope Fis assumed as ¢ matter of cure in phonology if 8 symbol in a represcition has no sensorimotor interpretations, the resentation des not quel asa PF representation. This is what we ‘alld the “imteface condition” The sane condiion sped to LE ‘entails that every element of the epesenttion have & (angae- Infependeu) interpretation There an, for eam eno next, m= cher cx vacuous quantifier at the LF eel. The principe of concmy of dei ‘ston reqeies that competstinal operations must be daven by sone “ondtion om Yepresenatons. as @ “lst resort” 0 overcome failure tomeet such a condition. Interetng wath eter peitipes of UG. sich sconemy principles hve wideranpngeflccts ad may, when Mars Fe properly understood, subsume meh of what appears to be the specie ‘Saracter ef pera penile. “The hfs in fcus over the years ker te tsk of inguin considerably sd yield ferent conceptions of what eonstites 4" sul” i he Study of language. Suppose we have sme eolsion af phenemcos in ‘tcl langung. Inte aly sages of generative grammar the task ‘9 find a rule stem of the preted form frm which tee phen ‘mera (nd intel any eb ould be dened. That harder sk than the ones posed in prepeneraie arama. but not an impossible fone there ae many penta role stems. andi seen posse to ‘evs one thst will more lene work though the problem of expla tory adequacy at nes aie 36 noe ‘But this ahieveren, however die, dees ot count a rel resale itwe adope tne PEP approach ao apo, Rather it rerely ss the pro Jem, The sk i ow to shew how the phenome derived hy the mie system ean be deduced Fem the invariant pines of UG wth pre Tes tn one of the permite ways. Ths i 2 far herder and more challenging task I san inportant fst thatthe problem can now be posed ently 2nd seed in ntereting ways in cme range of ease, ‘vith Fares sot af also interesting insofar as Chey po THE Wy 10 taster solution. The departure from the long ard xh eration of Ur utc inquiry i much Sharer an ore rial has ety generative ‘rapmar, with problens at are qste new and prespects tha appeae roising ‘Otter tradition! problems ao aime a diferent form under 8 PAP ppreach. Questions of rypoley and langage change wil be expec in tems of parameter cc (Lighafoat 1991). The theory of language scgustion wll be concerned with sequin of ecu tes, Ka of porameters dad parhaps ratration of principles (Hyuma 186, Rosner nd Willams 1987, Borer and Wevler 1987; Chen nd Weler 1991, (Crain 1991, Pierce 1992) 1 might tre oot that pares re beslly| ‘unform for all languages: he parsers for Engh snd Jopanes wou (ifr only in that parame are set dieremly (Fong 1991). Otier Preps nd Foams » sues would also rear soe sehihing. ith proach tars ot 10 ‘econ, Much of the mst Fit ini into generative rata the pst years has pursued te workin hypothesis that UG i spleen le ‘et heey, With fondamertal rnp hal havea inti character and broad generality. By doling he notion of eoscition and mov ‘ng omar "rule" stems the PAP apqronch cares this tendency onsideratiy forward. A jelaedassumpsion i that UG is “poms “dan,” the sere thet penonens ure expe by interaction of prin ‘pin one prticlr ay. Discovery That phenomena sre “ovedeter mined” has commonly teen ten to indent w torial deiserey that should be overcome by se of refined princi. The working Jypotneses have proven sucemel wt «guide to ngury.faing 0 He Sscovery ofa vast range of empire! phenomena in widely varied le ‘ges and to Torn of explanation that much excad what could be fonerplted wot many yeas age. Thee are eather arpiing fs. THE ‘iin ideas resemble thse afien adopted in the wady of ioxpaic ‘Phenomena, where success has fen bues spectacular nce the 17h ee "wry But lngoge i «biologal psiem, and biog sytem yp cally ate “mes.” iat, the esl of evolusonay tikerns” and stp by accidental circumstances and by physical condition har eld ‘complex sytems wih vara! Tunetons and eons, Regundency | oc anya pial feature of sh systems. Bt an expected nein ha i Indps to compensate for injury and defer, and to acormodate 10 4 diver of ends an functions. Language ese appears © have theo sted proper a noted ie «far fet th large parts of I ‘age ae "unable, he unl parts appt ofr a hoe ad “unpre pled spc ofthe al anguage Nevers, it Bat Bee & ‘nfl working hypothesis ha mts Baie tire, te lsage ae ‘ty as proper of simplicity sn elegance that ave not characteristic of complex organic stems, just ste infin dg character seers Fiolopeaty rater slted. Posably these concusions are etfs re Aecing a pariulr pattern of nga, the range of completely utr ‘lin an eppareny choc plenomer of anpinge lends ced ‘osauchshepcism, Sl. the propre that has ben made bythe contrary stance anot be onerioked “The PAP approach st sometimes tered Goverment Binding (GB) ‘Theory. The teinloyy i sending. Tro, ealy ots to see 0 Caper tent thinking these tems happened to concentrate onthe sears ff government and of binding (Chemsky 1981s), bu these medules of language stand alongside many eter: Cas theory, beer. ee 50 en. ay turn out tat the concept of poserinet hs a kid of unig role, but there is othing inherent 16 the approach tht requests. Furermore, insofar ax the theories of governent and binding del ‘sith real phenomena they wil appear in See Tom in every apoech to language: this arpraach tat no special ci on hem. Determination ofthe mire of hese and ether sens is «common projet, ot se ‘oft thi parila concepion of th ature of language ad sue. 2 The Leicon ‘A person who has a language hes aces to detailed information abst words of the fanguape. Any theony of language mst refs this Fst {hs any theory of anguage ust inode some sre of erica, the re poston ef al iosycrac) properties of partial lexical ies. These Property include a representation of the phonological form of eich tana speifeaton of te yar enepry. and its seman character ities, OF partial interest inthis dain ar te seman see tion aed thematic proper of lecal heads verbs. nouns, adjectives ad pre o postpontions. These sper the "argument sree” of & fxd ndiating how many arguments the bead Seems and what 5e- mantic oleh eee. For example the eb ie most be spied as ‘signing an agen ec, a theme rl, snd » gelrecpent role. 1a (3) ota bock, an Mary have these respective rematc(& os. (6) on grve a book to Mary ‘Te sssacntion etwen assigned Eles and argurem positon isto 8 lage exem predicate, For example, “agent” i apparently never a= signed to complerient Ard to the extent that de association is p= ‘table ater than ionyerac, need nt (hence, must net) be aed in paniculae ea ents. . ‘This conepion ofthe lericon i aed o that developed in he 1960s (@omeky 1965, bu depart from ti certain aspects. Thee, sob ‘cxegorization and scketonal cottons played cata rte. Te Fr ‘ner condone fr ical Head what phase enegones kes a8 emplerente for example, that Kick takes an NP complement. The ter comiltons speci insase semantic features of he complements) Principles nd Pasar 4" and sujet fn this case the NP complement of ik i [ concrete ‘us noted in section 1. that pluse sacar rales er (rey) rn ant wih subenepoizatin, ence sre Carey) elma, But now ‘ot that subeatporizaton follows alms ent rom ble specif tion, A verb wth ne ole to assign fo a complemen wil Po Be alo take complement. A ver wih obligatory) Ooles toes will ave to occur i = configuration with enough arunents (pws inttng complements) to racivethowe Doles. Further, at Hast in prt see ‘ional restriction wll ako be determine by thematic properties Tore ‘te «particule Grae she inherent semattic etuer ofan argent ‘mst be compatible wih that ole “These tentative conclusions about the ergarization of the lion ese importnt questions about he seqsstion of ene Knowledge. Suppose that subeategoczntion (cseecon) is artic is lcs derived from secant properties (selection) Is reasonable to ak whether this ‘sensequence ofthe acqustion protdur el (Patky 192). Peet (Geveoping ides of Grishav (1979) sogets tha this st beso. He mpares the primitives of eselection (ymactic categories such as NP, (CP, ec) with tow of theory (Ypen,” "pes." “peal” ee.) end farms iat the be, but not the former, met hat we my call the ‘nition of epistemotoicelpriovay. That they can play be pled by the earner Io provide a pretiminay, pings anal of eatonabe sumpe of data ane thas can prove the bass for develop en rom the seit site othe tend sat Ths is an tractive Ee of ‘essing, but. piven our crrent understanding of th is, i ot ‘concsve. Whe i doer seem core thatthe primitives of eseectin oot have episteaolopcal rity, i is not a all ler that those ‘of sseection do have such x wat. Although the notion “gent of an ion” is possibly avaiable tothe child in advance of any syntactic vowed itis les lee tha the @aeorese nation “agent of = tence” i. Tat is, before the eld know anything about the syntan ‘of his or ber language (beyond what is ges by UG), ean the ehild de ermine what portion of a sentence coitus the agent Fete, the ‘idence availabe to the learner ely consis of sentences ater than simpy individual webs in oaton. Bu such entenesexpicy spiny celection ropets: they exit verbs slong with ther complements ‘Thus, the cll simultaneously presented with cadence bearing on ‘eth eskection (given tat semences are present in conte, and ws ‘ming tha the revant coments can be determined) and esteton, 2 aps is reasonable to ase that both aspects of the evidence cotibte to the development ofthe knowledye. Alongside the sta of afi outined by Poetsly, he converse stuation with exaction evidence tfc po- ‘ing infermation about te meanings of verbs (Lasik 1990, Ghat 1950) igh lo obtain. For example, expose to a seneace contig {lua complement to sn onfamiar verb would lea! the leaner 10 Iypotheze ha the verb oe of propositional tide, "Tas serio not neces efit wth Pees’ ini pint out the onsnizaton of leucal eri. The mens by which krowledge is arved st & pt insriably refed inthe form tat the Know shimately takes For example. Grinshaw (181) ages ht he seq. tio ofthe syenaciceatgory of leaker based in par onthe rion “enon uct! realization” (CSI), The CSR of hyscel jects that of un action is Vand 30 0m I the sbsener of eines ‘he cid will assume that a word belongs to ts CSR tat 2, a word refering te am actions verb. AS Geisha indicator, wile Such “semantic botsrapping™ might consti prt ofthe ecqision po ‘eure, the resulting sca) state lexicon bas no such requienent. Lat fies commonly have aouns, ke dracon, fering 0 ato 8 elas verbs ike, tat dot eer te aetcas) "Note that ths consideration indicates tht lea enc contain a test some syntacte information in adton to tbe phoologe! od ‘sea inflormaton that surely ust be preset. Grins arpues that farther syntactic seication s needed as well, ction i addition to saelation. To conser ope example, Grimshaw observes tha sean category "quesion™ can be srutualy realized 85 either ‘dase, asi (6), 028 NP, a (6) Mary ashe vba imei as) (7) Mary asked te ise] The vec ack sentially svete» question, Grimshaw argues thai is alo necessary 16 specify thar it estes dause or NP i order 10 ‘isting from wonder which nly takes a cause (where ® indicates ovine) (8) Mary wondered fat imei wo (9) *Mary wondered the ine} Smee as sugested above it spose tore mest of eaeation 10 rcestion, she question arse whether such redichon mighesonatiow ‘be sill inthis instance as well. Psetsky argues ha ii, Ase wil ser in section 143, NBS must esive aract Case fom a Cate signet ‘wile uss need ot. Hetecorth, we wil captaie the word Casein isechical use J Given tis, Pesky propos tat the difleence be ten ask and wonder ped nat be stated i tere of easton, but rather follows ffem « Case difereoce: ask asens objective Case but ‘onder doce not In this rept, wonder pater with adjective, which ko do mot aespn objective Case. (1) Marys uncrai ae ietis} (1) *Mary uncer the ins} Pscsky presents Farber evidence fr this Case asiging dstinton be ‘ween vets lie wk and thse tke wor In English. generally oly ‘objective Case-ssignng verbs can occur in the passive, Given this (6) and (8) contrat in precy the pated fashion (12) st as asked whan time was (13) Si yas wonder what tie twa ‘As Peseshy notes, descriptive generation pine! au by Grimshaw now follows: among the verbs tat sselect guesions, seme eet ‘se or NP site otberse select ony lave: none case only NP. ‘There are Case-assning Uiferenss among verbs nd hese a relevant to esskction of NP (because ofthe Case rouirenent of NS), but ot kas, This reduction sms quite succesful for 9 wide range of cases, but it is important to ote that formal syntactic pcan in tenes! eres hve not been emily ebininate in favor of semantic ones. Whiter ot Dota verb asigns objective Cae a8 far ai known at preset 4 purely formal propery et deduce from semantics. While much of ‘celaaion follows fom sadetion, there ea rytzctie vsidu, sata, it Pesky is comet. in terms of texaly esynerati Cae properties We will introduce further properties of the lexicon as rguied By the 13 Computational System 12.1 Geneal Properties of Deiticos and Reprseatons ‘The generative procedure hat costes the I langue consis of & lexicon and 4 computational system, In teton 1.2 we oud some 4 Chaper proertie af the lesion. We now ur ote compusational system. Un ‘er the general asimptiors of sceton I. we conse the four levee ‘of eepesntaticn of the EST syste and th relations that fold among then, fects atti oe “harow stan” that the eration lating DStuctre, S Stuur, and LF DiSirucue, LF, and PF are irae eves, which sty the general ‘condition Flin manner wo be made practe Each Inel in a ombobc fysem, consisting of atomic elements (pines) ad objets onstrated ftom them by conctcration and eter eperasons. We ake thee objets to he phrase markers inthe friar sense epeseted conventionally by wes or labled bracketing): Each prime a eature compe, though for orthographic convericnre we will gntrally use conversion sya bole For conretena, take eatepris to be asi (14), for noua seb, jetves and pre-and pesipestions respectively 09 & NeEEN Vy) bVeI-N a) © AM[N AY] 8. PoEN VI ‘The feature [+N] ¢ the wadtionsl sbstantee; the Rute (+Vh sedate, ‘The primes consiuing the termina ining of 4 vase marker sre ‘dawn frm the lscon oes at projected from these hea by opera Sins of the computational system. Elements that pet no farther ae ravine projection: tn inleral nottion, XP the maxi projection| {om he terminal eatery X: tha NP is he mania projection oft ad Nand 0 08 So ston 1.3.2 he io base relations of = phe marker are domination ad fn- ‘eariy 0 te pase mater (15) we Say that B dominates D and E, © ominats F and G, snd A dorirate al ther ealesories (node) Fur thermere, B precedes CF. and GD presses EC, F, ad Gr ad 20 09 A x oN ZN, Cy £ S icp sn Parr % TEX is 4 ead. is “sister” i is complemen: ths, iD and Fare heads, then E and G are hr comriements in (15), We ssue that or ‘eng eclations are determina By a ew parameter stings Thus in English. ighr-vonching anpuae, 0 heads precede thet complements wile i Japanese. lef tranche Language, a hed follow thir comm pHements the errs determiged by one sting ofthe head preter Eyamples below tat absract frm partes Inginges ae uly 13 te interpeted independently of the order given. Domination elatons sr determined by general prices econ 13.2), ‘One findamenal eonzpt that applies thraghout the modes of grammar is camend (Khia 1964, Langacker 1949, Lasnik 1976, [einbert 1976, Stowell 961, Aoun ane Sportiche 1961. We sy that 9 “ccommands Bi «does vot dominate and every y that dominates 2 ominates 8. Ths, in (13) B ecormmands C, F, G:C ecommiands B,D, E.D ecommards E and comers, F ccommands G and convey ‘Where 7s resrcted fo manimalprjeton, we soy tht command ‘A Second fundamental concept government (Chek 1981 1980 Riz 1990), 3 more “ca yey f command to wich we retin seston 141 ‘Given ie language L, cach SD is a seqoence (8,0 hese being ras markers dawn rm the lets PP LF, Struct, and Struc lure, respectively. The creat 8 recs properties ofthe ems sete fom the Fescon as these are interpre by the prices of UG, with the parameters fed for L. The ements @ tnd 3 ae formed by suc ‘ye application of operations of te eampurationsl tem to 8 ey will Ine he properties of 8 8 modified by thee opertions. The PF rque= ‘enition x a ting of phonetic pres ith slic and intonation ‘srocture indicate, derived by a eemputaion from o. We assume tha the primes themsees ae not mes in the course of the derivation from 802, ‘Apical exe entry consists of « phonological six sod other feature, among hem the eategoal features N, Ven s0 0m nd io tne case of Ns, Case and aprezment Fears (Gerson, number, gene) ‘znetorth ectes. In piel. any of hese Fates may be lacking Tn one case of particle tee, the entre pons ti i ck ing. this case the element is am EC mp entegory). Atsong thes Cs we have the elements # of (16), (17 we ase * to indicate seers evan, 7a weaker aie Ps gir (16) John expects [eo hur ise) its commento hurt ones] 07) earned yesterday (he anived yesterday") We refer to te EC of (16) as PRO, an element that can be conrollad by ts accent Ua, i (16) oF eam be aba a itespetaton, 35 in 16). Pnsbly te ater is alo a case of contol by an EC eccurying ‘he same postion as fru in (18 (Epstein 1964) (08) is convenient for us for others to othe har wk) so, PRO's alvays conto. See section 142 ‘The EC of (17s a proserial clement, tenefoth po, Is not pe ‘ted in this position in English he counterpart woald be rama Sn Mins, 2 ml subject lavage. On factors relevant to tig the pe rameters see Riz 1982, 86a. Huang 1984, Borer 1984 Jeg and Safir 1989, This BC ute mich inthe manger ofan ordinary pronoun, Irving reference ine by context or by seme anteedent in wn approp te position. The ruc! Felton of antec pr) pie se, fr thermore ereral tke tow of omeedent, pronoun) ana nthe those ‘of congo. For example, na ell subpet language we find the eau [ent of (14-6), analogous to the pair (18-€} eb taken to Be the anced of pro, (19) a he people thax pro taught amie Soha . ‘pm adiited Jobe the people hat he aught admired John “he adie Sohn “The behavior of pro and eis simi, while PRO can ever appear in thee postions 'A thd ype EC, not denen from the fesicon bat crested in the coun of x Sein, tated in 20, (20) 2. wonder vo Jobin expected fe10hurt inst} dob was expected eto Ha msl] Wie refer to his EC a rave (9, telational neon mace-ofX, whee X Is the moved element seving a the antecedent bd the ce. Thus, Jeb Winds in (200) much 25 ebieds the reine of 35 ehey ids he ‘espa ia (21), ie turn binging the refx. 21) they expecta each otber to hut themes] Principe rd Parmeter ” In 0a) the Wace ofthe NP who. Te trace futons aa variable ‘ound by nha, understood se restricted quantifier: Tor which e, «| person. Here © tira binds hi uns each oer tnd thoeles 4 (21) and il binds hin iv C2), with Bul substiaing fo the ve able of) (@2) fot expect [Bi tout hist] 1 Qu) both «and Aone ncn ws variables bound by the ested quantile 0 tha the LF fon would te ioterpeed “1 wender (or ‘eh cc psc Job expect eto hurt ef. Note that ear ing the erm find ere vo cover he assecition of at antecedent wht ace ‘ule geeraly ncn he case of he (sya) bing ofa variable by aguante ihe elena we aso we heer, at Ln the sense ‘oF quantir arable nding In (208) the ve-b war is composed ofthe lene element be and the Jafletonat elements [pos 3 person. single} Assume now thatthe roces of competion niin the copa to the nfltional eres (ing. Real that there are two interpsasions of this proce (I) rasing of fe to the inflection positon of the setense Yo coberect the combined form [+ eflering. oF (2) ring of [t+ intons] (vas, dra rm telecon wth its fener ated assigned) to the Toficon posion, whee the features rechecked. she way, we have & ‘cond ace in (200) = (23) (23) del was expected, to hart hist] ‘The EC ey isthe ince of fe oF war: ci the ace of Jon, binding msl Ia each ese the tac decpies the poston from which sane ‘dent was moved. For cincretenets, we ado the checking theory So that we ave wasting in 3). Raising of wa othe efetion postion is neesay to check ees onal proper. The see is tre ofthe other ifsc ves fo xa ple, wand (2). whieh spree I person. singunr Ths, 9 ler (thoueh sil oni prt representation would be (24), whee ei he tre of wonder. (24) T wonder who Fon expected to burt ise] “There i reason 1 Believe that in English (24) ean LF rpreseeition while the cousterpare sn other sma anguagss te, Freeh) is an ‘Structure represtation: (23) and is courterpars are § Struct ep rsentations in both kinds of lnguape (Ernonds 1978, Pollock 199) » upte “Thus, Engh auniares reat S-Strotue but mn verb rise only CLF, while the comesponding Pench cements all aise at SStroture English and French would then be identil in leant respects DSirvctore and LF, whl dftring a SStructre, with Eopieh (25) (coresponine othe haul shared DStracure) versus Frese (25) (conesponding tothe Bascal shared LF for). (25) «Job olen ses Mary) by Jean embrase souvent Marie Sean se often Marie leformaly, dhe ace functions througout asf the antecedent were is {Hat postion, receiving and aspring sntactie and sete properties. ‘Thus, es isthe normal poston of the antecedent of a refenve both (2a) and 20). Ande 25). che race isthe ert head of VP, as: ‘png parcla emai vole and granimatial Case to is sonal objet ‘Note that PRO and tae ate quit diferent in ther properties. Thus, ‘an dament tht eorzos PRO isan iadependant argurontn the sense of Section 12, esigned a independet setae ee: bu an element that binds w trace fot Compare (16a) ané (200) peated hee: (26) a, Jobmenpect fe to hur ire) 2. dob wt eapected feto Bur hist] 1 (2) Jon isthe sujet argent of expected xsl ain (22): he EC contrled by Jal as s independent farcton a sbjat of art 1 (2G), in coast Jon has no semantic fle ther tha what i “inher” from its tac, as subject of fr. Since the subject of 6 “espero s asigoed no independent argument rec bea onan tment (an ele) 22 2 (20 theres expected to beam ti tomer Other dierences of interpretation fellow. Compare, For exaene (28s) and 2, (28 a. you fiends oped [to ish the meeting happy) 1 your frends sered [to ish the mec bps] J (8a) your fronts and e ve independent aeumenis,asdened thie serantc roles assets of hope a fish, espetvl therefore, the EC rust be PRO, enti by yur re Bu over as 0 sea ‘nero ot subject whic can agin beam expletive a8 5). rips aod Parton, » (29) ise your reds isha the teeing hp] 2 thee seems [eto te mistake in your argent) ‘Aecortingy, the EC in (258) ust be trace, wih its anedent jour nents tesving es semantic role a8 an argument ata were sn tat postin. We know further that the adstive happy modifies the subject Fis own clase, not that of « higher dause Thos i (30) Mary med meeting ot your rend: the sentence mens that your fends ope that che snospere would be happy when the mesting nds. (20) your ends hoped he meeting would Bish happy} tn (2), the, dappr modes PRO tn () a trace (6) Example (248) ths means that your neo ha crtain wih: hat they would be hap pj as he mctng ens. Bur (280) ha ought the means of (2). wth app) medying your ene Otter dierences of measing also appear, as in (a) and (316) urae 980), (01) & ene ransator ath ws expected / wo be signed 0 the ising diplomats >. cue wanstor each hoped PRO tobe asgned to the wsting Aiplomats 4n Gla) acer ene womr ach or its ae sf 3 postion with Independent argument satus, Therefore, the argent phripe ane one lave och interpreted as if were inthe peton of the tae with te argument satus of objet of apne: the easing that Hwa ‘expected that on tramlator each would bead to the visting it ‘as (ie, enc diplomat would be assigned one Wansaton). In (3D, It ne motor rach aed PRO ae indepenert argues PRO, not one transtrer each, that Winds and is ieerpected a were in that position The sojer one rant ed ish Be with 49 Jnverpetton very mh as tr in the sma consruction (3). (02) one wansatr each hoped that he would be asgned to the visting ‘ilomars Alibouph the argument status ofthe anacent of trace deter- red in the position of the tree, the antecedent may stil have a8 independent semantic roe in other espe, Compare the examples of 3, « haper (63), ‘tse to each eter lat your frends ae happy] 1b your fend see i esch oer tobe Ha © itees hat al your ends have ro et ried {all yor nls sem to ae not se are In (3u) yeu friends can bind he repeal each cer, bat it can in| (630) ths fonconig in its overt poston, et that of its wace: Tn 33) nd (38) the oven postions are revan for determining ops P| {ie tha, only (35) ean mean ta seems that nt all 04 end Fave arived, with no aking see over all, We se, then, that sco property and arpument satis are celeined be dillerent ays fe nteedenaraceconsrucions. Sch fics a these cught (0 fl ut 2s consequences of the theory of ECs and semi ioterpreltion. See section 142, PRO and ice alo difler inthe sytacti dsrbation, Thus. in (34) we sce the properties of contl ith the sntcrten snd PRO funtin- ing as indepndert arguments; bur the preperis of tae, with only foe arpuren. cannot be exhibited inthe analogous struts, 28 (3) irate. (28) a. ob ake wheter [PRO to ese) ‘akin expected that it would be fin [PRO 10 vst London} (68) a, thot was asked whether eo ae] John was expected thar it would be fun (vist London fe fe, ace ad PRO do not overlap im thir dation: the fet sould, sin, al ut ofthe theory of ECS ‘We aho allow fourth ype of EC, one hat has only the eteporiat Sestures 4, £V), projecting the uml wa. They serve ony at fa for movement oe filed in the courte of demotion, Sine thee ements have no senate role, thy will not saisy the econ FT UD Strecre (as we wil sharpen thi blow), and we may tentatively ‘sume tht they andthe srtres projected fro them are ised in the couse of éereation, in a manne prt bythe theory of phrase struts See tien 43 fo farther core Tithe kinds of EC are inde dinint, han we expect ser to die in featsre compostion (Chomsky 1982, Lasik 198). Opin the {estres soul be st those that chugs over: cements AS 3 HSL sppreximation, suppese tha overt NP fal into te eategoves ene Poi ond Peas 4“ eerie, reciproeals). pomaun, aud rexpression atm, the rina ‘Square root of 2, an the expresso that are "quatre in the internat sense of scion 11). We might sue, then, shat we have two two-tlued features, fanpher! and fponominal. with potentially Tour catezeis (06) | +naphor, ~pronasina 1b (anepton + pronominal) [-anapher = pronominal] & [anaphor + pronominal ‘Au expression tat s+ anapor functions eférently ony in intra tion with ts aniecden te reference of a expression that is (+ promo minal may be deterine by an antecedent but cs rer), Retenies and reciprocas thal it category (3) and pronouns nto category 166). The tid category contains clrents tha refer but ae not refer tnvlly dependent: neapeesions, The four Ee discs above would Te she ypotegy of 37. (G9) & Trace of NPis[-+ nophor, —pronemirall 1. Frois[—araphor. +proneninal. © Trace af operator (anal) 5 [~anephor, — pronominal 4 PRO'S +smaphor, 4 prosominal) “Thus, te Of NP is soneferets, ro hes the properties of peo oun aa vanes are “referent im thi they ate pct foe expressions. Cooled PRO alone category (7), here all PRO apparent uncontrolled PRO actly hss a hidden conto (e (18) ‘We would expec, then, that race of NP ro, and vable wont share felevintproperics of overt anphors, pronouns and rexpresons, Spectvy. Soc ements as English en, French am, German man eight ‘paral overt counterports o PRO, sharing the modal interpretation of abirary PRO and is ressicuon to subject postion (Chomsky 1986) “These exputations are meget satisfied, when we aback aay fos ‘ther Fairs. Tha, he stretra lan of eet it terse ‘nsicaly that of an anaphor ots antecedent: bath ea the arsed fn mus command the race, and oer sretral eons mist be ‘s,s asrated io (89) wit the examples hep sighly dierent Yo vod fotos that br the urate rote, 2 Caper (8) 8 Job ur ims fi John war hart “bine housht Hoh sams to be inttigent i. “rahoupht ohn ses that its rain 6 ob dei inst ef eal) 1. ob ak died ft eve er) “Thee properties sharply restrict the options for movement of NPs rising not lowing. bjec-to-subet ut nor conversely, and 50 on (engo 1977. ‘Siolar but aot que idenical conditions hold of PRO. Ths, she ©-Command Condition iatated by C9). 9 eon expcts[PRO of fms) 1. “o's mother expects PRO wo hus hms} Sob expects [PRO to el Mary's voter about Reet] 13%) PRO isin poston to ind raf But the Command Con tion requires tht its antecedent Re Job, not Mary. Sires, varies share relevant properios of expressions, as expected (40) 8. They dhnk od wl eave tomo} iT wender who they thik [wil lene tomorrow bk sitseens Hot tobe itelizes] wonder who i seems 10 be iliget} ‘© hedhiks ohn sittigns) fi Tere who he hiks is iteigent} Si John thinks Peis ineigen) '%. Twonder who [tins fh tigen Go (40a) she name aod the viable appear as Case-marked subjects of fre das, andthe expressions are well Fr, satisfying the Case _markng edition on rexpresios, © which we return diet Bo (08) the nate ad the variable pear es wb of initves lskng Cs, land the expressions ae severely deviant. I 40) isnot feel ound by Jl (we cannot ake oreo Tok, a we may (0) and in the parallel svuctre (4c) he nd the variable 7 are unrelated ‘ferential (we cannot take et bea vasable bound by te opersor sto, which binds «35 we may in ei). Agim, many conditions om ‘overeat fal out a pci aes, rips ana Parte a “These ECs als have ote features of over eapressons, epi, fetes Ts. the eae in (2) as thefts masculine, ga, ence the choice of overt sao. ’An EC lacking the typoloisl Feats of (37) or etre i i serpreabie. bene mperissible at LF by the piste FL. Such an ee ment denied only by its eatepoil Features (NP, Ve) may appear ithe cous of derivation, bat ony a6 poution to be ied o eet ‘wi eliminate. 11 an open qustion whether movement alvays lve «ice, and ‘wher whan it dows, thve ar independent reas for thi. Fer the putpeses of exposition, we tentatively assume that moverent of en ele nena lays leaves trace an, inthe splat cat, frre choi (@.2) where othe head of the ein, the moved elements i ie ‘nace, The chain sun char its ea asthe property X: we ern to ‘elevaol choices oF X. Te ements sujet to iterpetation atthe iter face lel LP are chains (Sometines oucrnenbers). cach ah abst presentation ofthe head of he cin, “The movernent operation henceforth Move 2) ia invariant prince fe of computation, stating tht x catepery can te moval to 4 tree positon. We tate tbe moved enegory and the target to be primes (excel items, EC tarps for movement, or potions fom the mini ‘ements, ith we options her the awed eager a races the target (substitution). of # ago to (adunction. sei (41) (ower relevant, rte tac of Band, te cccurrens mA . Any further consents on moverient wil be derivative fom othe pine ‘Pe including eondiers on representations ‘There ae to nstura interpretations ofthe eaments fred by a jmeton: we might assume tat ech ecumence of fin 2) a catenory ‘io Ms own sight (Lasik and Soo 1992) or that together they farm ingle eaegoey (BB wih the two occurrences of as its cps “ capt (ay 1985, Chomaky 19869), Eira! direnes Follow, a usu. fuer theoreti stucture i arculate “The sepmen-caesory dstintion requis a sharpening of the cow cept of dominance and those drive frm it (command, ee), Let vs ‘ay tha the eategey I. Bi 41) ches X, exces 1 and coma ‘x tand whatever i dominated by these element). We esr domination to inlsion, Ths [fu dinates ony X. We ay that segment oF ‘enlegory 2 corer ii cota nes Ror =P. Defining the cor trand eations as tfor, 2 ecommands 3a (4), sce it wot om ‘ott (oly sontsined) by bot ¥ ince in 9 des not. We eur ‘ove the poperticn of bead and command tothe pestadjunetion tras tre, Thus, iy was the beud of the predjecton category and e- ‘emis 8, then in the pstaduncion sete [Py Bah ¥ semaine the ead and ecorimands 8, Where ne confsion ares we wi feo ‘he postadjuntoneatepoe BB simpy es ‘Sibsutuion x consraned By 8 UG principe of recoverabiliy of dle tion, bch egies tat no iforsation be lest by the operation; hus 2 say sbatiae for only tere & 0 feature confit erwsen then ‘The tact of svn wil hetcore bays be am EC with he sme tego features st the moved category (te srctre preserving hy othe of Emonds 1976. saisr property hokts for adjunction, i oper se Sean 13). ‘Mowe permits mukiple (steces- sive rom he D Structures (43). with the argets of overen inserted (42) Jol seems [to hate bor expected 1 eve] Tender who Hon toveht Bil expected ft ewe) (43) 4 esses to hav ben expected Jor to ewe] . T wonder e Sohn tought eBilexpected a ose Io 2a) we have te esi Job. Fit he ks Ulin) an in 2b) the ein wha 7 ao wih wo irks. The ead ofthe hin re al ho. respective ‘We have so far sued thatthe operation Move @ forms a singe link of chain, Asatvely, we might ase bat te operation isnot Move 2 bar rather Fort Chai, an operon hat fr tefl chains (oF 2) from the Structures 2) 4 ingle tp. With her theo reveal context. the dsnction mayb moe than merely notational (Se ‘apie 3). We tenanively aoe the more conventional Moe inter retain. The operation Move» sais narrow lcaiy conditions. ‘Soppse thatthe postion of he intermedi ee #82) ied. a in (G4) 50 that he ehain rust he formed with a single ok, skipping the ‘odked poston (ocupied by i, wheter here. respective). (44 4 oto sens ht was expected [tere] bb eka he John remember eter Bi ed {© "how dd John ember [rer Bil ied the i “The ehns ohn. (shat, how, A) wae the fealty conditions, and he expressions are deviant, hough stingy diferent ways fat ‘hat demand explanation i ters of properties of UG. Note that nese e) i the BF for ith ths nrpetaion that yi how Con ‘raed inthe postion ofthe taco that deat if how is sons sith remeber ee no devine. The singe PF form has to diet ‘Ss, on sharply devas the ether no cal that each element rust have wifem, anguageedependent interpeciation atthe trace leet LF (he pip FD), Some Cements “re argumens assigned specie senile colts (Coles). sich 8s gent ld goa (Se section 12s over anaphors, PRO, and rexpresions (i= ing varibis) are a arguments. Expeves ea Empl there Hak inne ae axsgred no Gels. Some cements (em Engh, French Iteian gro) may ambiguously serve as apumens or expives. By Fl, ‘xpiciver mst be somehow eave wt LF Geto 1.33), ‘As argument must recive oe fom «bed (@ barking) Am argu ‘ment may also recive a semantic roe (wher be eonierd a ole ‘oF not uheorierea question that we pu ade) by prdeation by an XP (ee Wilms 1980), possibly an epen selene (the reative ‘use OF 69), ith variable postion 0, (45) the job was fee to Mary [who everyone pred ha the best ‘ushieatons] ther Ps (ajc, such a veri pases aga semantic le 19 predict. a ead o another june. AS stested i (eb), move ‘ent of adjunets and arguments hes quite dileens properties (Hoang 1982, Kayne 1984, Las ad Saito 1964 1952, Aout 1965. Rize 1990, inqu 199. postions position to which w Pole sane. ‘The lemenis seeing iteration at LF are chins. Hence cach srgurnet tain (46) ms conan atleast on postion 46) a5.) “ chop Fusthermor, oy he postion aecupied by a) at D'Strcur, mis! te f Gpostion, The etn les in the interpretation of D-Srvture 35 1 ramimatial retzation of kcal properties. Accordingly, Oamarking trust ake place at Structure: an element, moved er no, wil hve at EF cuacty the Oomarking properties (asgsng and ress 0k) that it has a B-Siractue: From the same comiertion, eDOws tat toting am move into Spontion, gaining» Bole that was Dot 8 Siena to iat D-Siratre. Thus,» chon car ave no sore than ene pesto, though any ur of semantic roles maybe ashe in hs pono, fn(4) for example the wall cies 2 tric ole fom both Pot a ed (40) e paid the wal rod “Te theory of Case (extion 1.43) requires hat eer argument have rac Cae (goeubly reaiaed vey im one oF another way, depen ing on spctie morhologcl propre ofthe tnguogs). ese. ana urent cain (46) mest have ne and only one Opsitin (amy) find at lt one poston im wtch Case is asiged (a Case pesto) FFtoing Joseph Aoun, we might sink of the fmion of Case as 10 snake an argument hain vale for Omerking The Last Resort condi tien on movement (ss section 1.1) egues that movement peited tony to satisfy some conn, in prt, to ati ity (hence, FH). Once an element has moved to » Case postion, it ean move no further, al tevant conditions no being satisfied. I fellows the, tha ‘ery argument chain ust be Heide by Case postion and mus ‘inate ina postion ihe Chair Conon). Note that these conclusions held only for arguments othe han PRO, ananomaly to whith we return section 143-On the tans of ins eae by expletives ith regu othe Chin Conon, see section 1.33. ‘We have ofr considers chine that ori from an NP argent postion of D.Stucture These al no the two types usa in (2), peated bee (48). obe sens to have been expecta to evel 1. Twoeder to Joti though (Bil expected oes} tn 68) we have, among others. he arguzent chain en 1) and in (42b) the opernornanable chain ho.) ‘Chairs may she erga frm non-NP pasitions. Oe ese, areaty inenioned, he riovemert of lec eaeyory Ged movement), an Proiglerond Peete © (22), repeated here ilustraing the sing of V to the inflectional postions (49) 9 Jobo was expected ro hur seh 1. Iseonder wo John expsted our its Here we ave the chains (rx. 1) and (son), te ter an LE chain for Engl Hed eovement is also involved information of compound wont i szany langges Suppose we were to form a caumtive verb meaning “Gusto al fem the underying D-Suctar (St) by aehoiing fll 0 owe ——_ i ee eae tothe V i ca “This operation yields the strata) the ace ofl ob ‘See Boker 1988, ere came is the head of @ two-sgment ves cate 00, if we assume a segment hey of adnction. A sezand kindof ean ciginating trom & noo-NP postion arises {Gon ovement of nonrguments (adores pects) a in (52, ($2) a. {to whose benef woud that preposl ber 'b. Mos: carly does he eae to theca ist England, Reever ot? [assur as Mary dont think hat Jn wil eer bee « ‘creer In cach cave the basket! ronarpuet isthe antecedent ofthe trace the ebting, then ae (Uo hase en (ow cars) England) le scsi ax Mary) respcovey. The quesione ele ‘nen in (528) ely whos the est "eased along” beers ho cam ‘ot be extracted from the D-Suucture position (53) pet-pping Ros 196), (69) shat props woud e to who + POSSESSIVE benef “The nasal interpretation reas the D-Strcture Frm; the meaning is “or wich person «that proposal would be ts benef” ‘ence thatthe LF form should inde be constued In someting Re this saaner ee seton 13.3. Case (526) might be ierpreted ilar, ‘ha the interpretation would be Tor what degre x he expec fx the ar fv earflly We tughy, then. argue tht these are not really cass St movement ef adhnet pass es sch, but rather ofthe question ee- rene wh, nt withthe adjunct pease carted along. We might con ‘he further that operator movement i the only kid of moverest 10 ich ade phrases ate subject, uke arguments, which ea form 3 fimeht chins The concuson twippered by the cservtion tha though adjnets can yp appear ss may sentence postions, thy fav not interpreted as i they had moved fom some more deeply em eae poston (sito 1945). Ths, (Si 201 gen the interpretation (oF (58), as it woud be if crf im (84) ad een moved from the DSirvetere postion of wei in (8), (58). carfly Joh tl eet te ear 1 Job old met he the ex ently) “This gets tha (52) might also be regarded a kind of pit ing ith the moved element how carping along the arse pase how eae ily, Se capers 3300 4 ‘Within theory of empty estepres ane chaise ca ete othe vesion of erection of nell relations rise in scion LA AS ota there, such questions ae obscure at Pes, and asne even more ‘ube under the wsumptions of trace theory” Consider again the S- Struciee representations (42) deed from the D-Simtur represents ‘ions (4) (repeated here) (65) x, John sems [to base Boe expect to evel 1. wonder ho Jobe though Billexpete (0 evel] Penal aed Parancters a ($0). eseemsfeto have ben expected oh to ens Tone fe Jon hough Bil epected fh 0 vel We now ask whether (550) are derived from (560-b). respectively, by movement of Jl, hs or water D-Strctar is deived from 8. Suture by algorthn (Spore 1983, Riza ABN, so tet DStraure is infect, derived property of §-Siructure: or wher thee simpy ' pondvctional elation betwen the pared expresions. Tse ar ale tative expressions ofthe retaion between $ Strate andthe icon. [A thee approaces are “aesformational” inthe ubact sense that they consider a eation between « “spaced element” and te potion in which such am element i stndorely interpreted, and in the case of (55) the postion in which i wou be over a §Stoctr inLguages ff he Chinese Japanese variety (e sttion |!) Sach dapacemet re tions ares fundsnenal feature of hua lage, which must be cap ged somehow. Apparent iferences ance slernative frmulions ten disotve. ov iui. to notational questions about how this prop ‘xy eaprese sila questions arse with par to apparent die cs Deticen “nulilee” approaches and “lee alters that tae plobal props of phrase marker in compen sito (Chomsky 195t, Herman 1863, Gazdar 191). 18 the present as the empl di ‘ingisabity ofthe approaches tans om ily thon-itermal eons erations We wil continue to adopt the derivtionsl approach of section TUL We asame that thsi at root,» question of tith and ft, ‘hough soe one ‘To se some ofthe problems that ais, conser he losaycontions fom Move a. A perl eoeiton ists in (4) thatthe target of ‘movement must be the closest possible postin, wih varying eft de ending onthe hind of movement involved. The condition vey st For ead movement, which cannot pass over the eles ecormarding ead (the Hend Movement Conran (HMC), «special cas of tore vert principles: see yecon 14.1). Ths, (5) formation of rom ‘he D-Sinture (ising lo the clase ntl postion, sats he HMC, but raising of rea to tis postin, erosing the possible target sition cocupied by wil ilstes the HMC, king the sharply evant Imerrogatie expreion (570) 7 & John wit cod the book 1b dot ead he hook (tread Join wil the book » apie But the teaity eons expressed in the sephysep computation ‘night ot be dry expressed tthe ovnput ees That servation tray stisy the HI an each step, but the cutpat may appear tn (ate that tbe condition is seated, Consider again the formation of @ “ausative verb meaning “euseto-ful By adining fal 10 case, 28 in (Gih Reza hats verb mut also be ised to the infection position ence, he reuly formed eategory cave must now ase this pos tion, forming the nuctre (8) (bere TP is ense-headed phrase. 15 thea of al, ad is tbe ace of eae fl). oo ee Here me have two chains: (afl) an (fal, 1) Each tp of ean formation sti the sre ecliy condition. But che resuling ein fended by Jal doesnot Inthe S Structure, the cain (fall) wots the FNC because ofthe ietervening bed a possible tart of move- ‘ment hi skipg™ by the chain. The fora soul hus be 28 devin I S70), bat iis wel formad. The lest condos are ssid ey ‘ssi te derivation, but are not sats y the ouput cain. Moa Chtom requied ter nonderivaiond approaches are net ently ‘srigiforvard 132, p-Stasure “The computational sytem fos SDs that express te basic struct Tacs oymtacte phonological and semantic) of the language inthe orm tt phrase marers with tena rnge draw from the exon. We ate SSvuming the sch properes of tara saguige a “placed elements” fre cxpresed hy multiple represtiona eve, cach spleen ‘cps Pacers and wit simple operations suchas Move eating thr. Fach fve ‘captures certain atest specs ofthe fll comple The Ration of ‘he computational system othe lesion expend a he ater ser face lvl of D Structure. DStructure 8 mapped ¢© LF, the interlace ith conesptual and performance systems at sre point (Stricture) Fethaps varying somewhat from langusge to lnpsnge the deivetion| “traches” and an independent mapping (phonoloey) form the PF rep resnation that provides the interlace wth the setsonmotor stone S68) ‘The east atempis to develop generative grammar in the modeen sense postulate a single lve of syntactic representation, fred by rules ofthe frm (9), where A iba Single symbol and X, ¥. Ze stops (and possibly nul) Ss the desensed nits, rd hee fs set of dina terial symbols that ar ten mopped by oer rls to phonetic fore (8) XAYxZv ‘The symbols were assumed to be comple, consisting. of two Kinds of leer categoria und sructoral Categoria mets were NP, Wad 40 on Sieur elements were fntures that cde global proerics of | Die markers for example. NP-VP agierent in the mam ae here 8 ‘ded by the [+ phar} ature asiend to § and “nheried™ by NP {0 VP through application of the ule [+ plea» (NP, + pal) IVP, + plural (Chomsky 1851. Subsequent work “factor the com ety imo two component, reiting the symbols te jst thet cae oil part (phrase stucture rks forming phrase markers) wed wing ‘ransformations! rues to express global properties of expressions (Chomsky 1975, Lexs 1963, Matthews 1968, Kim 1968). A ftr step festrced the recursive part of the generative proce to rules of the form (59) apd seperated the lexicon Iran the somputational sytem ‘(Chomsky 1969) This pronded » wove syst: phate acre ‘ules and Yes insertgn form D Strctize and tanto. Form the derived phrase markers of sufoce srr. thn subjected to pho te imerpretatcn. The Standard Theory assimed urbe that ony D. ‘Structure are subjected to smanti interpretation, a postin haborated in Gonratve Semantics (Lakoff 1971). Tee Extended Standard Thor (ESD proposed that eure stwture determines eral elements of ‘Semnantiinterpreation ackendo 1972, Chomsty 1972) Laer work Jed to the fouclevel conception of EST ctlined eal andthe PEP pero, which dents entirely with rule systems fr partir tn aps and parce construction. ‘Seporaton ofthe sion from the computational system permis Si plifention of he rates (9) x comets, with X.Y pu Thus intend [ST (59h we have the content re res). a Aa=Z, ber Here A, Bate nostemsnat symbol, Zs # sonnll tng of monte rival sbols or sratnmatial formative, and 1s a postion of lexical Irseion Bie a nonbeaching level caepary, and Z eontains at mest te lexical xtgory. Z of (8) i therefore as in cite (1a) of (18), thee Css none eateory X and Y ae strings of nonexcl eate boric sd Liv ee ep, (8 AAG bASxLY ‘These moves expoied the ere eeduntancy in phrase structure rs sendy disused (etins 1, 1.2): the form of Z is (0) depends on interne properties of lel tems. Further fedudareies ave also imine ‘haely apparent In (OP) the properties ofthe lena category B are ‘comply determined by the Haical element inserted in / Considering the posable forms in (6, we observe further ha in fb the properties (fA are deermined by Le thas, Lis N, Ais NP, i Li VAs Pr lind vo on The rule endcenri, wih the fina L of the eoasrution| profecrng ie dominating eatery A Suppose we sume that the ues {cla ae abo endecetre, aking A to Be x pojetion of one ofthe Cs Tan exprenion of des developed i strc! fngutis ines of onery proctor of content anaes (Haris 1981). We now have ules ofthe Frm (62) (0) 8 XH DOW bx ee me typically + | and X's yore set of eteoval Fetus (6 (Udy aod X° i 8 les entgory. The clement inserted in position 1 fkeerioes the fears of Nand, to 9 substantial exten, the choices of and W. At Tie pon phone acute roles are largey eliminated frm reieuirlingges hey are expressed @s gen propenies of UG, ‘hain the fewer of Khar er: Pepa Parr: 2 ‘A urter proposal estes he rales (623) to the fora (6). (6) 4 Xzxr* boxy XI XW For m maximal, we use the conventional symbol XP for Xs #0 sfien dropped where no confsion aise. To forma full phrase marker ‘exch X" isreplacd by a sical ment wit he eater fetes of X, ‘Suppor that n~ 2 and m= Lor 2 (63), 3 thatthe posible rule Forms are 68), la Kay bX Zx! eax exe oxw “The nonterminalelersents are XX? (conventionally, XX", oF XB XP SNP. Asus further that Z ¥ ar sngle symbols. We el 2 the “gece (Spee) of X2 he elements of W te corpements of Xd Y in (64a) an ener of X2- The sat oF Yin (G4 imbued ing on farther arculation ofthe theory et ws tentatively a ex an adjeret. Note thatthe notions spcier, complement, and adjunct ae functional (lana), not extegoi ts, thre eno categoria sre ‘Spe, but rather 2 elation pete, aso on. ‘This i eseatally the sjtem of Chomsky 18ts, and the basis for funher concepts defined there. We continue withthe seumpticns ming ter to modifctions required unde eres Musaken (1982) proposes thatthe bar lew are determined by he feature stem projected. axial. This, X°= PX, projeed, man imal, X"= OX projected, ~marimall X? =[X, + proected, +m ‘mall Note that ths approach perils # distinction between adjunction ‘rte forma at -Stucture end by ajution operation Se ako Jackendfl 1577, Stowell 198, Speas 1986, Fuk 1986, Belin ad Koch 189 Wii the move to Xar thor: the pate stucere sen fra pa ‘ear langage is largely rested vo spetention ofthe parameters Hot determine the ordering of head-complerent, hend-adjunet, and secierhead. Choirs above ere (peal for ahaa language ‘Te ules (62)-(4 themseves belong to UG lorder aide) oot o parc ar grammars. As iscussed in etions 1.1 and 1.2, the ciation of * hart hens sructure rues has vay ben plasible gal fringe he fry, bucaae of thir edundany wih ielinale lsc! properties. {Xhar theory can te sustained in ts most pene form, oie of items from the lnicon will etenine the D Structure phrase markers fora language wih parancters fas, Tees ofthe lexicon are f to geal iyp: with withoutsubstan- sive content. We resi the tem sical the foemercatepery the {er ae funnel ach tr ea eter set. Lexical eet head NP. ‘VE, AP, and PP, and their bcaepois (adverbial phvasc ete) AL DSirctue and LP, soch such XP est ply ie approprine seman scl, stieying Fas ised caer The hea of thee ateores une (1 catego Fentress (2) pramaticalFenures such as 6 fates tnd others checked in the couine of derivations, continuing teste ‘or ofthe imerpretations of orbologcl strate dines in ein 1k). phonolopet met, athe arcuate by the mapping 1 PF (4) inberen seman and jtacte features that determine emanti)= selection and foerrial)-dectio, respectively, Ths, permnde hss ores determining that the an NP td 8 propostional complement. th te specie oes, Ae cicaeed fn secuon 1.2, estonia least n prt determin by elem; he determination comple trecan et asention to scktion, We may now assume that 2 om pement appears at DStroctre ony in 0 Opsiicn, mated by is Fra. Since the cempttional res can ado farther comlenens, it follows that at every lel, eomplements ate postions, in Tat, ‘marked the sae way at each eel the Projeion Principle). The Pros ‘ion Prine und the related conditions on marking provide a pert lar nterpeeation for the general eorton Frat Dtucture and LF. ‘Fertional items also have fentresracture, but do not ener into rnrking Thee presence or absences deermined by princi of UG, ‘with some parametenasio. Each funtion erent has cariain Sle tional properties wil take certain kinds of complements, and ay oF may not tke a speci. The speierstypialy (hough perhaps not thnys) ae target for movement in be sense discus ear. Hence. they have wo independent semantic role ata, AB sugested im seton| 121, we may aise thom to Be inserted in the course oF derivation, ‘nk ome pene conion on DsSeactre eqs tir presence ‘We aesme th ful clauses headed by # complemen C, hence ina CP satsiving Xbar theory. C may havea specifier and must have a ‘Complement propositional pea thot we assume to Be ede By w= recplesan Parmacers s caer functional category | iecion, whch asthe obligatory comple. rent VP. Hace. a cause wl ypicaly have the form (68) (Brest 72, Fass Fest 1960, Stowell 981, Choma 1986). (65) fer Sle C Spee VPN ‘Spoofs are tially opional we assume his i tr of Spe. CP. The Extended Projection Pinpe (EP) sates hat Spe, IP oigatony, erhaps as a morphological property of For by we ofthe preica. tional character of VP (Wiis 1980, Rothstein 1883). The sper of Wis tie ater ofthe nominal complement of VP th bj of WP. We ck these o be functional rater thin exeporial nations for fer fen lew, Sce Brean 1982, Perimttr 183. By the Prjation Princ fe the objets Bpostion. The subject nay or may tbe Hy te fled by an expletive or an argument at D-Stucture. (Spee, 1] is Uerfore «pote! O-goiton. An acval of poteta pation is 20 ‘Apesiton, ober are A pastions (tar postions). A tates nd ths eit. complement and subject (Sper 1) are A-posiions, and [Spx CP and adjuoet postions are Apostions, A chain eaded ty 20 tfement in an postion ian chat chain heed by an ere in ‘a Rposiion san Zchon The distinction between A-and pions, nd between A-and Archies, plas entra rote nthe theory of move ‘en and other modules of grammar. We rues to xe problems om ming these notions Recal the tuo interpretations of the sate rule R tha esses lea tos with ther infectional entre word formation by acjune honor checking (ee secon 11). IF we adopt the former apprese, i fotlows th the operation K must apply i te D- 10 Sstrte deri tin, because It “feeds” the caer ofthe phonological (PF) component The checking shemaive does not suey imply that mombolopcl Properties must be deterined by SStrutre, but we wil assume tat "isi nevertces ue. Toms thatthe infected head of VP mst ‘nave it etores assigned or checked hy Hat Structure. either trough ‘oweing af Ito Vor trough ising oF V0 I Ge sections 13.1, 1.33). ln the towering ese the §-Srutare chai is defisrt. Thre nt tee fore be an LF operation that rans the adjunction stature [V-M to ‘eplse the trace of the lowered J voiding the petentil wonton and ‘viding an LF similar to what we Bindi a language wth rising a SSurocture (on some empirical consequences, se chapter 2, At LF ‘hen, V wl always te at nt a igh a (65). se Chae “The [V- complex may aso rie Fuster to C. fn Vcc languages such as Germanie generally V rae to C und some other phrase res to [Spee CP] inthe main ause (Den Besien 198, Viner 1990), The ne’ phenomencn appears ore argon in Eaelsh questions and Senne other consructons, We ssme thee to have the Fer iustated in (6, wh Bing in Spe. CP as asing to Cand ving the trae ‘ing the race of whe (60) fey ho asp Jon fy re lll By site of he genera properies of Xba theory, the only opions in the pret poston. iiraducing a chu, are PX of XS: X? may be railand commonly st tein embed elses sf Spe Cs oneal {the Duably Filled Comp Fier ce Keyser 1975) We assure thal is feral over movers of the queaton word to the (Spee. CP] pos tion andthe sume era of other eorsieucions ‘Stoctre of the for (5) may alo pest in eed postin. as inthe indict question (67a ote declarative ass (67). (61), won fey Who Ce Job has tll 1k UU bebe fy tat ok has met Bil i peter fp fe Jb to meet Bi) i. Gr was ede) fp Cy PRO to eet Bi) 1 (674) sd (TI) the C Bead of CP is min (OT it has and in Gb) for. The head of WP is [ene in (6. Tis {tens i G70 ih. (Sp, CP] wed in (67), Ba i x Be rea ina in othe etbedded comsractions for example, (6A) the rears ans (8), o the comple adjectival clause (60), where here good fessor to ttee that Op is an empy operation [Specs CPL. Cis emp Sm oth eases and isthe ace of Op (8), the mat ep OP Cl Jobe eA) 1. May is t00 clever ee Op Ce PRO 0 each “The embeded enue of (6) are predicates open Setenes with vast he poston: tn (fs) Op cout te who, an semantically vacuovs io Disease As a mate of (nontrivial emptia ut, F at LF incl the propery of srg Adi every variable mus have it range fied by rtrd quanter, or have Re valve detzriqed by an antecedent Since the operators (6) are vcuus, the value of te vale mst be fncd by the aecdents mon ary, the choke Bing dteined by ‘eealy condtiogs on predation ‘These properties sue to explain sech examples a (30) repeted here as (a), the aun ving the fre (690. (©) & if Mary's too deve to expect anyone teach theo we doe ‘expec aryone to catch Mary by Mary i to lever fey OP Cy PRO to exec faryone och an ‘The embedded CP isa rpc ene of tng sucesive-seh) movement analogous to (70) with eho pas of Op G0) wonder) oho be expect th to catch Theva mutt be Hod by anon or PRO in (9). at a At mst ot feta bythe nets thor eins urn tothe eperaine principe nf Hig tery In acho 133, 182 3 the ons diecast the ae rant hove Ay stisancendnt Futons PRO ms fe any. sou 1 Mrs nin Moy rc [PRO atch ite he vara si be ound ty PRO. wolsing the pone jt sd, We therfore hve he intercon (Note a he aun ses cra tha ing ed wo a capa tn selon va (On the sae asonpos. ne a edit probe of epi the evans (0a fhe evn of nen oer moveme sin the anlose carlo 2). 2)». she man ou me eps ho 8 Marys toler cpa ah (2) who did Jot eet pont that cous {nat ces the loa conditions on moverien at vice. ee tection v4 We have assumed so far that embeded infra ae CPs a in (xvi shor () | wonder wb he decided fer CTPRO to atch In such cases the embeded subj must te PRO ithe C bead is empty nd must be an overt NP i the Cate-assanieg clement fo, with aca varition. But there are other prepositional pes in wich Fetes PRO note Cassin comhenenin fr on fr * ‘ovat (74). Soko tteys ito be ietetignt) Jon considers (Bi nee) tha if made Bi ny end fri] “Thus, in (74) we cot hve for Bil or PRO instead of, Smit, insu construction asthe, he ebadded sujet cx be te, nike theft CPS. Compare: (15) a_Biisbeeved [to be inetiget] 1, it was dada [0b eel) In genet the embedded sbi of (4) tae very much a 1 He sh objet ofthe verb of the rain enue he matrix verb), though i tot Bracke complement of the ecb, but rahe the sbjet of an bedded cause Constructions of the form (Téa) are rather Wisi ‘atc to English in sina languages (ee. Germs, the eomesponding cpresione fave he properties of ii) (73), and 0 on, "The embeded cate of (Ms) contains I hence TPs thet i 20 eve ene for any futher struct. To wezoun for the dierenes Com the cenbedded CP infinite, we mst assume either tht the embeded SGhise 6 just IP, oF that thee am EC complemensize tat assis ‘Cas, Hk for (Kayne 1989, Ox the former esurpion which we wil aru er the embeded subj governed by the mats verb el ton that wes to aph Case, ens trace, ad tar PRO, a in ve ‘jet constuction. Note thatthe guxtion whether (7S) 6 ang totsructon (ie Join aeons tbe inelizen}) ora passive const ton ike clam wre bleed) Coes ot ai ese Cones avin ‘eco discarded es tnemorie artifacts (ston 1). The coasreton is formed by Move a wea "hat rie" the Caseassiging property of the verb having bean “ator” by te passive morphoogy. Inthe xa pics of (4b-<) there 8 overt Funtona ead Asuming the phn tunis indicated, ihe hei ap EC T,of the embeded pass se projrtions of thi peicats, o-alled sal laues (Sowell 197%, ORD, Eider way, Bi the subject f the ernbeded clase, behing 2s fn ia) and nt the subject ofan rbd CP. ‘We haves ar considered to functions exept: and CA nat tat exterion is that jus 38 prpesions are projections of funcional ‘Segoe, so are the tational noun phrases. "The Fonctonat Den ia ths cae is Da poston lied by a deter, a posesive apeencat tloment, ora prenous (Pita 19669, Brune 198t, 1982, Abney 1987). Pracps ad Pareto 5 “he phases tat tare of Bland Jo's pete of Bl wou heeore se he Foons 70, (08) 2 fo that ctr of i 1 Keetohm Pos ge pete of Bi] nS. DP] ing 15 td by he st” oe 1 fo whch te ai Ps adj’ by ponte oe fon The hed shan aed Post (0) see apps foe sane, Tursh-rnising vil peewee Kort i), Nowe prs neal sre ret se merase nm ge even conan “ome testo: enol (7) We mit expe thn, ne Nein 9D {Sekar to Voitig fo se Langer 954. Thre ar mers ter eoseqeess wich ne aot panes here We i i te lo Stalaraon Nowe Pra fo DP or konto woul ere ‘We might ak wth thse corto etn te or cotprn 0 ht AP an VP re ho corso etal tenet in ¥-¥P or Maal NPorsacion i. rel cha Seal bean senile in Ca ssponent sn genet ea dora cobain ef tata). Such poster get» ens fon ofthe fret een i ese sng peso beg obictendc the vesos oF Xero te a cong suring ht Tens) a Ae ange has aon Pa (096 ets seta ed Arh ear ain! eens. ‘Amine hat VP Cant AP) ms somone ef Age we no tae Stet Spee"T-Agr-VP or the pn we he ale TP cow 2 lem of emesis ot) ih ang Ag an ample ta WP-AP cing component ofthe Agr hod of ar). Poh ines fest rondo tesa sree [Spc Agr VF i i ‘State the pcr fT sot eran © omer) by ‘Aw es not pene yt Hen i son Uo) he cate eon nog nent ett en sch cl as thr oul eo nt egencn ef setae, The ‘thr eee tongs tht ord sell be Ape T (ee 90, ‘re Agr since sete sen nd yarn Ce fer The proper rent ofthe cnftng propos may te hati are to he con oI each elt of fetes fm ile ini pent and et Gas the te i oe Stren and ost Cos Ths te fa suc wl te, oe « taper ‘Aare 20d Afra st inform woations to dings the two fusion {eles of Abt Sore incites» Functional role o before. and IP = Agr. mF ere we omit posible [Spe TP), Embed in tis structure thee right aloe a pease beaded by the functional element Negation oF Fethape more trol. «eaepory tat inelodes an afrmation market Sint others a we (Pelleck 196, Laka 1980). We might proce 10 #8 sume tha Case and agreement pnerly ar anifistations ofthe Spec ead elation (Koopa 198, Mahajan 1980: aso se secon 14.3 ane shapers 2.3, “The sas of Spe, IPs anornlous in several espe. One i that it may oF may sot Be @ Bpostin. epending on fecal choices. Ths. in (G8) the sujet of hare postion ocepsed by the race ofthe arg rent Jo, taken to be the aget of; ba the subject of sem 6 3 ten -ponion, which ea alo be occured bythe expe i (08) 4, obo sens [re have re ise) escent John has hur himself) [Spe IP also the only postin in which Ook ot asigned within ‘he n-command dors xia ead ‘Sch iionyrarate properties would he eiminned if we Were to sine that thematic sbjet originates from a postion inieral to VP. the asin to [Spe. IP] Collapsing the infetiona nods to # for con niece, the DStoctureanderyng Jobr met Bal! woud then be 7) {Kitagawa 1985, Kuroda 1988, Spvtice 1988, Koopman and Sportehs 1991) ipsa Parrters A o © wa ‘The suet and object are now @enarked within the secommand do tain ofthe verb et. within VP. On the present assumptions. Jon [Spee VP and ras to (Spec Po rece Case and prodac 2 Vile fn By LP ew ave raised LTV rales tot $Stucute od {i subeetraes to [Sp IP] only at LF, we have a VSO language (at ‘Structure. We oe ssigned to subject Ube ester olen the Sense of Willams 1980) prt compostonalyéeermied (Marte 988), then these proper ight be eres eral 10 WP 3 prop tes ofthe pred ems (ube, ‘The astumptionssbetched out hee provide ceria veron of & “univers hase hypotiess" 2 non that ha bes expla om vi ts poins of view I they are on the gh track, ypolopca variation sould reduce tothe ordering parasters nd properties of fncionl ‘eens As used cake, we expect tat D Strutre and LF vary bide in evential proper. D-Suctire releting lec! properties trough he mechansns of Xba dary athe porate opie er Fsctionl cement, and LF beng the oucome o an nerant compa hal proses that maps D-Siructre to $-Sinctare ad the t0 LF. A Ferber propolis that there sa uniform tract representation of ‘rolex: thas. per ply associated with [Spex VP, theme o Be test wih complement co V, and soon. This appears more pause ae ‘idence mounts questioning the extence of erative lnghnges a the Jel of theory (Saker 19K, Johns 19) See eet 1.2. Wie hve so far kept tothe assumpsin of Chomky 198ta that all Jograt oles (ll oles spar fom he role of sujet ar anes 10 ‘ters of the hee. Ths wssumption has epentedly been guesoned ee ‘us lacey buen abandoned. Te mention few exes, Kayne (1988) peo- Poses tht ll ranching bias (elding“aramhignus path) so, e cuper scene ternal oes wl be ssid o nonsters. Keyne sugests for ‘nample hat double-obet verbs have the structure in (in whic ‘as g wll ack NPs poet contained within its eorpiement (0) give Mary books} Similar ideas have been parsusd in oter sues as weil Belt and Rie (1988) argue thatthe nderyingstrcture of "psyh-ver6”con trctons such he prom dsr Jobs (81, where she ister of tris asia the Ooi theme (as usa, then sing to (Spe. IP, Sole the sister of V' russ the Pole experience (6 ao Pesky 19S, Boucasd 991) eh ve anon proposes that doble-obieet verbs suck a give ener into D- ‘Strcrre te form (2) (Larson 1988, 199, fran opposing si, se Sockendot 19500} om ZN ae vo lL eN ow amok ohn | oe ‘aes tote pty min verb pion of th higher YP shel ile Join pave bk to Bl, ANerastively, operations smart hose yl ing the pseive consrstion could “absoro” the Cas of Bil forcing to cpa Parantos o 18810 the sbjetike postion of @ honk which in tan becomes an jc, yelding ofr gave Bil Bok. In (2) the et ste Pek ‘hourh marked as theme bythe ver, ots se, Larson also ind cates that des are the innermost complement of . Ths, the stu {are unde Jolt ed the Book crea woul be (83) @ ve ae vo lL eNS tetook ¥ crily read In this case he ster of the veeb fan ave ha ot mae atl nd the soe intral tei signed to 8 nonssrth Boct) With sich modiestons, the notion “8-osiion” i sil wel defied, bat "A." and "A-pasiion” are not These notions are formally quite ‘fee in chacacter. A partir cccurerce of category ina phrase rarer, ors noe 2 0 postion, spending on wheter its Oarked in tat phrave marker. The notion "A-pesition,” bowser. depends upon potential @marking.” whic is wo say that it presipposes an equiv lence relation among phrase markers eo A-poiion fone that i Inarked inthe equivalent postion of some member of the sivalenes ‘las This isnot an ene staihtforvard notion, and with modifier tons of the sorts atin st Becomes tepid in ay way tht sl bear the considerable theoreti! harden that has Bx id onthe versus A distntion, which enters crcl nto large stas of current werk The nitive content of the distinction to te captured is reasonably ‘kar. postions and species ef inBtional creme shere a range ‘of sretural proper other non-Omarkel positions (Spee CPL de mens odoined to XP, nonhmrked postions governed by & head) sre a-diferen range of structural properties These are the former “ eptrt ‘and postions, repecively, Thee are varios proposal aso how to captre this duction in ers of paral elasse, and how to exend and sharpen t (eg, for Spee. DP). ‘One approach se ehaper 3) is based on the observation tat ceo fupetional elements aren effet features a Rea. thatthe oust Be adjoin to this hee to ceck its inherent featres aerativly. 10 as Sen these inberet features to if). Tense and the Agr elements are fe {ores of in this sense, but Ci not Given esc end Ly we 8) tht 8 pstion is Lorelei is he specifier or compere of feature of The toate postion ate the former A:postons, with the exception ‘of non marked erent such a cf (83), But this exception ‘sil not be problonati if independant comsidrations block movercrt tsch ements o any Lelated poston (ising) econemy consi trations pert raking only when it regu (only Last Resort, ‘ovement. then te se il not ase see saan, 1 ‘Along ths Hes, ene might reconstrst seting ike te A vers EK distinction. The count wow rele on ploperts of occurences of 8 ‘ategory in phase marker without reference to equivalence ass oF phrase nor. Other ine of thew otons, an ining theory. peat {ofl io place without to mich elif, We eve the ater ith these info indications of dietion to explore. merely noting Pere ‘bar certain cones tht serve as foundations for much current work ste ripnlly dened on the fais asmptions tha have teen ‘dehy atundoned and therefore mest be reconstructed in soe diferent ‘iy. With these qualifications, we wil comin Tote the nein with ‘heen conten 5 standin cures chia work 133 Derived Syetatc Represetatons We have adoped the EST assumption thatthe desvations from D> Siucture 1 PF and LF have a common pare DStructure i mappa © ‘SStrvtie by Affe 9, and the derivation then bres into 180 inde endert paths one forming PF, the ther forming LF (tbe PF com Peent und the LF component, respectively). These are the two enteral feterace ees Since our concer er is pmax in the nareew ene, w= ‘esti eustes othe computation fom D Srucure to LF. "The pat ofthis destin shat saps § Stature to LP sometines tial, But whenever structural properties relevant to meating are not sedy pnssed ot Stace, ts mapping sbstantve. Following ‘Chomsky (1977), May (1977), we assure dht scope of operators i icp on Pare « _sevtrally epreseled at LF ip tems of econ For interosative ‘peters. a5 wl be discussed below, movement fo an spc ope ration aes place someties betncen D Struct and S-Stoture nd rmetines beween Stature and LF. Movement of quaiers May's ‘quantfer rain” OR) generly an SStevetre to LF operation ‘The exaps of ioversely Lnked™ quantifeation dvs by May, 8 ‘n,n deat that Srucre configuration doesnot sue (4) everbody in soe tatan yk it. Here some Hallo city has wide scope, even thovgh a $ Stace ce conainad within the universally quntiisd NP. The core interpret tions tracturly represen), wth teenie subiet NP ng derpone OR. and ie extent expesion having raed il forthe. 5) lyse aia ciy) everybody in esi ‘See May 197, 1985, for ute motivation for QR. ‘Sisce it isan interface ll, here ae farther tegument on LF, ‘bauer cerca hs tals tha thee should be no ra expctiver oan CF representation. ta Such expressions as (8), then, the expletive len here at somehow ‘br climate mapig rom $ Structire to LF 6) theresa mann the 0m ‘One posit that cam be realy dissed i tha he expletive is BY deed. The EPP damands that a clase havea se a very {acti ve. Deletin of sor wo tae thi requirement at LE The expletive also appears to have estore thal enter into apsenen ith ‘he inflected verb. He) hos eatures ar [8 pen, singular (87) they ste person. purl (7) There are men in he room ‘A sarong frm of rcoveality ofdeetion would presumably prevent the deletion of an element with festires Given thal there mes te eliminate ard cant be dete, the remaining possibility stat i {he target of a movement operation, with thease of he expletive (a ‘ma in (86) and men in (ET) moving tothe positon of the expe Whether it eomirecd as subsiton or adjunction, we mn) sume ‘tat this operation produces new element combing the relevant le {ures ofthe eapeie and i soit: [ere marin (here me cs cuper (87, Let cll his am amalgamated experi, eavng open ts exact fee. ‘We now hive at account forthe apparently roms rihward sererien in these cases, thai, the Set tht the infeed ver ares tv the NP that follows i fad are cannot Be interchange a (2), (9), The LE movement analysis erly predict (hit para. Tare mast be replaced ut the prose sralamating with mus be mond Tit rm in eure, I the operstin substation, his equrement ‘sl flow fom the recovtrabity condo. If he epeaton i june tion, wil flow from a fextiesmatchine regimen. Aleratvly, swe might assume tht shee licks @ features aed thatthe vert ae neat isan Structure refs of agreement atthe LF level Between the feted verb id he amalgamated expletives agreement ates Peo ed hy the ama. Noe fark that one of te cel gropets ‘ofthese consracions at thee an argumear assecited with the pleine een flows since FT demande thatthe apie be replied From an $Strcture covespending to (6) then. we dese the LF representation (8), ve ace of man (48) [there ma] is inthe room Sie the expletive occupies a A postion at S Suoctre (Spe. IP. ‘he LF movement forming the arspumatd expletive A-movement. Ie Follows thatthe rdsion betwee the assxate and its trace meats he ‘artow conditions om Asmoverent. We new have aw account for 1) fect that inthe over expresion, the expletive and it asso conor to the lca equtements of Ahan. Ths fliows rom the fat that ‘t LF, thy ae amalgnmned to form an Achaia. We therefore have ‘apleinesseinte relation: ofthe ind sted, ut not tose 0 (9), analogs 10 (0). (99) 8. there sears that aman isi the room 1. tere seems that Jon saw a man © thee was thought hates @ mar were on sl] (90) a. temman sees tha isin be room 1, emor scens that John sa © ems thought hat [pcs of Were O95) [Not hat the locally condition athe expltve- associate pis tha of ‘Amoversa, not Binding which permissible in the aralgie fo), rinipleand Parte, o (01) we thought tat pictures ef each eher wee on se) {We etn in ction 1.31 some probes fetes oF this analy TInsection 13.1 we aled to an approach o Casein ters of vse iy for marking. Expleves appear to contradic the nip ice they ae not Crarked but appear only i pesions to which Case _ssigsble—in fc, only in asset of such postion (subj, but this follows from the fact that -Stactre complements sre present only i they Have a semantic role gypsy, a oe) Thus, fn (92) with nominative there and (02) wth aeciative here but 92) s imposible (62) a. ection hereiya man bere] Bb. Teele [thr to hem man te] ©. ier to hea man hee] ‘But now these cts il eetly dr the vsbiliy apoach, AC LE we sw Rave 3). where isthe ace af ¢ man and EA Is the wnalamaed ‘apletve (99) a. hetieve fq shore, manish] DB. believe lle sere. man to be re) © “Ted feg teres man be ee) ‘When at expe i ina Cases position a § Stet, ts asoeaed srguinent will cess bei ha pston at LF and wil a cone uence, te vinble or raring ‘The analysis jst sketched suggests that Case checked at LF even ough manifesta $ Strate: Ua tgp thr conditions eg tng chsching or assignment of Case ate LF condos, not $Stctre ‘ondions, despite appearances. The sme concen i sugested bythe feral approach to Casein terms of wisi, which inks Case sin Se to Saeory. As discus eater, thee a preference on grea ‘oncepual ground fr inerfceconion ater than S Stare con- ‘Sion’. The various consideations so far aoe point in he sme rection, but serious problems arise in trying to pursue this course. ‘Werctua tothe tp in ston 143, “Twig tothe SSectre repesetation with parameters fi this ‘cerned (presumably uniquely) by the choke of DStraure and LP ‘Rpreeatons. $ Strate sulk the ve bac lvl (D Sect PF, LF) i that sss no consesint external the computational System. Ie would therefore be rwsonabie fo expt that cantons in volving the intarfce (poral, codons eating on Ihe seman ft Chater Jerpretation of SB) shouldbe resce athe itesace Hel hen ‘ives not appying at SStcure. Nevertheless, thre may be cond tion of UG that mst be satisfied atthe Stare ve. “Tere eee crosetingvistic tition inthe character of Situ: in prc, funtion ements vary the ways they ane aruda ‘Structure and here are resized exert. Langues may sb ir, as pote with regard wo the placerent of S-Srucre inthe deviation of LE from D.Sc, that the poit of tranching 10 PF. One wel Studi case cones the sppleston of Move a th determines he trope of» eueston rise (commonly called the "wh phrase,” by Histor ‘al acident, moving to the piper of he propston. Te Engshnpe languages the efits of the movement operation ae vise, yeling the §-Sirotre form (54, whee 1 tbe tac of whe (6) wh do you wank ba ove 10 Bil) ©. what do jou want Joba 1 get how) Ina mile question such a (Hb), only one of the question phraws over by $ Structure Tn he counteract wo (MU) in & Chinese ype language, the arson to ar isin stu” at S-Srsctre,oocpyine the postion of the tae in 4). We mug ellowing Huang (982 and much subsequent work, thatthe pea moved to clause peripheral poston at LF, ying a EF for resembling (04. More generally bth ps f lang lh ueton phrases wil ave rowed to cop postion unde his operation Inthe cone of the devvton, within the LE component if ot before (iggnbotam and May T981, Rous, Hert, and Spoetihe 1981). ‘The D-Siructure forms ae therefore alike i elevant respect in En- ih and Chineseaype languages. as are tbe LF forms, the standard fxpectation (ee sstion 11) But the §-Struture forms ifr. depending fn wheter the operation shat pices the question phrase in the positon that ternines scope applies befor or afr the branching to the PF component at § Strstue. One typeof nape (English. Freeh, ee) employs overt moment of gueston phrase in the cour of derivation ‘or Sirature fow DStractre, feting the phonloicel eoraponent, nother pe (Chie, Japanese et) eve all question pss ins at SSincture Both types of langage employ cower” movement within the LF component Tor nny ity question pha. A thid type of language (ee, Poh) as overt mosement of all question phrases, inipand Parmer, « Destructoe and LE seprescoatons are aginst (0 the other wo language types. bu the SStractres fer (Lasik en Saito (984), Given #narow theory of parametric variation ofthe srt decised, thse thee language types shoul erin properis Teton fe tures. Cheng (1991) argues thit sod (inerosaive, delete te) ‘ast be indicated at SSiractre i the pee pein, hence By choke Gf cther C or Spe. CP. he Had f CP and especie ths serves fore indicator” in seretiog Whe the Frege sees. 1 the lenicon ontains an element Q rarkng ero qusions) them this element wil fc to iemtfy an expression aan tmerrgative wheter or not i “ont a inst question phrase. There ne ee the, forthe que fon phrase to rai Spee, CP at Stature Lacking the Semen. Engage mist employ over movement of question plas to [Spee CP] vo be dented as an interogative n §-Sictre ‘Suppo father tha eonomy pics Fer operations that do not feed the PF component ever oters that 6; hen if oferaions need ot ke overt to sty seme cendston, they wil beastie to the LE “omponeat ppg a Late” he derivation as posible tthe pos here they art ced by LF colons {i the eae nr deus, ‘onditions of scope). These assumptions lead ws to expec wo Basic eat sores of language inthe simplest ese: (1) languages with = Q element and the question pase in stu (Chinese, Japanese. a 2) nstages lueking » Q cement and with x singe quston word in (Spec, CP] (En- ‘lsh, German At LF all question pases wil have mene 20 Ca "he ust quantifier can be imterpeted wih ts Scope deters sod = ‘ound vaableheding av argent chin. Other typolial ferences should hen be eduibe to nicrmal morptology ofthe quesion phrase for instance, languages of the Polis Hungarian «ype with mule ‘eontng of question phrases at SStrucure (Chough perhaps 16 [Spec CP. see Cheng 1991). On assunptions such ab these hee are ‘onions that ust e sia by § Stcture representation. Over and covet movement might have diferent properties. Huang (0882) proposed th the Bounding eondons on overt moweent are ‘else in the LF component 90 tha we have stzh pats a8 (98) in Engl nd (958) in Chinese (85). whe do you tke Bock that eis ni ubuun [piping shi de sho you tke erie who RE. book] » coaper Both expressions fave the interpretation “for which person x you ike ‘books that enize but ony (95) i wel fone. The English ex ‘bl (95t) wees locality condition on reverent (Subjoene its Chi st counterpart sf rom hi costa foe waning approaches sinong eters, Hranp 1982, Lasik snd Suto 1988, Nishigaochi 1986, Feego et 1988, Watarabe 1992) ‘A sini phenomenon i found in ule questions in Ens languages, To, Engl (96) is well formed with the ineepreaion| (96d) expres inthe LE fers (6). (95) 8 who ike books thc ete whom 1. forwhich persons js [x He Boks tht erie 9} hor, who] fbr Books hat erie 1) We hve assumed tha vert movement a in (8) oa}, places the vetion phrase in the poution (Spee, CP. Possibly covert movement, tot required for mood spesfintion, may ajo the question pase to TP resting it ike a quant prose mssaned scope by QI. Typical. such queion phase as whn, hone share semantic and diebutonal Properties of quar phrases, and might be composed of an indent ‘Gosnte, a vfestue ad the ereion an he quantifier (Chomsky 1964, Kezodn 1965, Nistignc 1566, Ki 1950, Watanabe 1991), Ace endinly, who would be comporsd of [tome x. wh 8 person and ‘von, Ke would wot then be sping such question phrases wre £0 ‘are propertice ofthe indeite cuantier, acing to TP in the LF ‘empancet by OR, toxgh it remain o expan why hey ove 50 fee "iy, untke QR, which is ype clase bound ‘in Engishype languages, relative causes are formed in much the sane mmnner a ineropnves: en operate plas, which may be ithe {an EC operator Op oF morphology ead oa question phrast moved 0 (Spe. CP, kavng » trace that fetons asa vale in on. (97). the peopewho Joh expected to meet In cither case, the elie aie i an open Settee uncining as & (pre Sinturs)sytsk, as shown in (7a), ad sts the bounding tonditons on over movement, 8 strated in 8. rips ond Poon, A (68) a. ‘the man fisto you lke Hooks tat xi 'b. “the man{Op (at) you ie books that ee ‘White Chic ad Jaane have quton words in ve es tow she pops oF ov even eng 16, Wann 1 Ii 199), Ths oberon ugg ht line dau ee oer ‘overent The reason ih be tha prediaon mes eae Sire (Wile 198 Irs, ye ne ane ele ea 3 Siar condo I oul sino ned te sao tenses ‘tt form eae ith ns ronoueunpeve prone) an ll NP ns inte penton oh arate stew (eh , Dorece 0 These coierions no ther censors with Caper. suche he cml sol ict i con 1.22 (89). ‘he cl rope oer move tlre) (7) Nao OPC a PRO WP ened hast sof OP Clan PRO tot rs wo ‘ove Ge he eat pops te open ete ution a red ‘ats mu bve been orn by xen move res Ste Sone sont proprio lng chess epee Je inl by §Stmure configs nependety epee the LF compovent et Pe sich «propery. Then twosome reste (0) Phot a Siar 5. Phot Leander econincton hati with he mowed Praeueated “ast tween he potion iste Ah forme corte te propery P nhs conden SSacure Thor sve varios wapt of conning ie nen ‘A goed dt of ist into tee quesion dives fom he rine fining tery ~cl it Connon spain at yom cot {command ts ameeeder Go seone 132, 142, We oy formate {Ns aa requiemen thal an eexpenion 9 tb Aft no {commands y 8 pronoun aan poston heed tothe nag Aor ere. Th (0H) ad GE) Tl Ate he poss (im, hs) Sos nonmarket or tsar 2 apie ‘But in (1016) fe ecomnands Jan and must be asigned reference in same other ay (401) Joe thous Mary took pice of him B. Ihit moter thought Mary took «pictur ofJohn he thought Mary took a peur of Job “The prinsple Command aylies fo expressions general, ec 10 visi ics unwell J, ase see {12}, analogous to (10D, wih the tne of who in the position af Join (10) (402). he an who [thought Mary took a picture his} 1. the man wo [hismetbe] hoa Mary took» ptr of the man who [he thought Mary tok a pct off 1 (022) aod (1025) he pronoun dos not ecommand &, Even if he pronoun and vrible are referentaly Hnket, the variable is Acree. though Abound by is operator. The variable ad the ronoun can ro tr comsirund as vnables bound (bound) by who. The interpretations fe then auch hat «thought Mary took pice ofthe man x Sch that meter thoweht Mary took a pte of especie the ‘evince of (1020. any sgh (Choesky 1982 Higginbothsm 1983, {snk ad Stowell 199. "Bt in (1026) he oman? ne therefore canst be linked to tis ‘variable twill ot be Ate: (1024) Useforeeannot have the inter (petatin ihe tnan x such that thoopht Mary took picture of 2 “Thee wething"erong” with his irterpreation i at is he inter rotation of (102), Bot it eam be assigned to (102), by vinue of ‘Command (the property oF strong erossover: Postal 197, Wasow 1972, sank 1976), “The prcciple Command lk enters nto the explanation ofthe mean ing of the compen adjectivals of (9,3 seus cari (ee (68-68). “We nov sk at what Jevel Command apples. Consider the examples aon, (009) 4. yous he tike he pct that Joh tok} © how many pctrs that Joh took] di yow sy he tied who [sai be ied how many pictues that Joh ok] In (10%) fe ecommands Jon and cannot tke Jor as antecedent in (4036) there io ecomnmaeeltion and Jo can Be the ascent of te Inthe mulip- question phrase nnsraction (103) Jon et can- Ponce ant Parameters 5 ot he the antecedent of It mst be, hen, that he e-commands John atthe level of representation at hich Command splice the binding Properties of (103) are tose of (1033), ns (103), Returng to tke two options of (10), we seem to be ted here 0 adopt the fist that Commard apps at SStrcor, before the bracketed question phrase is moved te pretasal postion st LF at which pos (Q0%) would be fom silt (1030). not (108). Atermaive, e ould sue, in the Face of examples such a these, thatthe cond op tin, reconstruction hols for LF easing but not overt movement More simply, we could dispense with Beth option, resting the ac ass tion that LE moverent formed (18) fom (10%) the trace of the LF moved pase (008) {how many pietures that John ook who] si eked] Recalling that LF movement doesnot meet he srt oly condone ‘of $ Structure movement, we migh eet the sssmpton tba the entre [NPs pid-ppod when Sve man i reged to the Scop peston, a ‘suming ater that how many is exacted tom the NP yielding a LE ‘orm alo the ines of (108), 7 the tac of hor many. (405) (Dhow many} who] side tha etre hat Jo oo ‘The answer, shen, could be the ple (12. Bil), ani that Bi side lied 12 pictures tat John took, But in the LE fora (03). fe c ‘commands Join so that Command applies in (103), Parsing sich ties a these, we would ot be Ted 40 Sdop the assumption thet Cou mand applies at Strctre,lenvng us wih the preferable option thet ‘onions involving interpretation apply only atthe interac tee A fuher consequeece woul! e that {103b-c) have somewhat diferent forms at LP: the pial fe i oles (Hoenstein and Weiere 1850, rer consrstions iets the proces of resoratmcton erst are ‘ths consstent wth the eateton ofthe conn on inept to {he LF tee. Consider (100, (106). they sid te adres Jobs ater who [sid he acini Joh ater © (gues) whose father hy su he adie In (16a) and (1064) he cecmmsnds Jo and cannot tke Jo a is excedent, pve Commacd la (1058 he doesnot ema 1,50 both ” Chaps ‘in be taken as variables bound by who, sieling the interpretation ‘or Stich ponom id + admires Job's Tae’ tn (106) he does not ‘Coorimand who. ut caanet Re taken as «variable bound by who. ‘en though thisintepretation would eave 1 Afi. The complement of ts is interpreted a (107) wth euro, aalogoust (106). (07) for which pts they sid he adie father “Tha, we have roconssetion: eaten of os fr ai the pase st in he postion of ts teace 1 (106) (Chomsky 197, Frei and ask (981), “Questions prifrte quickly wit utr inguity. Corser. or exam He sch eonsracions as (1), formed by sueesive-ecic movenent| ff the question phrase from the poston ofr © IRE pesion ef f, 10 [Spes. CP] ofthe mati nse 18), ich piste oF mse] at Say shat Aa ed Rest 1s which ites esc a] ey sty that we kd test ars (1986) observes thatthe anapor can ake ether Of the tiie [NPras its antecedent. But an anaptor ean only be bound bythe cloves ‘eommanding sb ae we so inthe earesponing expressions (19), ‘rthout -moveent (4109), John si tht Bi hed at itore oft] bs 1 ey snd [tht we kd ose pices of each cer Het) Here the aneseents mast be il we. n (18) the sume binding cond tion requires that cach of the traces be “se the quton phrase bing interpreted for biing asi were In One CE the eer oF these postion (hai iin). “Another problematic example is (110), with the interprection (1106) sand. on our current ssunptions the LF representation (110) (Higin- ea 1980, 1985, (0) a guess which pct of which oy hy sat he aie for which boy x wich picture yo they su he adie 3] tit toy wie picture hey sid he adres) Reconstruction inthe manner of (104) and (107 oes mot yield src ture fared by Command. Neerthaess, he cannot fe cosines an ‘cecutenr ofthe bound variable seit “The foal property entering into reconstruction her secs to be at the oir eenneetion 2, pronoun f) ae referentalyCosomneied at LF ici an Parcs 8 there is» y sch that y contains a and B ecommands y oF is ae Pot thot principe aplyng aS Strctre, els incor rss or (U6), baring binding ofthe proneun i (OB). The esrepany suse that he problem wik (11) He cirware “The problems are more general Conder (11) (UL a the cli that Jo was alep he wont ss 1. the dai that John mad, be wor dacs {Case (110) is anlogous 1 (110 es (1116) to (1036), On our current isaumptions, the peozoun must not take Jalna antecedent in (Ha) fr (ID the corcason is correct for (Ila) But not for (L110). St farther complications arse when we consider difrences between thee ‘amples of A-movement end “scrambles” consacions i which he normal rubjetabet orders nveted ‘We leave the topic in this unsettled sate. For further dicaion of | (base ad related matter. om various points of vw. see Lak 198, Reinhart 1976 1983. Var Riemsaijk and Wiles 1981, Higginbotham 1980, 198, Langendoen and Batstella 19, Bass 1986, rein 1986, eben 988, Sato 1989. and chapter 3. ‘Consideration of LF A-movementalzo sugges that there is 38 S- Susur condition teensingporstic gap (PG) consuctions such 85 (9%, imerprte a (112), (112) 2, whic book aid you fe uthout my eeding est 1 forwhich x, Book, ye Bled thou my eading fest ceasing of PGs by Achains i site rene at thos formed ty LF movement do no liense PGs as iestated in (113), with te Structure (0135) and ue LF form 1138) (113) 4, tho [fed which bok ithout my reading 1 istics boat who Bled 4 sithout my readin el “The icepetaton cannot he fr which book x, whe fd without ry ‘ading PG conaricions then, provide some evidence for the exis tence of §Siractue conditions ‘The condition tht Beenies PGr must ako account forthe fet hat these consructions ae need by Aches but not Acai, Ths, he ‘chain (ihe Bok, of (18) doesnot zn the PG yuk the Re hl hich oko (112A), withthe same re relation (118) the Book as led fithou my eager] 1% haper For further discasion, ce Tarlden 1961, Endoh 1983, 1985, Chomsky 1982, 1986s, Kzyne 1984, Longcbard 1985 Brosning 1587 ingee 1980. [Note that cten the aceptable PGs are somewhat aukoard: asin caer cas dines, we fe intersted inthe flav deviance of sr fous consrutons, which quite deat and demands explanation. The feral erature om POs reulay uss for ifastation sch pais 35 (113), where the fst is completely grammatical and the second sharply deviant, ut these ene o bt sulle to show that A-hainsHerse PG aie A-ha donot, because (1S) i ued out for independent re form of contol theory. as dlstrated i (116) (Lasik and Unageeks 1986), (15) athe book hat you le without PRO reading “the Bok that was fil fithout PRO vending} (116) 4 the ook that yr led vitbout PRO inking} 1h te Book that asd fthout PRO hiking] ‘The question of Structure conditions sho aise in consetion with clement eal iertiied ss alae eg, promorial cis, vel in econs. Case Features) Sine the propery are commonly overt PR, they must be manifest SStrctore (Last O81: we omit Hee te possibity tat ies of the PE eomporeat might be nich enough to Handle the pheomencn. As indicated eater, the question becomes ‘nthe sue if we eum the checking interpre'on of iftional es- tres Suppose ngs th Englah walked s inserted into D-Siucure ‘ih the proper wath. rst the ater being checked end cers ty a sntnctic rele R hat oi [past and mae. Suppose Crter tat sich fetional slements 2 ene] ack phonological matces and are tus inl t PF We need ot then stun thar i overng rule coining past to alle, to be revered at LF: nm alieraive pestiy [thatthe D- and S-Structres ae ale, with raking the verb to the ‘nfsctionalpostun at LF mirroring the poses ha i ert with x fais and in Frese languages for theory sternal arguments ear ing onthe mater see chapters 2 and 3). The same question aises wth regard to Case marking. Even iis over the concep possibility temaine that cents ene the compuatoral system wih thor Case Features aro ineted, tes Being chested only athe LE eel. Any spparent $-Siructrereirment for Cae would have 10 be satis in some other ay Sen etion 143 and chapter 3. ices Parstne n ‘Other sheoryintenal conerations sagas that empty categories must be fcensed at S Structure, in partul ace in argent chins (ani and Sei 1984, 1982: see ection 140 I the relation of pred ‘ation htding hereeen an XP and is (mac) abject mse satfy Structure conditions, a5 sugested ea, i sso natural (howe ‘ot neces) to suppse that ease ofan EC set of prediction ‘shold also take place a this eve. Thus, scoring to Riz theory, the ull sujet parameter reders to property ef the eter ofthe veal infletion: i Kalen, “strong agreement (Age) cee pro sje in French or Engi the “weaker” Agr doesnot. We mht expect then, ‘tat hs condition must te satin bythe $ Struc confgwation, The plasty of this assumption is enhanced by conseration of properties o expletive pro. Consider the Structure (17). wat tlen a book 'b.eseems{¢ 10 bea book ming] ‘nano subject language, the expressions ean surface in thi Frm, with ring epletive pro and ete, ere po sles by some ABS Bat i 2 non. mul subject language, e mst be roped by ©Siactee, fier by an oven expletive or by rng of ako lh poston 85 incl, (U8) & 4 here was soem» book A a book wae stolen theese [19 Hea Book ising) | a book ems [fo be missing Sem S Structure property, it appears must ensure thatthe options of 18) have been taken by the § State lov not nthe LF component ‘he problem becomes mare severe if we adopt the strong vero of FL that regis that expletives be replaced at LF (sections 13.1, 1.33) ‘hen the Structure frm of (117) wil appear at LF essentisy a he (i Tora of (118) would follow, then, tat the relewantdstinetins| ‘ust be eb $-Stuctese: ro Heed at $ Stace, pert ‘ing (17) in Hai at 2 Engh, For arative says soe chap tes 3.4 Was ako been propored that some af the conditions that have ‘een assumed co aply at LF actually spel within derivations from Stuctre co PF (ep 1980, Aoun et 1967. fecannot te that he ‘itons apply atthe level of PF repretnation itll beast the imerfice lel PE we have only photic fates with forthe evar siretire. The sscmption would be. then, dat ese conditions| tly ier at SStructre or ot some level intermediate between ‘SSirucore and PF. We have sss o for that Xcbar theory applies at Structure its properties being “stnic! mver" to S-Stracture and LF by the computa ‘onl process. Suppose that X-bur theory apps at SStuctre as sell Van Riera (198) argue hat on this esumpion, merce test! not be renricied toil and animal phrses (X® and XP) #8 ‘fr tay seed. Movement of X'(=X") coud be allowed, t0 be Fotowed by a proceso “Yepenertion™ tha forms a proper X-bar struc tore atthe $Strctre lve in nial way. On ths analysis (19) trol be derived by moverent of he N eatery Lana elowed by eration of eine to sat Xebar theory at S-Strature, ine being 3 peting out” of the fests of Lan. (G19) five Lasung hat erfene bese alsich 4 sludun hase n beter (ou) thea Xba theory apie at Structure, Emons tracts prsersing typottesis for ssstitton ation 131 follows in essen since con Fie of etegriat ears wil late X-aeteorte pial. ini ter conlaion wil ae ld for adjunction. Seppe or earl, chat an X element i dined to the YPZz Forming (12), (020 (eX YP. “Tie sructre violates Xureheory, which equ that X" head an X sxrcture. Adjneion of XP to YP, however, would yield & structure ‘ontene with X bar theory. Aunt af X° 1 Y° yes» two-ep- ‘unt ertepory 1¥, Y°) thm aie structure “invsbe” to X-bar Uheor, Parsing tis fine of hnking, it may be possible to deve a ‘erson ofthe rruture preserving hypothe for adjonction: stent, the condition tat «category canbe ajined only to a enegry of he same bare 14 Modules of Language 141 Government Theory ‘We have refered severat nes to the notion of government. more “ca yrety of commana (action 131, We ase tentatively hat Pape an arate » the levat notion of command is ceoenmand, The concep of goer socot bas entered extensively ito the study ofthe vaiout modules of ‘prams. Hence. ight modcations in frmution have witecaning epireal consegures (Se, among others, Aun and Speriche 198 ‘Chomsky 1981s, 19864, Kayne 1984, Lasik and Saito 1984, 1952, Riz 1980. ‘Wersy tha a governs Bi ccnmmunds Pan! there cn etry y that “pete” Brom goverment by a. yprtets in ths seme i {eommanded by a and eter (12a) or (1216) Rede (120. yiss bare dominating. 1 Yintersnes etwoen ane Government is conn f the near omer of (8) seconde with the sal of the he prameer (Kayne 19, We speak of “Xogoverament™ hen the governor has the peoperty X. There ar to main exes of [overnment to he considera amecedent goverment fo by an sre ‘ent of and head govrmmon of By & Ben. We refer to these cle: ores 35 prope government. “To make the concep of (oeuty pov, we have to se out the -otons "bree and “intervene in (120, Consider the wo inten Wie tes bare o be an XP thai ota caeplarent. puting side tow the amipuous satus of sonccmpleents to V unde the varius ‘amifentins of Kaye's unambiguous path theory (ection 13.2) Tham, (022) the bracketed expressions ar all XP, but oly thse subi ‘Bore bares Tor the elements they cota (022) «wonder which took eh dl the students that they show ed MI monde which book [ti et renee fy who re "I wonder how ho mt fromecre whofe the ae <4 wonder wich book Hohn eft New Yoo before here) AT wonder how Hoha eft New Yor efor he ed ‘heen 4a cach case the ace indiaes the positon of eateton, under theo tend stepretton: hus, (1222) asks how Joi fed thee, not how ele New York tweens from within a Barr, the teh bel ‘wl not be atteden governed: other it wl ie, When extracton roses barr, the expression is devanindistng that smteceder a haptr1 foverment is a condition on propery formed chins, In (1221) no aries are crosed ad the sentence fully ratmatial.t te other ‘ass hurries cessed athe snteneet are deviant. The ilains fre more severe i eases (172) and (122), stating x characterise ference betwee argument and ade etacion. ‘Wappen that not ony complement bel aio a specie exempt fron Beserbood Balan Riri (ORI) ole thatthe feces of resistin in align etacs re fons he bjt ofthe verb ut et From its suet. The obit the complement of he verb i ota bowie te goverment the ite thus governs the race ft hy m-xtacion From the objet, a required, Bat he trace of me-etractio for the b> {fe wll aot te antooter-yernd the subject isnot a compere Tene is barn, whether goverment i based en e-command of corzmund, Hence, we have (23) but not (123. (29) a. proneto visto fet) 1 ofhe-have seen many ‘havescen many of hen” Mimo nesone meget) any octbernare ineliger! Bat now conse (124), deriv rm the DStrctre (124). (124) a prositergo [ft ne} ineligen 1 bebeve "many of them intetigent bs neieego [rch intel firm bine many intligent “Tv many af them 9 be) lige Hee the complement oof roo i anal awe. The phrase el ne i the specifi of the all cla, ence i aot a complet, Bute ‘tation is nevestels permite, We return to her iusratons of he se point. ‘Weconlite, ten tha XP is ota Barer ste complement of & shad H or he specifier of the complement of H. The cofguation of properties not surprising, given hat the Bead gical sates the fax ures of ts nail proton and gress wiht specie, so theres an indict agrement elton betwen mma projection and ite SP ‘fir The sane cbsraton sugges that we generate peop f= thor iis the complrsent of den the dauehirs of (ts spss vl is Head) ae ot burners, When the ead ison the question of inci nd Puram 8 extraction from it doesnot aie, tit could arise in other congue Tors oppose tat ina smal elaise (125, YP — XP. with XP beng the ead of VP and NP its pecier (tbe subject of the predict XP) (25, VigeNP XP] tn (124), then, = YP = AP, and its hea is the AP ineligem. We ave already see that te spcier x nota bari, Example (126) as ees the at hat the saree tr the en (126) whor does consider Ly Bi ap ay] “The satu of (126) i 00 dierent frm that of wo she erat ‘Thus, neitr the complement nor the heed of» eoipemen bart Sian, (120) the main verb pase of the embedded else snot & bari, a its VP ead is abo nota brs, so that who esr eh (127) 1 ender bo ob apy et ls ih Note that in the ease ofthe small le (1265) a8 wll as (127, we ‘ht alo appeal to the segment thewey of adjunton (secon 1.3.0, ering hat baer be «category. nota spent, and taking the rd 10 be seers. Rene pt posable brie. We have desl ina preliminary way wit eae (9) of (121); consider 0 case (), wth he configuration (128. where serene Between ap (8) occ B call tht « ecomeands the intervening eet, which we assume Faber to command thus, tefo-rigit onder in (128 exqreses the {Scemmand reaton, Two eases of intenention have Beso explore fetloming Rizzi (199), let us eal ther rid minaity and related sina, (129) 0. Ri: yie ead H (arbiter by Relay 6 ofthe game “ype as. ei ninalty cen et nr of are ing he cay ‘te denne a rr Toe oe coe no mii, Nm sh a pes Pe ‘ren in (130). = (1) Wien yisaba i. risinan A-poston, then ya pir nan A-poson {© Haisinan Acpostion then isu specifier nso A positon 2 chapter Recall that se consents As and Rpostion are not property define in ‘uren heory: we suggested a way to appecach the problem a he end ‘of ection 1.37 ae contin Yo assume it her, The thre basi cases of relative minty sre stated in (131) for heads postions, and Arposions, ripen) it pal (ee em. (031). how 6 Vlohn WILL (he cae) 1 hon sce [ha [FT era [10 Fi the xl] ute [er ow (ohn wondered [WHY Ie fied the ear) In contention einlogy cas (18a) illseaes the Head Movement Conant (HMC, ease (13D) sperasng: and case (131) the Whe Toland Contrai, ks te stuctreneates (3le) 0 be understood as expressing Job's puzlement as to how we fic the car nots 2 ‘ery about how Re Wenders, In (34a) wil intervenes betwee fx snd it race, and oth i an vet sre heads. Un (Tb) interveesbetosen Jon and its tee, both it and Join ae in A-positons, andi the specter of TP To (Ie), ty intervenes betwee rw and tac, both why and Bow ae in pesitions bd why ithe ecier of CP. In al thre eases the expression [severely devia. ‘We noted carber tha aunts and arguments behave somewhat i Seren with regard f extraction fom bases (ce (122). Te sae is tro in ete (105) of intervention: eorapare (13) (adjunct extration) vet (132) argument extraction) (53) gues cg what Jos wondered [hove xed ‘Wie uracepetle, (132) much est serous wolstion than (1340 “These observations havea wie ange of deserve adequacy, but fl short ofa satisfactory explanatory poncp, We return te the question tthe endo tis sean ‘We have disused some ofthe properties ofthe st case of preret ‘government: antecedent governent. Let us tur now to the second ene: head government, Throephout the modules of gemma, we find tions (1, XP), where H i head anid XP a phrase with Some property| sspned (or checked) by H. Tet telations met fat coins tha freely narrower thin eer variety of command and ave thee fore een Been considered ofall under the eaegory of goverment, We sowed carer tat goverment By ver sues €0 assign Case bar ‘Prosi and Perms 2 PRO, and Heese trace (tion 1.3.2. fal casts the relation is nar rower then command. To Case theory we find tat ave Vea atsign for chesk) the Cate of an XP oly if the XP iin ala relation to V. The web ind sens scusaive Casto he Bok in (133) Bu notin (13), (033). e found the bok 1. we found yy he Book incompretenibie) (130) a we found fey thal fy the book a ncomprcensiel] bb we fund the sewer, when the book rived In (139) no barrier pote he hook fom goverment by fn The same is tru of (1343, bat hee the intervening ead C® (i) brs sneer of sh ook by find te (13a) mis 8 Bar. To (134), then the book must resive Casein some ether wn. the contrction in which it sppers is ifival, wi ot resve Case aed the oom inion ungrammatcal ain (138, (25). We red fe te Book 1 win pez) we ound Jot when the book 10 aris 1 (135th intermedi head (e bare government of the Book 8 in (34) satura spon, pnconet ene ea ‘into Case theory. ° ‘The positions ro which a ee can sien Cte ares, spill, those in whic «trace can appear. sugssig that governs by a verb can ee tac. Ths, longi (133) (13, and (35), we have (136) oe a3, (36) & the book nas found + 1. the Book was four [ay Fincompreersie) the Book Was beeved fo be inochi] 4 the book seems tobe icomprchesii) (137). he book was Found ey hat (+s ncorpebensbe) 2, the Book was ied Ly #10 8 peal Torning to PRO, we finda similar configuration, PRO cetnot appear {in govern pons hou in which, with the rope form ofthe ver (Care can te aspned or race Heese (19 2. ve found PRO we oud [4p PRO incomprebesi) PRO is ako excluded from postions that are governed but in which Case cannot be assigned, a9 (139), (199). tity express thet PRO to Be integer) 1. te ese there to be Fund PRO} ©. Ss ised PRO to be elie] 1. itsees [PRO tobe eigen) {As discussed insect 132, we assume that the verb eleven Engh fakes un TP, not a CP, complersent. Thus, PRO is governed by eli in (39a) and feted in (1390, though no Case markings pst, The entation are bared. Ths (193) docs ot mean that they apres the betel thet stone or other is inten, wth arbitrary PRO, or at they expres the bee ha hey are inlges with PRO bound ty they. Silay, (1396) does not mea ha it was eed hat some me or oer is intliget the ponte form can ony beter with {rrane, levine 8 tace inthe postion of PRO. And (1398) does ot tren that we expected there to be fourd semeone or othe, with abi vary PRO. ‘A leciyshtion tetween ahead adn XP alo is found in Oahe ‘ry, Ths, 2 ver Osarhs ony XPs thi the VP that it eas. On the “ssumptions of sion 1.12, the ye Oanarks the specie ofthe VP fa sere of V eations tt do not sil fl udergoverspent ‘hor. slang wth he complemen, which does ‘A cess lok a end governess shows tat C(-—C°), wheter overt xn tchaves either dieremly fom tbe heads we have considered. ‘Thus, PRO isnot barred rom poss governed by Cas astrated in 0), (140) we dei fey PRO to eve at eon) Simiaty, € dees not apear 10 lense trace Ths, we find that XPs move fy tel nl WP and CP, but IP des not (440) «fp admit that ie was wrong. Joh eve il op 1 tei ry] wa mad ey tha Jain was eon] © “ail wil visit eroerow} ink (tat) Cato doesnot iene tence of subject. Ths though € governs the tree in (14), extractions bared as ewe gan, MEUREES ave vae tous special doves vo ovteoe the probien ce bow). (042) 0 id you say ey hat yeserdy ripen Paras «s ‘Properties of Care uch ikstraed in (143). (089). Hot was decided ef 10 eaves noon Swe decided fee fp Jb 0 leave oer © weeded ey [y PRO te eave at 2000) IWabe bende of CP were to cease the race in (M34), ring of Je to the main clause sujet postion would be permed, Note that e does ‘ot intervene between Join and is tace we adopt the ations of re ‘nied minima ( oes under the ssamptions Of inimaiy). Examples (1490) and (1%) Musa the fet tht e does intervene be tween the marin ver and the embedded subject locking « porernment ‘sation between them. Ths, in (1436) Job ca recive Case frm 3 rs verb, and in (143) PRO i allowed, ater the mati verb non © rope governing it Thus, C function sn intervening ead, ba no ‘prope govern, ensing tase. Sinariy, while eter X% ypelly rise ea-governing ae ths ‘oi the race et thin, det not re of C. We nd ting oY ‘, N-ising to V (noun incompraton. Frain we C (Waser). soon. but we do not find Ceasng tothe mati ver that povees i (Gs. incorporation nto higher verb of «verb thet hasbeen raed tO Vescond postion). These facts loo woeld follow fin fae of C to propery gover, CC ato differs tm other heads with respest to bericsood. Recall ‘Hothead typical res a complement an daughters operand ‘rd fom batretbood, But the station is erent ithe ese of C ‘Consider the fllowing observation of Toro (1985), who wots the conrast between (La) an (145) n Spanish, (149 [,dequeauonlfnosaher fp ave tadvariones 1) by what authordon't you kaow shat tations (isan ganado premiositeraccnate} havewen ntratsnal awards 2, Festus Inston fel del quel C fy varias this istteauor({ “by whom) several tradusionsf) han guzao premio ineracionaes]} translations "ave won iternsionl awards 1h (445) CP is the complement of safes andi therefore not 2 baie: ts specie P is sso ot e barrier. and eteesdet government it ot Pocket, so exacon is parmie. In (last). however, etscon caer Mock: even tooth isthe wer ofthe eamplement of Cit sa tue leckngartecadent goverment A plausible conetsion i oes not ets eempkten or the suger of he complement) rom tareiood unlike other X°s that e hate consiere, though pursit sts se takes nto comple tht we wil ignore bere Cis unlike other heads that we have considered in other respects vst Unite sflestonal cements isnot a fear of the verb ths {ape snot Lcd, and is therefore an A-poston, not an Ac poston av are tke specie (setion 1.3.2), C al lack the seman ote of some ether head Th pect «good fs pprcximation ist he proper governors are restricted othe sel Fentures (seal eter inflectional Fears of {he verb and petbap ota) and tot nly proper povermor ee thir ‘orpltent fom barerood. “We kave sen tht © dos not seas the requled head porno of a subject tee In (1435) she ol complete failed Yo iene the trace of Asmoverent. The sme aire observed with an overt Cn the similar configuration (189) (45) “Stns important ly fr 10 fave at 000 “The paradigm with Armoveren (as opposed to A-movenest) of the Subject less senpvorward, Wife (12) 6 unaccepeabie, it Boscmes| ‘efoto formed if te overt compleraizer saben (446) who dit 00 89 Fee ef peste] In the proach outing above, the question show the subject race ead governed. Seppe there 5 « moll complenetizer andthe moe nent Of proven aucesve cyetcaly va the Spee of the lower (CP Ther the representation ould bea in (14, (4) bo dd you sy Lp ey He esterday] ‘Spectheat agreement taker place between 1 and ein this confiuration ‘We tcnaivey sug thet thi green poids wit atures alow ing it vo beenge she tae + The ungrammaticality of (12) Commonly Caled the thse fer, on the oer ban ics tat we etre Sharing snot posible wth te overt compleremize th. Not oo that thereto derivation slr so tam {147 aveable For (13) sce, ‘Apostion cantor hou ‘ute enerly, movement t0 an Apo reed [Spo CP) Sich “improper tnoverict™ esl nam iit Abound Principles and Pare © ‘arab, an eonstruios tha fl nr the principle Comme i sed in tion 13 eal sestion 142), ‘One eoncem of some of the ea Ierture om proper government (Pang 1982, Lasik and Saito 1988) ws the absence sare fers wih adjncs. Ths, (148) zea wth or without ha, (148) wi do you tik [ha Jt ee Sine adjuncts ik subjpts are not complemen che qustion arses how their aesare head governed. When a alent the sume meet anism i svaable as we posted for (14), Bu when shat preset no ‘ch metas ent as demonstrated by (142) (ee Ria 190) The Framework of Lasik aa Sito vas sto dierent so at she techni fal problem was actually appurnt lack of emrtdnt goverment Ba ther sohion can carry over under present stsumptions. They set that os a consaqunce ofthe Projection Prscne argument tates must bericrsd (r-marked, in her terminolgy) at -Sroete, hie adr aces are cess ool a UF (12) wi hus be red out ot §Strorture ile (18) will not be. Then in he LF compet ht, Ping seman tly empty car be eirinata. The testing cotsurtion wil low aoverment of the adjunct race in ast the same way tht i allowed ovement of he subject race in (147), the hed goverment ete ren olds t OF Tn he examples we ve been comidring. an adn races posible |i a stuaton in which 2 suet trace ot, We bo fl (near) the ‘opposite sate of ans. (149), ith movement the adc ons is ‘completly impossible, whereas 80), with moverent of» subject i ‘uch les severely deviant. (049) “how do you wonder faethe Hn i (Mary solved the probe (50) who do you wonder [whetbor John ti [sled the problen]} ‘In bot examples the intial trace & eppropriatly goverme, inthe ‘anne jst dined. The diference between (149) ant (10) tut fe thse Consider the structores of the examples in mee del. We essume ‘hat where occupies the Spec of he CP in which appears (1 aw do you wondely whither [JO Si ep Maty Solves he problem i « chaper (52) Pao do you wonder fe whether ‘he problem) Lassik and Seto argue tat wo just inital trees, but intermediate tees ust be spprepsnely gover, Bu he intermedi ace tov anteeden-governed nether (151) or (15) i the case of (152). [Ls and Soto argue the inermotite mee anced overs the ‘nn tie and en dete nthe LY component. Such a divation| ‘Snot poste for 131 sa they spent al Hens of adj races iso the level of LF. Ths fis present inthe LF representation of (5), 1 wil be propely governed but wil not be. ANG I not preset atthe LF level, then will not be antcnden-governe. Ether fy the, the epson contains tice at nt pope gover "We have jst sen how (145) and (15) ean be dstinguised ip ers of proper government. tn (19) thee wil cewably Fe an “offending trace" but there need be one in (150. However, atbeuph (150) i ‘uch beter than (149). i snot peer, and that fact reins to be ‘rained. Everly, wh-moserent 6 not pete to bypass an inte Frente Spe, CP a id im Bo (151) and (152 TS ene eons ‘rence of the wbjceny eonstxin’ os movement proposed ty Chomsky (0977) as pana unietion of several earlier consis on move Fnciang those of Chomsky (1968 ant Ross (1967, Subaseney wks ‘on are characterises severe tha proper goverment vctions, all ebe equal Another property of subjaceney that distinguish {ror proper govcnment was allied 10 in scion 1.3.3, Selscency ‘eomrsiae ver moveren, i apparently doesnot cones emert tmoverent betwcen SStricire ad LF. TAs is cen in the following searing pai repestd frm 8,9) (053) wo do yo ike books ha ei (058) who thes books tha exc whom) “The $Strcture position of whom in (15) ste LF postion ofthe tase ‘of wham afer UF rising. which yells astute that iin relevant respects enica 1o the SStrctte (and LE) representation of (13) Yer the io examples conzast sharply grammatical Silay 5 incssed by Mazo (1982) in lagiaps wth interroatveexpresions| ‘pit sich as Chines. the LE movement of thos expressions et omiraind by Subjarncy. (158) (C50) isthe Chinese analogue of (053) bat ie accep rch ike (134. Jon sid ere solved | Piped Pate ® (155) miu piping she de sho) oUlike [entice who ae. took) Wile LF movement seems not o conform to Subjacen, it does r= ‘pec he proper goverment requirement. The following Chiese ok ple allons LF movement of shee whit into the higher ase. bu ds ot allow such movement for weseme wy” (456) sangahido Ls weisheme mse sheme) youwonderLisiwhy bought what (156) ean man (157) bu ot (15. (157) whats de thin sch that you wonder why Lis Bought thet thing (15) whats the reason such tha you wonder wa 5 Bought or that The ee of he LF moe fed the higher de not be rey gover ders pcan ta ah based ne prein (15, Msn evened sone ae of he Bry of movene tv tum ote tc enc of geverret tet cry ah ‘spp eter mae prmar We ed ia eer & Ses fore of cman eat ake i operon te Se cccmmand Toe cen of ay wer eae grees Sled ty erin Rr and yan meng ergy ie ‘aly Condon. The Misia Conon hs we wa: Rid td Retvied Mana Wek oh slewng Res 30 Forte hon of movement, we ok the ve erm of aneronet, foe sovermen to ent govemet ad a oe Iya a ies et ch ‘As dus eater teens ave como deste we 6 fhe prerty sd ery twee hope a oo Sxtaaery ay of tung Ge ion) a pa the ase 1m! speig toner ep Ree Me iat ot ey cpt ty he wean pon whe es tiny est a sexed compen Wie eof ‘pc for tte cs 0 hehe innton at he Opraon Move» sol sys "contrac "ie hos ik rsa pumae ara ef arene 5 aheay cad the estate Ge Rs 0 9 os 0 up compe 2, Westy pd is apr of per pn of Say a deaton, Condens gute independent of Reset Mi Sy rie that eny heads on move to hed postions, and ty Tema pam ie a pson Fathers atin oer Serres, XPs con oe nl Yo speer postions and 2 ah Mowe ‘hy toa pion tt ceonan i Hence. he spl prope ine {rise beet em he formulation of cdo whch sede (159). (03) Mints ns {Tn ogoach vile e can lint tention onion of Linnie ats pret condion on eorony ef esas nthe deen of oneent {oy ever fi comands ad tee i no basi For B commande by we nat gover 1 eco hy te st oi cond sn haapens in ding try td eee Thus am atc Toran Senor june the ort deta es 0 ‘ons byte bing ext 142) Sit ge ny Fare sno permed by 2 and govering Ths conn & 400 viata Tor anezesene everest the stoner eonton (38 fur as ators hs fo teste fo bead ferment. Tat ‘Saat wh he perinet ain TRSnsS nme 190, so sce caper 3. We wil posed on ‘Srtommpion a i agit, noting the problema apes of is TU ake thine acu more pee and ssp more actus we tavew espn nw esc 4st aes ae 0 SEU eat mots and why volo of th cow onion Tvs sre flor adjuncts tha pen, oc hvougho, AS re mechani at cued, we wt suppose at when a ci Ek iSaemed by Mowe o he vce crested © gna * she economy onion (19) wed ae ented owen ft main onl Lasik ord Sto 1984, 1992) “Foto that ony erin entice etinate LF cic. st sty Ceres ptt PF obs (a ah Teor arene comonan enc 8 ema PE bt ad a et TEENS Nodes ea form a reper SD) We tere ipo ond Pare, o fore need some notion of festimate LF object. Suppose that he chain © of (161) isa leptimate LF objec ony if is unform (0 Browning 1987, (69 C= ty.) ‘The ony other lepinate LF objects ae operatorsarihle constrictions fh wire asin ae Aposton and Bhs ete or) Uniformity relational nin: the ein Cir wth eget 40 PUNIP) itcach 3 as property Por each bas non. One oboe hole for the relevant property P is Letednes, which we have sug ‘led ground the distinction butmesn A and postions sce eton 13.2. A chain is UNL] sft uniform th wpe to Loratedncs, Heads and ajunes are pon-L-elated and move eno son-Lvebted prions; hence. the his they form are UNL). Am argument ein ‘ons: only of Lele postions, hence is UNL}. The esi per ads, argument adjunets—are therefor uniform chains, ieptiate objects at LF “Tehing this a» ft approsiution, we ow repad the operation of een. ke moserent. a ls reser” prnepe «spa ate ofthe _rincpe of econemy of derivation (ake deivtions as shan as poss Ie, wit inks s short as posite pertions im peal are pele ‘rly to Form a epmate LF object. Deleon i impermicibl o form chain since these az lead lepiimate, Deleon i the cain Cof| SI), however, perms foe an an A-pomtion, were n> /> 1 and 6, iin an A-postion that is, the ese of uecaneyeie move ‘ent ofan argument. I his ease stared trace can be dled at LF, ‘soiing he voto: in other esti cane. ‘An epresion (an SD) sa Subjacency vnation if its dertion forms 4 stared tac. is an Empty Category Principle (ECP) volation if ‘terme, his starred race fens t LF ence, ECP woaton ae ‘more severe than Subjaeaeysoltions, which leave no rede at LF. [Note thatthe concept ECP i now a desspive cover teen fr various ‘ids ofwiolstions tht are marked at LF, etong them, iain of he ‘Soneny principe (Related Minimal) Wie eomtinae to sssume tat tases must be properly govern both Secodert- and bad governed by a lexi Fate, 28 C). To emily ‘he acount, fee us ay that race i rhe * a ls eer of these ‘conditions. Thus, a eae al be mathed * iit fi bth, or fais 2 omnes cons alongwith the economy sondion, and it will be marked "9 iit fais a toe, wth mute stanig indicating icreassd deviance. We fhve Faire of eectentgoveraunent ip the case of movement over 8 taro inthe eas af lowering lation of the C-Command Con- Consus the offending trace dts, the viltion remains at LF. ‘ie spc carer tha oly proper governor fie their complement Fan arise wl flo, then, at BP the complement of © wil fete from bared only Cbs ea entre: hat wil happen itv tries oC ‘Government nom is Uh special eas of local e-ommand when there sno barer,Subjceacy wits fail he economy condi ttre ‘ues chain tink to be minim. There is generally further device if the sila leaves a reside fn the LE rereseraon. Traces mus be fwopery governed (heod- sed anise pvernd). requiring sing Father than lowering, with device iasingcrosts aries The spe al props of ernie many repel a we ave en, pone further constraints on extraction of subjects. Deon, ike owen. Given by FE te egurement tht derivations must form fepitimate LF tlgets, Te guiding pte economy of derivations and repre ions derivations contain sepertoous sep, jus a6 fepesettons| Contain no superious sjnbots See capiers 2 and 3 for Furter Sacssion 142 Binting Theory ‘Amote the imaginable unaphorc rations among NPs sme are posi ‘he some ae necessary. nds thers ae prosrbe, depending on he ature of the NPs involwed andthe sync eouigortions in wick they occur Foe example, (16) hor eam be sefeentaly dependent wp- th oe (ca take Jf antecedent). wile in (16) anno, (462) dob aid Mary eid irs (63) Job exitiind bin “That (163) bas no ening in which hn ref to Bb, the way tha sein (168 does. (068) Jot vice Nstt -Apnaret. prenovn cannot have a wnecedent sas “ioo le” to ie Not Bot in (162), where antecadence is posible, 9 cause Boundary wel a Proclear aretrs a inervenes been pronoun and ance. There i 0 sich Boundary between pronoun and antecedent i (163) ‘AS we hve sen in scion 1.33, distance in thi sense doe ot alas sufi to make anccedece probe. Consider (165), where a case boundary iterenes Beseen he and Jal, eta snaphori sonretion is impose (065) he sid Mary eric John Inport its tthe linear relation beween pronoum and name tat inhibits anaphors, This evident fom conscraton of (16). in which Ineonce again precedes Jol. ye anaphora is pst (66) afer be entered he room, Job 8 doen Simin (167) sean tke Jol a its asec 67) bis ssid Joba “The peeratization covering (165) (161 apprxinately ae 68) (68) A pronoun cannot ake an cement ofits (com) domain 8 ‘The ecommana domain f an element i the wininal phrase contig ‘Thus in (165 the domain ofthe pronoun isthe entire sentence, Sines, ivy, the putative amecedent is icici that domain, the ahs hove tntertetaion is inconsistent withthe generalization (1) Tn (M66), onthe other hand, dhe demain ofthe pronoun i he adver ‘fois, which does no ince the antecedent Jha. Silty (167) the domain of he pronoun isthe mjet NP, he fs, which doesnot ‘che ohn ‘There are a umber of ways tat tbe generalization in (168), which feats espe of the srt and meaning of an wernt, might be ‘expressed nthe theory. One way iin tem of consti (171) oF ining. a vcr raion Stn in (19), ad redo defined in am), (69) hinds it ecommands and a are cxindeed (70) 1 Bis pt tound, ten fis ie. (17) An repression ity eer exresson not pronoun or an anaphory mast be Fe “Te fndamental elation a this approaeh evindexation 8 symet- icone. Foran altrative in term of an async relation inking, ‘Se Higsnbotham (983, 1985. Consider how (17) often ealed Cont tie C ofthe ending theory, wi tet the examples in (16) (167), Rep seman (172) fr sentence (163, il be excodd, while represent tins (173) and (174), or (166) an 16, espectvel wil be alowed. (072) “esd Mary cried John, (173) ier ered the rors Jo, sat down (178) bis boss tid Jn, ‘Note tha according wo (UH). 07S) spermine Fi #3 (419) be, sid Mary eric ohn, Hence, if (171i aly to play role in opting the generalization i (68), an intergvetation must be provide For the indexing in (175) tha apy pretaes the inpowsbl seterpecation. (176) sues in (his (176) Fe inden i tine frm he inden of en ier noe isthe anesedent of the oer. ‘Shorty we will se reason t senor this consent on interpretation of eoetaindeaion, turing noo the phenomenon i (163, sven that hee 00, found constrain on antecedent resonable to suppose tha (176) Soul gai play role inthe aceon. vides, ltt s mecesay i that he configuration (17 be allowed and (178) protbie. (07) ob xed Kim, (078) Hobo, tc hy (271) wl pot be fesve in exlading (178, sne tha onsen i im ea no ccumstaners wher the bine isan expression, whi in (178) the inde isa pronoun, Further, we do ot want geerabize (171) 10 Frc pronouns as binds nce that woul ncorcety prec ante: dence (162) by stoning resentation (179). (479) Jon, id Mary tice him, [As noted etki, there oct fet volved in this paraigr. A rovoun scary able o be within the domain of is antessdet ence, tomes vo have «binder fut nat pot be woo owe” 101 (180) i ae Priscipsand arse * Fough statement of the necessary constraint (Condon BF the binding theory) 1150) A pronoun must be fee in aoe domain, ‘The presse nature ofthe relevant local domtin reins t be speci. ‘The examples under consideration suggest that the loca domain is ap ‘proximately the minimal clause containing the pronoun, We il kn ‘uratenton hereto purely siructural approaches. Sce Wiles 1989 for 4p acount in eas of eles, and Rebar nd Rend 193 for one ‘used on predation "Note hat as predicted, a pronoun ca have an antecedents ease Jt as longa that antesdent dos nt command i (181) peri [ible represeiaion, (281) Solus, bosentan, Anaphors,sach as reciprocal and refesivs. recite decedent that ting ther. tn ths, thet tehawioe is quite aire fom tat of pro ‘ouns, which may have Binding antecedents, ut ned not Adtonaly, ss Engin anda numberof ther languages the etc ofa ‘anspor most beloca to zh anaphoe In parca, me ave (12), Ca Aion Aof he binding hes (082) An anaphor mast be hound infos domain. Under the nll hypothesis that the “lca dori” isthe sre for Contin A and Condition B, we rect complementarity between pro- ouns ard aapbers. This prison is confirmed to a substantial de ee. The ilormed (178) becomes prammatcl iis tou! prone replaced by an anaphor, sin (183) (085) Jobe, xed hit Conversely, the wellformed (175) becomes ted, if i pronoun is ‘replaced by an anaphor (488) oh sai Mary xz hme, Alt hat remain for this rough approximation ist sey the inter= estonia, That We mst paranteat (RD) anoe ‘neon that Sohn ercized Hany. The necosry prininle of itp ‘on snot emirely ebweus. Forth moment kt seu (185), ay Isyng open the preci import ofthe ton “anteatent.” ve capt 485) It the inde of identi! othe index of then asthe aniesden of for Bethe artzeent We now hav three syntactic constrains. repeate a (BGA. C), and he wo pipes of terretation (176) and (185) (426) A, An anspor est be bound nla domain B.A pronoun mist te reins lea! domain. An eexpression man be ie ‘Before considering Farber the prise natere of the local domain n= soled in Conditions Asn Bw ten bay to the semantic mp Sndening relations. Ear, we ine! tat (176) woukd seed 0 be Sarengthened Consider, hs ear representation (157), (89) afer Sh, walked in. Jobo eed him, “This ecprsentation i uly consistent ih the oy evan syntactic onions Conditions Band. Neher oeeutrence of Jahn i bound. ot hi is fee ine clave. According to (176) ef, eavoot be the Seteceden of hit do, fan propre antes Ii hs ‘ear why (187) dos not Have the iterpetation (od stats) of (88). Srereeorfereia nerpeation forthe two occurences of Jn com Tebutes enya ino degre of devises (488) afte John, walked in John enticed himself, ‘Given the sharp conrast teen (148) sn (187) the eat intr pretation, the exe deviance of (187) cant e abated to ree Son ofthe name, bat ether uxt em Fam the relation between the ‘eco occurrence ofthe nae andthe proaoua. We must rule oat (it {ended} coserens betwen these tao NP even when the cond ds tot ake the fst as is aneceent, Wo achieve This esult by strengths ring (176) 1 (189, (189) AF he inden of is tine rom seine of he ad ate oncerefee nti (085) mas no be med a comepondiog ishion (0 (190 (190) Ir the inden of 2 is ides ote nde then and are ‘oreteenial Consider the conrst tween the milly devant (191) en the severely (epraded (192), both one eleven sterpretaton ivohing only one ‘agi Posciplesn Pareiers ” (291) tater Ju walked in John se down (92) ob eitzed John Condition C exces representation (193) for (192), white permiting 098, (093) “Subo, esiz Jot, (198) Jobo cried J, (189) now comely guarantees noncorefereoe fr the two NPs in (194. But now consider (191). On the dese interpretation. the two ecu reoces of John cannot be conrad, sine (189) woul demand non ‘oefrence fr sucha rprsemtiion. Congestion oo, woe be prob onatic ander (8, sine (8S) demands ateedence none aeton o& the other. yet aman, te fly erent iat vn ih, pesonby ‘anol have an aniecedent. Ths pablem docs not aie ce (190) batted for (185), “This fa we have Unied our tention to anaphoric telnons among siogulae NPS. Conan compbitions arse when we extend the see of the iavestgnon to plas. The configurations siving rie to nonce ference etlets, by the edn oun abn, tm to ete to dha reference ets 8 well Portal Eka). ak os a covert Interpetation ofthe to NES ie markedly deraded in (195) ois over lap degraded in 196). (195) te ies hie (096) they ke hin ‘Comespoingly. (17) whose NPs lenely demand coeference i ba and (198), whose NPs lenisly demand overlap in reference stan {ily dead abo, 97) “Tike we (09%) oe ke me ‘This suassts ha (19) should be firth stenthoes, (19) Irae inden of is din ee the index oF then an ate isin in etree. 4 (195-1198) Condition B eacidescondering, (19) then demands ioit reference f the nearly conrsindened NPs. But «proce 7 chap 1 rs for pono wot in ofan bet 1 Conon B Cote Tecra ane on. (ht thy hn He wie vitro (cr) we tnk 1 wake iors Inconra ith (97 ad (5,200 and 20 alo an rerio Wises etre second NP tlc nth cc fe re out) mind saat Ra he oe ps rrr (30 a 18, thee x Bite ngrectaron atl for ae ears, Nets 22) Fan it conte prea fr 200 (a7) we ink be wis (20) we thik il be vont 2 1 eto (1) we a ust be dn eee, eae ache with as meaning of be te romne ASE snp he tw pronons mabe ee seat asin mest wth he se! earns NO ts TEMAS at dow mete 50a ot ands Oey rites rar han corre Toit see in 28 fr ea Bie nc demand ete he us room 8 BE Besar bu anor he hypobeeed weaketne, oelep Sod aioe {20 tay pa hee, . en, equa ise otf tao obits ion we ve Sete "Acca eccrcustnes cores, 908 fren Saf een mat he scommodte. the pty ty tao) casing en anc Hh Bee Sp ts fron ov. To everett, ove Sree ees dann ht a aspen ete cramer Gyre 1985) ah me emi 1 ‘Gk ttny ino eoeapond to carina of te fet of (Penge a move ane fea! sand fs insures ‘isan Ses Migr 6, Lash 1969) ras wi Smenon fe een flows commana oth 2s) ine ith eg 0 eh does not cement (7 cao eisai ‘repos Pacts Correspondingly, we mod nserpreive re (19), (206) Ifthe intersection of he inde oF al he nde oP is atten ‘rand ae dsc in reference. ‘The problematic contrat between (156) and (201) i now sag ward handed By Condition B. me (198) mist be fe, a in (2078) (0%, 207) 2 wey, Bee meg, bh tea: thew, 206) then demands of these representations tht the sujet and object ‘sot in reference (201), onthe ther band Condition Bip ie ‘vat. The indies of subject nd abject are theafore git to over- Jap (though sil aot be dental, ve (190, which we mai), (208) se 5 thik fy wl be iterious “The phenomenon o pit anteedence is inryaccommedated, a played in (2094-). 1209) a. John old Macy tat the souk eave 1, John tld Mary, that be should ave Several her prstes might lo be considered, Thos, in lace of the ert to st ries, we might enc the iterpretation provided for simple inxs ofthe som considered ere. Consider the following i {epretive procedure (210) a. Suppose NP and are conden, Then 1. Was an araphr, its coreferentil with N 1 asa pronoun. overtaps reference with NP. ', Suppose NP and are conndened. Then hey are ii, ‘The standard cases of corres, dint reference, ad dijon ee ce now fall tte place. In (195) 198) conesnderng is requied by Condition B, x8 the proexns re nterpeted a cit. (200) (29) indexing permite, and 20a) yells he tended interpretation of| "overlap in reference, It remains, however to del wth the phenemeton fp anecedene, and farther questions a inthe case of move ex. ex construction tht we Rave not eons, ‘Another possibility woud te to unify the ining and iterpretve Brocetufes along with the binding conditions thems, dapesing x0 taper ‘wth indexing and stplng (210) 16 21), whee D isthe relevant lee domain U1) a. asa anaphor, interpre it as cortreia with @ ‘ceommanding pase D. by. Waris pronoun nerprt ts sce rom every ‘ceommanding presen D. Following Lani (1976), we restate the Former indexing equverent for ‘express along the se ins. aisan respestion interpret ita din fom every cccummanding SE. ‘Nothing id about interpretation in other xis. The standard ex ‘Besar interpreted saihtfrvardly, Spit antactenc is now under ‘ood to ea spel cas of fee erence Ths, (209) any interes tio speed. ching those ada in (08), ad also cers. “rape an interpretation in which the taken to refer to Joba and seme third parry. but not May. ‘What nat ore comes cases suchas (213 (Wasow 19727? £213) the woman who ov him tld him that John 8 ingen Here, we have to exch the iterpetaton in which the (vo pronouns tnd Jo a corer The problem i thatthe binding contions permit ‘oth Join and hart creer with i, 1 aben follows incre tht Join a him, ci be crefeental. nthe theory ound ear, this ‘vos eactat by the fat that cindesng is an equivalence reltion, 50 That eoindeaing of both Join and hn, with hy exis that Je Ceindened win, which bared by Condon C- But we now have to coindeing, bene no equivalence relation, ‘However, te sane rest achieved simpy consequence ofthe inrpcionisalf (Lasik 1926) fy (212) om soit fom him, Free nero allows the te pronouns to coer and aos Jahn, to corer wih, Ie adopt ese options, kin and Jol, coer, ta we bave a ncn mterpettion, with Zn, both corer ‘vith and ijn from hin, NothizgTarter eed be sid. Many other ‘Sond cases oon the sae Wy “The theory culine exer, wich i she standard one, invoNed a indexing proces that sats the binding condos and exp OF implica interpretive recede. The apyrcoch jus steed amie Pricpsand arate o al three info a interpretive procedure, Whichever approach selon inom rena to consider he “focal domain” in which aaphors oust ‘ebhound and prenours fe ‘Thus fa the loca domain hes ben the minimal las containing the anaphor or pronoun. Bu this characterization sradeqat for wider ange of phenomena. In (214) the anaphorf fe ints minimal nse set the example wel formed. (218) John, bees [hms ob clever) Similarly (215) is deviant ven though the pronoun is ree in the ome stent ate. (235) Hot, bteves him 10 Be ver) We take the relevant diference Heten thee examples and the em: ede clause eases considered cali to be interme ef govemmen To (214) and (25) the ean ver gover the suhet ofthe ifeitival omy ‘emen,a seven from the acesaive Case that hows up tht subject tn 216) om the ether fun, there ely no sich genet ‘elation and the grarimatiaity fgets ar the reverse of the in 13), 215, @16) a “he beeves hms, sve) 1 John evs fics) “Te local domain. or govern category sii gun cao, niles ‘Rlerece to goverment, roughly a (21), 369i approximation. 1217) The governing category (GO of ms the mininal sue containing ands governor of. 1h 214) aed (215) the GC forthe anaphor or pronoun teenie se tence, sine the governor, elie, sis the Higher elas. Since both he 'tapoe and proroun are bovad i that demain, te former exo ‘red, in obeence to Condition A, and the tater bad, in votion of onion ®. In 216) the GCs the lowes cae, since te sujet ‘tsgned nominative Case by «governor internal t that clase, ite ‘suring that governmest is dened in ets of ommand). Sine ‘bin the lower cause. her tno bdr for the sujet of hat cae, {2160 sn voltion of Condition A. and 216) in conformity with Conon B, Nox that (217 cones pest the free te {5 fine and initval complements is Knesset postion, With copet to objet potion, Mite and nite cates ae pole le teen (29. Ho, Bag eh i ti (290 a. He tee Maye ist ‘ im tna for came the GC fr he aaphr or ponou she a cee eae sb ofthe ened cee even os or ca domi or Contin A ae canbe NP a5 we oP nin) (2 Heh ey Bite abut ise) “Ths seg nt 17 sul be xed nthe cus way 10 TEL Nir thc tage NP weld ten beh GC fer Mf ese se Ne aera Hwee. ate ashy ore comps Pasta ancy 2 arato (21) Son kes sore shou Bs ter he sgeied xen, 2) sala ea Te a aa nos uh 20 here NP con ot jm apo bat ao pti er, hat ater NP ne te tno Oo fol noteton crore tho. Soma ce prin tom Rand I cmp faci Om iceo. nha Cres a wocton cotningal samata ‘inchs compute thie. er coined ero (2 THeCC fer 9 iia CFC Treanctn whch bing cnn ou. pile be se “this coeaty dings 20 rom 2D, As wed bone, ee TESST ina forthe voto he ge NP in 20 Ba meena) Ts theater eg the GC rhe nar ths the ‘ie en Cow A aed. Urs te types a SES SUNicn 2 tn stp vets arena VP SEEN ee esa he uae af he je ih a ot towed ih Sper Ings he Bn edgar atc pet ides ones 1 we ty cat py mo oe for Condes sce he 0 aa or enous havea nr atl Hewes nial ea ES gover le ete ie prove) Sele Prnapsant aries 1 slays be the minimal such CFC in which as Kinng conan out In principe, be satisfied. This pedis tht (22) and (228) should oth te eood. infact the NP object ies n (234 gua ws 9 CFC. (22%) Jobo, ke [is tries about hi (224) Jon, ks {stores about bm} ‘As vxpered, (223) is pele. (228, while props slighty worse, sil ‘reasonably acceptable. Thi lner exarple thes provides one content ‘whee the use isincness i dnintin betwecn anaphors and pro ‘ou scr to sl down. 2 with nsf place of, wes a ‘ef coure, gramtial. Note dst, a6 predicted, cain detibaion i ‘aitained f there san actual Binder within the lege NP sn (25), (225) a Tike obs stories abou his) be im, “The NP Je’ stories about —_ isthe sratet penal CPC in which Condition A oe B could beste. Whiten 25) Condition A i {sft in that domain in (2258) Condon Bis nt. “Thee i seme eves that the apprent over in athlon sean in 21), €24) 5 ony losory. In 24), wre hin i conse ws John, the stores ae not taken at Jon's, Tis becomes even Carer in 20), fics i thal eanple the caning ofthe vr vr for the storie tobe Jone’ £@26) "ohn tt srs about iy) ‘This sopess tha (224) actualy can have « strate simi to C29, ‘bu withthe subst of the NP phonetically rll. Im that case the NP ‘est of ies would early consitute a CEC. tn (226), on the exer und, even f the NP obj of roll has al sujet, wil il be ity ound inthe minimal CFC, sins ta atc is derstood ws However theres one othe situation whee th sul icin dsb ‘on deftly breaks down. Engh has, a ited enter. confi | allows peiuing “longisance” naphors, (27) i # resentative Sammie | 220) May thinks pts of here aon ila] __Thouah fers is ee within both an NP anda fie couse here iis twound in is GC. the enti clase. Thee ie no potential titer forthe he apie snaphor anywhere in the lower lve, £0 Contion A could no be iafed ever io pone within the ower cause. Thus ere spe te 10 sek is inde inthe upper ease, wherein fat it ids it, Now tot tat pronoun posible in place ofthe anaphor. (228) Mary shins pictures of ber] are on ply) “The NP icnwes of her iit bas a phonetically nll subj). othe em fae cause(oerwit) i the spallest CF that conais eran 8 iovernor of he of pleres, depending ca certan assumptions aboot Shipwment of petive Coc ce ection 143) and in which Ar cou in inci be free: And her in fc, fe in that domain, The Sted vei in dstibuton thet exis ths ore accounted for By the “elativind” notion of GC in 22), “There sone remaining problem 1 conser befor we eae ths op Recall example (26), repeated seus (2) (229) ‘Hot, bles [ise ice} der the ear slate notin of GC his was correct excluded By Condon A. But wer the characterization in (22). its not. Thoveh Fans fas governor (nite Inf i the ower 8, there no potent tinder_The OC should therfore Be the entre seateocs sd lin should ‘eval seal bir. Assuming te basic eerectress of the for tmaon of binding theory we have Been devcopng. something eer than Covdtion A must be responsible forthe dormedeess of (22), ‘We sens tha the relevant conition one dicased i section 1.6. ‘tic eats aces fom conigrations im which they ae not propery ovemalOn the fac of hie condition mg seem elvan, boc thor smo trace evident in (223), Hower, i pase to repr the Iestion between a retexie and is aieendent as invelsing agreement. ‘Since areement is preraly sil fest phenomenon, the reflexive ‘host neve to a postion sllcmly near is ameceden. This might bay min the yt, a in the citation processes ofthe Romance ln sees. Ir, thn must happen nthe LF component (229) hs ovement i ene ate that not pepe governed. Tis preach dc accounts forthe faa ebservain that binding reatios ee ovement process fal under abseoty very similar consis, Fur ther iT inceed, the requirement of agreement tat fring the (LE) movement ofthe eflenie, (250), stich otherwise cou have bern Problema 5 aed out Pringle nd Parmass (230) *hinserreh ath ar opt Kn he eso Conn A dex nt dete cumple. gen he formate of Go Gh Mower. m te abe of acd In agen eae a it ea, Tes gp ls Sioa apeement tere wl be me ay resent fi ice 1987), il an ‘ie it he Condon A reuicet on efe th cn etn ths pny saisued unde te opr govern eqremet tne tc gos tho rs of vir eto cosas al oper oan ee try Hain, Lak and May C991, sxpening amen» repeat of enon (1983 sogast a teak seuenenterneer rec sprenion and hr andes ase w eonhean oo ao ten Toe S Sie tee (an LF epeon af ok evens din he Sur rc ance” 6 tontaesme 2 (31) Themen sa cach other £22) Kot the men} ech 4 og ote 4230 this LF overeat can be lng intance. Ore rut 3m be ong ditance. One reading (he non actor one) of Us sentence eps as (238) (230) chy sai that they are tale thn each ther 29 Ina hey ott ty tae han, ‘When the ver ofthe main clause a “nonbigee” verb, however, move ‘mens charcteriially locked. Compare (235) with (236) (235) who aid ey sy hat tey ate ar than ¢ (236) Toh dt they mtr hat they ane aller han ¢ Coespondigy, he “we scope” rang for ech is wave with Sonbriée vb. leaving ony the contactor reading (230) they mustered that hey ae ale hen each eee “Ts, th ajo ass fel nape ef ccs, fic ananhons,refexis and eh ‘isplay constraints sugyestive of movement, me Me io fay wo the question of the oe) of rpeenttion ele 281 ote bing conor. 9), whone dann vos roy 106 ‘ rite amecodent wo he appropriate poston to Bin the refi. 305 (fateh D Strate need not met Condition. (238) Joho, seems to Hts 10 be eve) “Te sue is ot ene lect, given the considerations ofthe pres TT decom, btw wl emt ass tat ds comet, Now i eae ey, fom a Dieta Hk shat of (240), ates th Condon €thewise nul not be sais at D-Strostre. (235) {who that Job, knows does be adie (240) tends ho tat Joba Bows) ‘Compare sentence (241), a sandard Conéton Cwoation. (241) “heads everyone that John ows ‘Furher, (222 indzaes that LF satisfaction of Condition C woud not rare ie LF represettion of (241), falling QR shown is 242), Pa sully sey sila 4 the SSurvctre (and LF, present) of 2%). {242)leveryone that Fb, knows i he aie “The clevant dflerence between (238) and (242) seas to show Te attr nor at Dsieotue, bs acer, only a Stuer. Aer ‘Relea dared ection 123, econsiiton coud bea Bee Tae nul yori tha the Bding conditions apy i Mocks wereuar er epeesenation at wtih they apply ix S-Srstare 884m ing reconstruction, LF ‘Warr apar to Conttion A, we have considered ce dsibution end ieseprenen of reeves Te empty eatery PRO, whieh was ey re soon 1. is ety sisi is istepretation a So serif im astibuon. Conroe PRO general he jot the itr ‘sein haa rete woul ve Ti icy was the metivaion reat ycxsin arly of those eonsrucons ofl Se Chom) reer tesnik19TT Further the einsipesrvat 10 he col of PRO aaa inspection, t0 De iar ¥ thoe involved i he oi ar ancdent 19 anaphors, For exarpl, 8 aready sessed 2% Tear rec ofan ef lase can sce be ound bY aa eu san (249) jst 2 RO ca e ound inthe pa ‘el coniguravon in 240, Pitcipleand Parnes 7 243) Joh, betes hielo becleer} (244) Foti, ries PRO, to be lever] ‘Ai as subject of fit clause, neither is peri. (985) “Yoko, believes [bimseficoer (6 “hin rome PRO, ens ay psec cas Fortier while both ae allowed asthe subject of x noni ea 2 weet the antecedent must be the neat sub up for bok Go cnnannnomenens one a position for an anaphor—direct a 5 = ie ar eee —_—- 23) “John jure PRO Fags, even _Bececnn ein trun psi twin bt PRO Rate 00 and 0, he cide sn ga | ct than itil. Sen in the contrast beeen ___ 249) 040, ononehand an 51 25, on hee. “ ___ 51) “oto beleves PRO ote cer 52) “Soha vies himself o be clever} oo or pce mee ants | See ens (Soe eee ae | Serato te toy of Cues whch nce opi inca La Stops for example. tha _ preg tl ha in aus Ce, es, ot. Then (50) es ted for: PRO is Case-marked. (252) ix stra ‘ one Seer suena tte Coe ii en Cana s component NP bt cent Caer te mere dase And G3) edo se eee Sas Cope crspionlty Cae mark, as seen in (243). Ba there are aspects of te ‘Gsrbation of PRO that cannot be deduce inthis way. Consider (259), (251) fob belive sincerely Mary to te cleves) ‘fbn lees inary tht Mary i cever tw 253) Mery fis 0 secrve Coe, perhaps because of the aceney eguirment on Cave asgnen. Bul 254 0 beter thas 250, (05) Hohe bees sine [PRO 10 be eee “Ths, iter prsebing Case for PRO is insicient Trt samples ofthe sort tt tave Beco widely seus, inate sions dicen of «por Case hoor aecount of the distibe- tion of PR. Since PRO in (255) 01 configuration of Case asin nen ( exe NPs impose hee), ha exarpe ight be expr (0 be grommateal, prey wth an “arbitrary” interpretation for PRO, sin 230) (255) Pie ely PRO to seve se problem] £256) ivsimportan PRO to soe the probe) ‘And (257 might be expeed tobe grammatical wih an arbitrary inter pretation, or possibly with PRO contoted hy Job. given the gett Tock (or at lest antioration) of Coniton A ete in canes with celeivesbjes a iseated in 258. (25) “ohn bees fi 0 Be ikly [PRO to ove the problen (258) John elves fit to be ike that pcre of his il be on ‘spas (259) 260, dese in ton 141 Ge (139), display one fier foniratin in which Case marking ppbabl yer PRO is none theless impossible (259) ty be [ary tobe nein) ‘oh my beet that Harry sittin (200) “my ba PRO toe metas) In Chomsky 198 itis argued that the crac fect determining the ‘stbition of PRO ts poveement. In particular, 261) is llr as 9 tescipvepenerlization aba section 1.1) (061) PRO mnt be wnencrned Pris a orate @ Under the standard assumption tat Case mishng requires government. {bis wl ental hat PRO wll notte Case-marked Bat the regarement feo brader. ince thee f goverment without Case marke, Tf nha ve fn in (258) (259), (257), and 4D, The drbuton of PRO Fehr corey dese. (261) can tel? be deduced from more gener proper, nal, Conditions A and BLIP we take PRO to be stiancouy both a anaphor and» pronominal as sygusted in seton 12.1 i wil then Fol tht it wil never have a GC, sce if it did, coaditory require fens would bein fore, phen that fee enti not hound. (2) ne Fallows, since a governed eereat wil alvays havea GC. The evince it Bo be independent. dence For (pul) a hors, since o exist at al, PRO must evil say Contin A, hy Niue of ving no GC. “Ths i widely ewe as at unforanate, oF eve intolerable, conse- ‘quence, and substantial amount of reach has fcised on retin ‘ing PRO andor proving alerative cheracteniations of “governing ‘ego. For porteulrly interesting cieusone akon thee Une, see. for example, Bouchard [988 and Manni 1983. We sient here ‘at control efferent enough fom anaphor binding tat sparse ‘shin fr anteeem assignment i fi, justi. Conder fst te mili observation that in ion 1 the instances of corte by 8 suljct Musrated above, a eonoler can regularly be an objec. os men, (262) Fokn old Mary, PRO, wlan} ‘Ths fr dhe is no evidence for eistngising. contol fom binding oe binging 00 can bby an bie. (263) Joh okt Mery, bout erly But at iss two difernces emerge on sloseriepcton. Fis, contol is ‘Seneally by a spel designated argument. (Se Nihigauct 1964) (268), wih conta bya subject nse of an oes formed. 268) Fob, old Mary [PRO, to eave) __ Binding on he other hand ha ach constraint in Engi as eon in ‘he grammatical of 265). (265) John, kd Mary sbout hse 0 caper “Thus, there an opsionsliy covering choice of bins that oes ot regulary ens fer ehoce of contol, sinicamtdifeencetetween ‘he two phesomea, "Nowsit wel Keown that there ape languages woke English with expect 16 this propeny’ of bung. In putea, there are lanes ‘abe, apparent. only subjects can be binders. Pls sone su lan [Smee tid in the fowing padi. fron Wii 1982: (266) 20, cpeniads! Maio swim, jeu Johieling Mary about srs ater ‘John wa eng Mary bout his father (267 tan, opowindst Maize. suai, jeu John tling Mary’ ahout sels father “ol wae ling Mary about her fae” “These languages dpa 8 cond iferece betwern binding an con trol. For mile aaphortinding by @ nonsubjec is impos, contrcl bya none i posible (or even neces) jus as in Engh. (268) Jon, kaa! Mari (PRO, saps arta {ohn told May wie aie “John tld Mary to write an “The precise nature ofthe prance dtinguishng Enplshtype anaphor binding (any commander a the Binder) from te Blt tp (nly sub- Jjvtas the bind i a fom Gear Bul what oes Sem la is tht this premeii dileence doesnot cary over to cont. For this nd cies ‘eases thee considerable eidence forthe exsence ofa sins com ‘nol mode inthe theory of erat, tide! 143. Cove Theory a some languages (Sesh, Lan, Russian.) Cases morphol cally manifested. wile neers. it has Kile (Eng, Frech,..) 0 0 (Chinee, .) vert realization fe ite with our general approach, we sue that Cis always present abstract. Fo sominatelacostive Tanguoges, the abet of rite aus ir asgned arama Case: the jest ofa transitive ee 8 asigned acute Case (th some pare fcc and leva variation, as decused by Freda and Baby (198). ‘Nestle (1988). among ethers: ant the obj ofa pre postpostion is {ssigno!obliue Cone (gat with substan vation. The base eas ‘of Cine cory Lew cutef the avestianton of the dsiibaion of overt Prnsples an Parcs im [NPS tbos with morphological conten. Chomsky and Lasik (1977) proposed a set of surface fiers to capture thi cininbution, But ‘erent (982) observed tht mest of theless could be nied i Cae assign cated jon bose and f Case is required former plologiel ealizaton 2st in (205), the Case Filer (28) Every phonetically elzed NP mut be signed (abst) Case. (Chomsky and Lasik’ ers, and Verprau’s spleen, wer lags ‘concerned with subject posivon of ifn clei, By aod large, lesa! NPs potted ts posion, 20) Hi ses Sno be Pee] (201) Tam proud [4 0 Be Bers} Feit counerparts of thee construstion ate pose. 2) ise tat Susan i ere] 273) tm pow [ta Bier) “Tiss a predic, since in (272)-(273) he aint NP i ase emintve Case, wie no Coe fn for the corespentng BPs nam) 270 Get ey egos ar perience fhe eal NPS in €21)-€ In Gs) we te the ae of ant Sn, nea te [NP Sua ite an in 75) we nd PRO in pla of Bl (2) Suan sees tote ee) 251 wm proud [PRO tobe el Toe, os dcused in sion 12.1, 4 the Case requirement tat foes ine novenenprodicng 7) lnm an underving stir ke 0) (20) ten oe no be sais at Sire, bt thr 8 Condon ona dened lel of eseration {G70 dopa acter conration peng PRO a subject of an ite whe dsowing sal NP (2%) a il ed PRO tobe ere) 3 steed Mary tote ee 4 rising conse with she complanet of ty, we fd jst he ‘Teverse behavior with Beiew. " ‘ (277 4. “Bites [PRO oer] 1 mittee Mary to beter] we Chapter |Asseen in section 142, (276 ers (272) rcives an account in tems ‘fbn theory. The CP comperen of ei bose to government fof she suet ofthe complemen, so PRO is aowel having no GC in this coniguation. Under the asumption ha elie. lexi prop ty. ake jot a9 UP eaplement, PRO in (277) governed bee as 3 GC, Either Contin A oe Condition Bs then neces violate. However, (775) no et explained. In at example Mary ot the Suto of nite case the objet of transite yr rte objet of preposition. s (269) shouldbe violated, Te fat tat the example i Icepubleindiates that Mery does resive Case; 278) incates that that Cae acest (or Oblique ther han nominative (278) it eve her (she tote he Frher there is even thatthe Case assign & the matic verb Feline), Pebaps becuse ofthe meager overt Case syste in Engh, ‘Case asignment generally confrne to. 9m adjceney requirement. a= iseaed i 27) (279) a_i sincrely tbe Sem wl Beeved sncrly Same “The sare reverent emis Sl with respect to the sujet of the infil complement of bf. 280) a, i ice belived [Mary to be here] 1h ft bebe sincerely [Mary tobe hee] siden ee can ani acestive Case sot ony 10 its obit (he ore suaion) a in (2784), but 0 tothe subject of Ws inital com [erent pheromenon often refereed to as excepinal Case marking (ECM). Resling that (277) shows that tere ga government relation fn this conigutton, we conclude that Case i ssigned under over Imovt (end, parameticaly, aden), a sight, weaker regutencet than te head-omplerent raion ats coe. We tently ake nom tative se also to fal under government, ths stance government ofthe sje ty the iesioal end of IP (assuming an m-conmand ‘einen of porernmen). In Engsh tell heads Vand Papper tote Case esgners wile [and A donot. Thi why NPs can occur as cet comperent of the former. [categories but not ofthe ater, [+]. eaepoie, despite the ft that Xba tory wow! lad us (0 expel the seme range of Penspls and Paramcere m amplerens in both sustions Thus, while proud can take » atl epee as seen in (273) and (275), ic eannot aka bare NP. (280) “Lam prout my students Lies, while the ver erie takes an NP complement its sor: taletion icon doce not (22), ob ext te theory (283) Hohe’ scram the theory In plac of the NP complements in (281) an (23) we fad an apparent rposionl phrase with email ol preposition of (244) am proud of my sents (2A5) Job's erin ofthe tory sums that of i inert provide a Case signer fora lexical NP that would othernine He Candese seriou of »plecmastic clement to fala morphosynacic egerement rather common proces. "De. ‘ppor sages an inflochoal afi slated Soon V by movement of ToC. asin 86, (230) did Joba tease Bur theres some reason to question such an secount of 284) (285) m pancut. one ofthe xker Case Fler woations enumerated above at be sage ty the ineton of of 287 “seems f Sion 0 be bare (270) 288) Lam proud of ailto bee (e.27)) (289) "Bi wed of Mary tobe ere ff (2766)) (250) ha bene csr of Sim (27969) 2) ative sincerely of Mary to Be here. tf (2806)) ‘To the 271) versus 28) pursiem, with an adectva head of the Sermon ould te aed 25) vert (82H, wha the head (232) 4, Sy proot Join tobe ne 1 “ny proof oft be here That proof ca ake a last compleent i esdenced by (283) (299) my root that on here 4 mer Further would be expected to take an inital complements a ‘option since the ver to which cs elated ex, (294) 4 prove hat Jon see B. 1 proves John oe here 11s important to noe that under oer circumstances “fnsetion” x valle wth poo as ithstated ia (295) (295). "ny proof the sore 1 my poor ef he theorem “Two requirements emerge fom the data examined 50 far. Fist “ef Insertion” tes place inthe content ofa [+N] ead (N or A) ard not ‘there. And second, of is avaabe ely forthe complement of an props ead eis x posible in "etcpsonal™ circumstances This Sores» ifrent perspective on ofsertion. Instead ef of bein i> fered, a5 son of las eect, before the complement of an A oF N. Suppo A Naren ft, (emive) Case asgers, as i overtly vs ‘Nein German (Van Rem 1961), Of eam then be eared the ‘ealaton of ts geiive Casein hi cetigurtion a Explish, Follow {ng Chomsky (1956), we then stings he sacl Canes scesative “nd nominalve, which are aserd Solely in ween of§Stretreconig- Uton, foc fuerant Cus, Snliding gestive. which ae associated sith Bsnarking Tht i inboret Case assigned by 010 NP ony if {marks NP. In (292, then los cansot receive inherent Case from roof ince rsvp Ode rom it Strstural Case has no suc the ‘rate reeitemen, but pref bing 8 noun, has stretoral Case 10 tion Thus, fen reece no Case a allan violtes te Case Filer. ‘Not hit unr the intent Cate approach o insertion, the bss (Case reed for the satisition of the Case iter can be either struc tara or inert ‘Passives ae another ortrotion in which Cases eiemly not av able but where o-nsertion, now viewed as inherent genitive assignment. ‘des not obi. (296) strates this for “exceptional” Case. (296) "i sete (of) Mary te Be hee [Compare ory i Bete be here adits Beled ss Mary i bere “These cimpls show that psalve ver uke a preposition or ative ‘eb. is nota sructral Case apr. The imposibty of er is no Sarprsing given the tema regurement we have sen, 297) is more robin Prcpesand Parts us (1 “itis etn (oN) Mary of Mary belived AAzin,sroctural Cose is unavsilbl, inating that, as suggested in ‘Chomsky and Task 197, panve verbs are not [Ni But sce Bory ‘athe marked complement of Pew, inherent genitive Case might be ‘expected. The fot that i not penal inde tet pusie ser ale nota verb (H+ V, ND x not a adjective ((+V, ND cite ‘athe. tis neutralized [+ V]eauegory with no ming er the Feature [M) Attematively asp Baker, Johason, and Roberts 1989 the pasve ‘morphene is actly an argument reecing the subject Cole ofthe sesh andthe aceusative Case thatthe vet assign. Assuatie Cas is ‘hen unavalbe forthe objec ofthe ver, o or the subj of nus atta complement ‘The Case Filter as orginally propos us 3 morpholopical regi: ‘ment, and wile such » requirement might wel teat eco, thet are Flevant phenomena tha do ot seem erensble to an account in or ‘shologicl terms. The trace of wh-movematgererally mist conor to ‘the Cate Filter, nue tha vial alo the contents examined hes a there lesa! NPs prohited aio sallow swi-rce (298) who dose certo be he] (@95) wo are you proud 10 be hee} (4.0) who id i ry a] (001) ‘who ate you proud (002) whic theory di you underzand the prooty 603) Sto isi heed # ‘Though aces have Features, they tave no morphological reization, 56 (258) -(009 are ueexpected. I ght be thought that i is actually the ‘iepiraseanieceéet of the ace that mst satily (269) wth Case Somehow being transmit fom th race vie the leks ofthe movement hai in well-formed wh-questons sch as (4. (004) who did yousee¢ However, the paradigm is cepcnted in contractions whee even the Moved operator need not have over morpholopnal ication, a in Mee dan 09 oe ol tn coe 06). wes 6 ‘chaee (208) x the san (no ae ster ho} iano ee “inh gorse podbe re “than ed ote re Steen oh) a road Sie they (ou deta he rot ‘hein tele Mayisoo doef forse state esp ort oe he “Bans oo upeplr Oper go oy to be heel Eset th panel eld MP an varies Ges foe tenemos lve suet Cs, reat r,t el po ‘uj bt npn ta Sanat ls ace Shady sews ote sgt os Fede Term pews nlmemstog, the shee NPs sn ena as ier ha rs ta we tad mig stb Cae Fer eet 0 theo Arun st 1 we ane that 8 gen wee reicaezonen an ese ate Ts Corey dogs ort Novara ad eon eee Ban. tram Noe on te hn Only th oe ewe te esse a he Ce Fr fi te, ort of the enc a ring stan fr Omang oof hs 2 anton corny condone (Lat Resort Mek fer ve tat cnn os wan tate cr oman he Sere conan oy Dts ee dove te Chin Coon in angen chain (ph a 4 Coe potion a 8 “e docsing the Chin Cedion io win 13 we ste two rnajr problrs concering expletives and PRO. The former var ds ‘ose ection 133 eae to ea! with ce Fact That ape PRO appears in non-Cate positions, a fact that apparent compels ws to ‘opt disncive venon ofthe Vistlity Condon that falls shon of 2 tre generation (G07) A chains visible for marking iit contains a Case postion (nce ite Fad o headed by PRO. “The peoblems concerning PRO ate in fet wore serious. Ths, PRO is tie ator argument in that it fore o ve from 8 non-Cse pos (606) rps and Parte w tion, and cannot move (fom a Casemiked postion. fits kf! unex laid ever bythe unsatisfactory dunt (07 “The firs probem is hustrated by sch conseations as (308). (908) we never expected [hereto be ound a isan indetnite NP, he courtrport to (208) b grammatical any Ibngusses and marginally acceptable in English (reso. with “heavy IPS" sich as shrank los by Shakespear) a LE resto ‘the postion ofthe expletive, ging» chain tht sates the Visit CContion. But with = PRO. the sentence i compe cali, ‘thong al relevant condtions are stiset: PRO ooeupc -posion 35 ‘jt of find and choice of arbiary PRO shoul satis the “Seite ts contion." giving the meaing we never expected that some ab leary person would be found’, Overt rang of PRO tothe postion of theres poste sn (309), bt with an ensirely diferet meaning song conta Py ne (305) we never expected [PRO tbe foun) ‘Asa descriptive observation, yt to be explained, we conclde hat PRO ‘ust move from & noe-Case potion at Stace while other ake ‘ments mst move from such poston citer at $ Structure or at LF. ‘To ber (38). we might appeal to the require that PRO un severed (eesti 142). We must however, ow assume tha hs Surdtion apples at § Struts if the condition lls fm Cond ions Aan Hof the binding ther then thes foo apply tS Stree “To account for GO), we migh mily Lat Reson o permit movement AT PRO feom a governed postion, Both te assumption that binding theory applies at S-Strcture ne the extension of Last Resort ar open to question. Furthermore they are ‘smpirealy inadequate, beats ef the second pote: Hke oer aE ‘ment, PRO i not permited to move from Cased postion en to ecapegverment. The problem i usted in sich Forms as Gio, (810) a, arora about p Ba tostnke [lat the prbles are isohb) © eto seen to Bhat te preblreae insole Suppose that (30a) = D-Strctre in the content iis le —, with | ¢and fi —John, Last Reson ts raising fw poston ying ue pte (6t4a,tecase the chin Jn already vibe for @marking witht tnovenen, Suppose P= PRO. On the assumptions now wider onsicr- ion PRO must rose tothe postion 10 sasy the nongovernment esireren. Bt that mwerent is permis, even though is a Jeptinete postion for PRO eer contractions sin (1) (210. Feiss Lobo 1 tak about 1 tis eft (PRO to tak about) itis nfair PRO to tak about Jn (Ove might are inthis ease hat there is henry vilaton, the sulyec being an ebigatony Q-marked poston (a dobiows move. sh Tasted by nonsnzatios in whieh external oe sion: ee {Chomsky 1981), But shat srginet wil oe sue Fr (3105-c) (Lasik {992), Here is in a-nondpostion, so that the sentences re well foomed wih x ~ expla = Jol sin (122-0) (B12) «itis areforit vostrike J that he problems are noble bitin rae frit to ser a Jn that he poles are inslube St b= Job eannot sn to the poston a leang trace. 23n (13. (613) «We wnt John to strke rat te probes are insoluble bse want John to sce 1 dat the problams are note fn the ase of f= Join, Last Resort accounts forthe phenomens case hing sine in the tace postion ae therefore baring fuer ‘movement. But suppose that f= PRO in 10), The request of no government fue ovement o ye (14), (214) a. PRO tosrke thatthe probes ae insta) 1, PRO to sem that he probles ar ish) PRO is now in an ungoverned postion. heating a Omarked chain. Hence, all conditions ane stsfed. But the consructons are raically ‘ngrammatial, whatever the conte ‘We conclude then thot the proposal Lo impose the nangovernnent requirement For PRO at §Stractre sd to incorporate tis conion| in Last Resort oi pot sl the penbiem. Even ith thee questionable rove, te disjunctive formulation of the Visibility Condition remains Cnpincalyinadegune, a well at unsatisticory Some other picile ‘avis tut PRO behave ike cer argoment, moving from non Case poston nd Bred from movig from Case poston cps ent Parcs he Notice that thee anomalies would be ovetcome if PRO, lke ether argent. his Case, bu Case erent fom the far ones om ‘atv. arurtine, abd Soon. From the point of view of interpretation, ‘we might regard PRO as x “inimal™ NP argue. lacking indepen dent phonetic. erent ober prope. Acording ft ws 4) tar itis the sole NP that can bear nll Case (heh ay have eh ‘Cases as el. in nonstandard conitions that We wil pot reve ber) follows that Last Resort apie to PRO exactly as oe o any ae sment: PRO is permite to move fom a nem-Cese postion 16 & poston ere its Cae canbe asignod or hacked. andi nt pitied to move from a Case postion. The Vtbiy Condition can now be spied to or. (G15) A chain sve or marking it comes « Case position. ~ necessarily hen by Last Resort ‘iserve fre ha in some lures apresmen ply the same role ‘8 Casein renering chains vibe (Baker 148), Thus, abstract Cae shou include agement along wth standard Case phenomena, The re slzation of abstmet Case will depend a paramere chices fet fan tional categories. Case is relation of XP to H, Ham X® had that a5 oo hs he Cs of XP. Whee he tere agen bth XP st sea resent: where i appears only ot XP, 4 Engish. Spanish, and other languages with minimal oct Case marking apremon ofen manifest with PRO as well x overt NPA a5 i to, whee pretence eer hte to he G16) a tant hem tobe fees sey want fs tobe ocr) © they want (PRO be aces 4 Juaneree [PRO sar exferma} Juan beteves [isl te sick ‘Thus PRO includes tues or agreement, eens of abstract Case ies sms the yi the mame jo nt. 9 futher stp, then. co nuppos that ike other NPs, PRO contains san died Case a wel os green eaters, lem then nll Cae asp er hacked (sume the lot, for conereenes? Recall that nominative Case i stndardly checked in 0 api {Spec Ph, whee Hovolves he features of terse ard agreement (T. Age Tel thus reslzaton of a Speotesd eatin, wih the head ~ I the ead of PI stra he, to ake ml Case to bea eaiation of the same elton where Flicks cee and apreunont features: he eal checks null Case, and the nial NP alone can beat. Mere gener tly, we my esume that the infra men (with ul agree) fe dhe ba! Ing of gervadive nomial cack nll Cas, 50 that PRO ‘appear in sch comecsions 38 31). 17) 2 PROC VP Hobe sek) by PRO Ing VP (hing ik) (One skin anomaly sil esis in Case theory. We ae taking ab- tact Cueto be an expression ofan (XP bead) flation. But we stil fave two distinc lations of hed to XP ving ssl with wy aati factory disjunctive Formation: wile nominative (snd now mul Cae the relation of Speed eatin, acusative Casi assigned by V to an NP that t gover In ican the miter ener we extended soverement to comand to incorporate nominative Case asinment, tut aport om te Cate raion. ecommand appears 10 be the appro priate basis for goverment. I would be more natural to suppose that Structural Case prea the eaton ofa Spec head ration, while inherent Cases wach, ws we have sen, osncned wath marking. {sere by lel heads, We have already touched upon hs possiiy in casing the infects system in section 1.3.2, where we 10k it 10 ave the form (318) (=(7)- om oe ke JS ay Ws € Re “yy sre age We hii an Pacaters Py {As before the notations dar aod Ar ae menor ther is only one flement Agr. a colton of fears. We contin 1 omit a posse [Spec T] and negation, and to assume that at D-Structte the sujet cies the [Spex VP positon. Recall frter thatthe V head of VP smalgnmates with the heads ‘Aato.T ed Agr: and atleast by LF, V wit aici 10 ‘innate all aces nt ecommanded by thar steelers, Verbs tay oF ‘ay not have the aby to asign Case, whith we may asic to be Fntiatad by a 106-alued feta [Cas] for accusative sa enacusae ‘bs Pevimutter 1978, Boro 1986) EV as [+ Cas the the ame an [Agro V wl alo have this faare and wil cask cctv Case nhe poston Spee. Apr if V has | ~ Case an NP in Spee. Aol wil ot have is Cos checked end mast therefore move to Spe Aga. The [aera Tamlgam check either nominative or mall Case the positon [Spe Ags). depending en whether Ths the vale [+r] or tre Sinctural Case in perl is spy 8 mantation ofthe [Specs AS] ‘lation, wit ealvations as Case ot apreanen, depending on lange: purser morpho ‘Ar we have ser one standard kind of parametric variation anon languages hs todo with the postion of Structure inthe derivation of [EF from D-Stuctre. Thus, cram operations that are nee for satsying LF conditions may apply before or afer the branch pont {0 ‘he PF component. The same fe te of the operation that vase NP {@ the [Spe Ai] positions Tor Case checking. Suppose thal he (NP-asing operations ace at LF and the language is tended. with V Fasing overtly to the final position. Then a ote care ection 13.2, we have 9 VSO configuration at SStrctue, V and the if: onal eens having amalgamated and ace of V adn VP in (318), ‘th subst and objet remsining thir VP-nterlpostons, Subject Vil ew Spc, Ags) and object o (Spx, Antal at LF Suppor tat Sabet tang is overt and objec raising vert the LP componett ‘Me then have an SVO configuration at $-Srcture, withthe VP beaded by Vor ts trace depending on whether the langue lower ifsctions 0 Y¥ ike Engi or raises V 1 ineton (ike French, and Englsh ih fais se setion 13.1). Suppose thatthe lguape figs ith ret cet rising and covert sujet raising: we then have OSV order a SSinvcure (rambling) If bth bjt ad objet rae ove in 3 "ihesesded langue. we wil til have SOV orden ut with eas in ‘he orginal pesitions a VP Otereptions re ls pssibie m taper “Te parameters vole ate much ike tse tha diferente Engh type langues that rie overt rising fa question phase fiom Chi ‘eye lngunges tat Hevea sich pass St As discus in| Sston 13. ake se econemy penipes to pref cover operations, wich doo ed he PF component, o event operations that do. Hess, Tries longunge rire that ewoverent be overt it will pp a LE, ‘in Chinese ype interropatves or tiple wh-phases in Enh languages We might eure tht whats involved condition oa S- ‘Swuctore Spechead sprement, were the ead i the C to which the ‘replies rises tat condtion on Cau, in the broad sere row Under consideration. The conditions on As and AB are sia. Only 4t$¢SurcureSpectead agreeiont (Case. othe broad sense required Ja overt sing permissible in English, for Agr bat not Apso. For 8 Formulation ennating the S Structure condition, se chai 3 "This approach, which reduces Case apremen 10 w eton ofthe Spee head ration. reqies ha ve ody the formation of x amber ‘tthe tic principe dissed earl, whe leasing their content esen- Tilly intact. or example the Last Resort condition For movement and the astcted Chun Condon, Coder he DStrctares 9) (019). wettine fe 0 hae yy John won the eeton we bles fer ave [yp tee led Joho ‘Assuming the VPntrmlsicthypotess, fn is within VP n (3198) fe st raise tothe sobjet positon e, ms lo in (2198). ing the SSirectre forms (20). £620). we ie ont have fr won the elton] 1 Ne belive chr o have teen eased ‘he siandard went, reviewed care, expisine this in terms of the Gain Condition, suring en Stace requirement on Cases tnont Movements a kgsimate st sor operation. ‘We now cennot appeal to tis angumeat for $Srcture movesent “The problem is thatthe Strate fores (320) sil do no sai he Gain Condon, esas Case checkes only athe LF repeenations ea. (G21) a welloha believe to have won the eeston 1 we ot bev 0 have been ected “This is one of clas of problems cating 10 the sujet poston [sper IF nor-poston ha eam be exci her hy am argument ‘exlesive canbe pro, ithe language permits nl aets. In sch eae. ‘nsloges to S19 woul be accepable in principle at S Strctre with ¢ Tein pre, assuming the stisfaction of ether cetions (ie indefinite ‘es condone). Then LF moverent would replace pro by its as ‘nein the novel fashion, Note that these problens arose in a diferent way inthe standard ‘account In part te problems were conceptual: the stndsed secant ‘wag hosed onthe dutious assumption that Case most be cect at $- ‘Structure, though on conceptual grounds we woud exper he Visbty Condition, hence tke Chai Centon 10 apy only atthe LF interface part the probles were similar to the ence jist risa The. im the eonstrtion (32). for example, the pase an err is aed at S- ‘Sirsa evn though the age poston is no assigned ractral Case; {Wins checked (oratigned) oly ot LF after expe epleemen (€22) there was an eror made i he eakdation “Tae problemi smi othe ae weno fae inthe case of (15)-C20), ‘The EPP (ix suction 1.32) requires, for English th the (Spee 1) postion be presen ihaugh the couse of derivation hee occupied by expe at DStuctare Other option) postos fe. Spee AD) ‘aay be asuned 1 fe eerie i he cone ofthe derivation a pat of ‘he movement operation te nsering a frget for movement in ma ‘ter conforming to Xr theory. Where the expletive fist fos ‘sly the EPP, it mus be citer pro or = vaciohstrgt for movetent Foglsh lacs the frst option and rus therefore accept che second: the -stcuous expletive, which only = age or move ‘A vacuous expletive, being only target for moverent st be l= te “as Soon as pose” Esther i eliminated by the very move ‘ment operation that inser i a8 target, om Hit wns serie at DeSuvctre to satisy the EPP, is eiinated at the fire opportunity inthe couse of devin, ence surety by Sirus, inthe sours of ‘te application of rls from the most deply embedded sructe 10 the higes estegory. Indirect, he, (320) asthe ony on for English. Ii neesary 10 ete this reasoning to ocr contin at exhibit a simise range of proper, a mater that requis «cles ‘oasis ofthe notion of economy and the stats of expletives. For di ‘usin within acersideaby simpli framework, ae cheper 3 ry nner “Turing 90m to the new version of Case theory, We can acount for the fact that sng tes place at Store in such consetons 3s (019) Ar! since Erish does not require $-Siucture checking of t= Saive Case, vert operations cana For (121). I ean to provide 2 ‘ew interpretation ofthe Chain Condition and Last Resort confer {othe new asumiins. “Thee revisons are srighforwend. The Vsbity Condition took ‘Case (nw ining agreement) to be & condition fr Omatking. We sunt before that ths wa 8 ondtion on chains (he Chi Cond- thon). We now take to Be contin on lake coins whee kd ‘hain i formed by king vo eis C, and Cy oF 23). we y= By (29.2 Ge Genet) B= BiB) “The new inka cain Cy, bended by sd eminating in ys the LE tlhe that must sty Chain Contin. Un the exarpes (319) (221) we ba she linked chain Zoe 'f) t LE, in each ese. The 26 ‘count canbe simplified Frter in way hat we wl to explore hee “Forming now to Lost Resort te ntuitneeoptent was that aeration shoul be permis onty if they form lpate LF bets. We no {sla that feurcment. faking am operation 1 be peousble iit sa rere to the formation of legate LF objet; hd the operation ft taken place he deation would not have Been able to fr such an Shit. § Structure sing snow 3 pensive last est epration b= ‘aun, were it not to app, tbe derivation would net il eptimate LE jects inthe cae f (30), (372; he ater caseiccates ht his ner ‘retain of Las Resor yas already neesay in the standard acount. Tn presenting the stander acount. we noted hat the Cast Perf et ssid at D-Sructue, But rar a condition ona derived kel ‘Of repreentaton. Apart fom expletive costuctions. thet eel was S- Sirucur fr English, We hive now moved to tbe concept prefer He nsumption atthe Case Filer i satsied oalyat heme lee S-Siruture movement, where required follows rom te eccnomy cel tions, the EPP and propeies of expletives ielaing the null sue erate) Tevemtins tose mur other questions (Se chapter 3. But he Basie sxrctne ofthe stem i rensonly clear. andi flrs some prospects foc unifying the properties of Case theory ad imegating it othe gem fa amor nara 4. Peis an Paani: bs 15. Father Topics ‘The review above sheishy and incomplete at eves many important topics virtually or completely wnmentioned. Auber of examples have ‘een noid. amos them the situs of morphology, a uesion wih ‘bead inplcations, however the probes ae sett. The dsenson of ‘he computational system i also ruil to trom in tha encludes the PF component. This restriction of spe wot ony ns sor tops (6 Chomsky and Halle 1968, Goldsmith 1976, McCay 1979, Clem nis 1985. Deland Elena 18S, Halle nd Vergnaud ISRR, aong ‘many eter), bat ako begs een quesions a billy noted eal, ther reopen questions as vo wheter een operation ud pooped sme have apne tothe LF component eno fat tong tothe PF onponet (ton 1.3) ‘Siniar questions arise ateut the actu “division of labor” betncen the PF component andthe overt syntin. Conier for sumple He parallelism requirement” cal t PR that hols of sch expresions 28028. 424) Jon sid tha be wes looking for a, ‘was looking for eat ed Bi sy tha be ‘The fst conan ig sevrabnays ambiguous Suppore we sole the ambiguies in one ofthe possible ways, 8, by taking the pronoun 12 refer to Tom, aa interpreting a or nonspeciicly, 0 that John sid thot Tom's quest would be asi by any ct. The constant PR re ‘quires tat the scond conjumet be verre inthe sme wy 8 the ist—in ths ease. wth he referring Yo Tom and cat nderstond noe ‘peat, The same rs ofthe ipa constrain (229, (025) John sai hat he was leony fora x, and so id Bi ere tco the interpretation stisies PR (Lasik 1972, Sa 1976, Rist 19, ‘On our assumptions so far. PR apis tothe LF repreenation. 1 025) is generated at § Strate, we unt asm ta some LF process repcrerates” something ike (326, whic i then sujet to PR A. simple erative would be to deny dst (228) s generated at SSuuee tur. taking it 0 be formed byw rk of the PF component that dts ‘he bracketed materia in (328) to (orm (325), neater Versions of ‘eneraive rama. That akeratve i strengthened by oburation of 16 capt Trpeoniae Scr colton, id we hed ol De once if the ‘ae ser porte "wi plains cre hat wed eve aes eproperySetered sno he words of he sentence parted. Note root mp aeppentonr av bare yt eine at icy ced fsbo atc a St foun Pte’ prac ae song and weak cig ben Te [1 fn] chow for | Ged) mong ae he [fe “hols Gn) a weak. Agr song in Fesek weak Engh. The base fats fol. wth some esto of he dt eck tc that eave sg of Engh wee sey much He French, sapien Pes at chong tbe Apr prc ed 10 {tecalr fpherent ht lem then tcc ‘tren Sowe of te forme set Dru Gea er ap, {Guctyim Pron a the Engi eure Other Ferme rect he numer fang oY 06 Ag Fe ay state. Peck pins Ser unt treatment ofthe comprsve dvi the ay oF Tats ving tres, neon and advet, vrs nd i fhe ree rca enanon of Tere ané Aeron ees SXScpemte sac ei am bat lee of epson tana D Sure Toe ange, he conte rove super for the rig Xr thee coon of single bends a te co Stove vincion tween Ag! a8. a fo the diincon bw ‘eand Sire repraentton 23 AMLesst Bort” Account 21 Minimizing Desvations Tat us now set how an asalsis of thi mature would bear on the puieines we have ben considering 1 wl put aside the elation of S| Swuctre 1 PF end B-Structur to texco. Thus we are considering the tony of Duation td Repreition vs sation anang D Sra S Sutra LP, For exposiery con evens wil reer to the ston of Dee Sitar ator at {Gre he consequences ofthe operons reiting thse lec re om only rtased a FP Te aay f veal infton ont i sion 22 ee crcl nt pice hat aig ney posible, Th wold Thin {ran te asurepton tat sorter deration ne alays chosen one Ibe ns The reso a lowering a an fea ne I init cof Engh uue verb pean proper chain Ge ey ‘eee If adn to Vat nt oem VT] ei ithe tueeof In, wich econsands Subset LF ing of al {0 te postion of terete rege cate 4 pepe chia, The Jou esx) the sane ss wold hve hen chewed th the Sts desvation tat votes enya the oer sta Te fas by et it condom, oly he later ere ‘A coe ok shows Uhl Ie has ft conten eect ede simply ote mater of counting ses eaten, Coser Ensh imeropaive: Lt we sume hata interop onto has te enmpleenis Q (wl) to stngush at DSincta tom the eneponng desire, igening he apr nonatos sr lure at PF andthe proper mteretaion LE Qi fates an finn must “cette ove tates by Nei The DsSue representation (1) wl yey lowe, §Srce septation wh verb [V-Agr-" nd ccs nthe poston ef mae. (0 om Agr i books ‘The resulting form i indisngushable fom the dotrtive at PF ands funhermoreiegiumate ft §Stroctr) i Qa real element 38 pte Inc To perm an ouput from the legitimate D-Strocure ruse os (10, Engh makes us of the daanny cement do tobe heal, 0 that lowering does not take place: rather, Agr an Tain to de ‘Let eal this pceeedo-sopport, «language apie prose contin. Een upon the weakness of Agr for expos parpose, suse to Be 8 rule ofthe overt span inserting ds the Modal postion. benee = fasion, auracng the ras affes ad then rast to Q. Given ths eve, we can frm Jr write eo rom (10) ‘The sme device, however. permits he leita form Job di write ‘oaks do ensresed) alongside Jl rote oaks, bath deriving fot he at caper earn fom conespancg 1 (10 ating Qn a i option is fot ely vue but nat pay Cbigson ahrr cena eal pred The eas that egitim fom res nly he Fuk ofdovnrion and en. wher he coe or ues ovet oweing and siege LE ang "To fod he ces rete" ft” condom ast Be ier ected stat UG pimple ae pied wherever pele, with = ge pera ke ony oe» DSimatereesation Jeking wr expe interrogate ferme wit mo cr nn--maring rie iu, UG pe dan fe pincee We a tank of the, nan EEonetied ron he soe cameo agg, whch eat P Gonsde 9ow & nepheexroion wih he DSi rereents- sont (0 tnt Neg Agr wit Hooks The cont derision ims do-ierion and ng O grt fora TRecomls ver do Apr ithe © Suction (2 (1) Jon (rot wt beh tu agsn ve eo rola: why doa {owe 10 ABE tbe © V, Selig te comer verb [V-Agr asin the azar otha Tsssmectre and PF we fave en nor wee (wre oo? Then LP ‘ang wep tinting the improper chin, sacle inthe a ite omega court. Ths pe mse ony he UG pe Soa overt Tong end LE ting seng henner ‘tel doseron lis theo no ony «peri Saon,b ‘Secu roped by these cto endion ue ‘Rrra sluton tothe rtlam provided bythe HM The ro co TLE ring eto crons Ne. tus ling the HMC The is therctre on on pina eration the on ang dnc, ch ctor i the ee "Nears thu eog tt gen a wc epson mt save ence api the tes coy evan cht ei tte toanSSnctre ond ata PF outa Dut ser Toner quinn medi ark: Conse the French counter ore he Exp or 3) (43) Job 1 Neg Agr have writen books Eextomy of Deaton ad Rasen “ ‘Here the correct destin requires that the verb howe rise to Aa hen to crossing Neg, yl (1), (8) John has nt writen books ‘The same willbe re of mai serh in French, a in she counterpart to the D.Strciure representation (1). I the HC books the unwanted , Pl era repeies. “The reson is hat the obec Fr rales in (3a) ie not in the approrite overmentfelticn th A he zlation bared y the Miia ‘Condion on government, sine the patie inerveres), whereas in (G10) the cle Has reed to 4 poston powered by Agr. peibaps Economy of Deaton od Repro ta (2c ABIL Kaye ae fee a stout te eo agent fro (ih neve and ch aes Saee e Frencconpertve snc hrs at hy an fe ee an chs na en AS “The econ sry the NP ee cot een in he pest lac eosson scat wit aay So sear eons 6 otis ep se section cacti tke Barking 10 Be wo te ih in Freehorn cet wrk tse bea VP 40 [Spe IF we mg see that raring mnt Be eral he [pojatio ofthe 6 arkng hes, hs npr 3) 2) lee NPAs oe MI Foie ofthe nomic obs ori 10 te pit i (2) fom the Chain Condon he price ws Coe et ote bjt oe in ths ferme Ray aes Without roving te forte congue tat Kane dep, seit hearin spy these tt an Ag poston tees etven andthe V. ad th hs let i fo he st armen chet. Frerotte vere ho "en ie sb arene tbl pon» pve aton Se {reo Aun hci. apy and ihe MP Koopman (567 as intend propo tht pee is evans Serelton of Spe ed rton We mop ras pont rd ack apn what NP aaa en men lain ete he head Apr and the Nei he ead ean ote etn ofthe end eon aj cen the At ‘aly tog ected wih te verb 31 8 Sete hy he pose Se fve fer deus Koopman sas for hat ts en ay {ee hr he propo! ht he ene eam oe Xr eich wo ndeendon tow Goal of se ang ok Sesacing Recpnn 198 x ah Ti 9) Irae aking ao Mele atari tote hn NP peat ane Saal eonpienetnpsthead poss We mii cary the roa ep fr, apoig hl srt Gare every i correlated with agreement Sad reecs 9 goverament ‘ation between the NP andthe sppropritte Ag arent. Tm mje ‘eb epeeentissociated with nominative Case and is eermined by bo coger thereon of the sei othe Aare of Ag (=I in 2). ‘entero nenent tel ith aie Case a ‘asad teow ofthe NP roche Aer etd of ABP. he ‘rapes sono gine to A, Te eats ih woos U1 ported a Sse, ih Case ching ad Cos = isco "Sey a he pops istcindi era with AS stim Abe ten ooh robles eer nan oma cumple ae: (34) Jon Neg Agr it ook . ‘Tie poem sensu doierion a aig of Ae or th opts es Ap tlh sist ofthe HMC hie aig Sa'tncrnve dain wit nee loweig we wee to adopt the ‘toc rua ian (arin Ap om Aa thes A SI'S i actety Apt. which won not rake ener Net would Ta CA ua anscam Leigh pation ef the tre of ‘veto form proper han Thee hen, 0 vlan af the MC segment Toe mar eee pote cued er eh. ence tment argon pre. 226 Economy of Representation {eta teen sugested awh that moves ave only os “tren Te ecefing docu supgstedta deetion mit oe rede a "tt eon” operation appa whee wes. Baretta ne eo weer mre Ceesoppore rion ta ete pope ay Wo mere the pheno ‘fon More pent tiny ea the pence Al wis ny wre ech evr pepe en pee Sed pepe of tansormaenl sor mor prope. te feet ea actualy + prise of UO. Tie inithe ein is fh drhaom mun fe comer at pose thew i 0 HP fous respon The tae coment fT fe, oc fect in toms of sete notions oft ht tings UG pi guage arte propriate ty oe Sion and won Wes faves pale ies elt inp ba 8 nie thaw apparent spe ose neues fac ins ca :onany of Detrton an Ropssatin st Foweulstion. This ita familar concasion etcwhere as well ne that eats onthe natue ofthe language Fac general ‘The analogous principle for represenations would sults tha jst a there can be ao superuous sep in devnations 0 there Sam be no superfluous symbols in representation. Thi eth nine conten of ‘he notion of Full Imerpretation (FD, whch lds that an element can anpest ina representation ony iit properly “heeaed” Let ws pro ‘e58now to ask how his itive notion might be refine nan Sort ‘te mow to fom the satus of spiel toward that os princi of Us. i oud be mtural to expat that Fok at each ofthe thee func ‘mental vel hat onsite ap iterface betwen the computstona sje tem of language another systems: Hence, a he evel of Stee, PR and UF, then “ces” under FL expressed in tees of ‘sontons lating the ax, broadly conrad, other stems ofthe ‘ind trai, ACD Structure FI holds by defiiton ths ve! simpy beng pce: tion ofexcl structure in tems f th notions of X har teary 2 At PE its aniverty take for granted, witht dheson thatthe condsion usin strong form. That is condition on phonetic representation that cach symbol be erected ister farcry and peep ‘mccanisns ina language invariant mane; a represeetaton that ks ‘ths property is siply not considered 2 phonetic representation oti Sead considered “higher-level” eepetsenttion, tl o ke converted {OPT Like DSrucruc, PF is understea to be ded by some version fF. The corresponding notion at LF would be that every cement hat ‘sepen at LF must have a lngungeinvanant interpretation interns of Jateroctons with te systems of conceptual structure and langue se ets explore this ies fore 261 Operators ad Variables (re consequence i that vacuous quanifction should be forbidden ‘Thats angwoge should ier fem ipa fora systems that per ‘it vacuous ghanieaien freely, with the weltonmed expreson “&) 2424)" reeavieg the sme interpretation ww "222d" Formal spies are designed this way fer eae of desrption and com Ptmion, but the design of human bnpoage diferent. Thus we cannot ave such expressions ns (5a) interpreted as olin sae Bil oF 5b) Imerpretcd a some person ket te Choper2 (95) & who Jone Bl thoi Sb see Bil even some persone ‘Sint langage permits such tracts as (30 the vous ope ‘Mor inierpaation sexe (06) who dt May eis 1 the man tat May si “Thee expresions cannot be interpreted to mean ‘Mary saw the min Jpauch tat Mary sw i expect. A seme teary a eraimar sp tates space devs and rls tobe sch construction and incerpe- ‘ons, weconlude that the wrong theory is peverating expressions ld structres too accrnely and terete inoret. The nothing aradonalsbout this conluion The unvanted constructions ae &- Utded om geerl grou, ters of the everarching coon Fl there wo reno f0 suppose tt the mecanisis of language include Superfueus devi and rs to achieve, redundant, the sme cesuk in ‘pest cases, Sal, the phonological component contains no rues to wens ei cass of pene! properties of users phonetic oof ‘onetc epresetaans “Aeleted question bar with ce variables. What isthe stats in aural guage? Typically. formal sens permit well ormed expres: ‘Sons wih fr arte, interpreting them ws university quanti oF fh the fee variable teat as an erbirary name, ain the course [tetra deduction and intiive matbematis genes, One natural Tanguage analogue ofc variable woul! beam empty eatery bound by sn empty operator. There fit siong evidence that ch centro tions ent for ena, enmpes ject consrstion such e837 G7) 8 fonnisto0 ever teat 1. Jon st lever to expect anyone ocak oi sto lever o eet anyone who esht Mary epeced Join tobe too clever to catch “The several properis ofthese ant many ote constrains oiow hom the ssurpuon thatthe undeying D-Sructure representation i finn (i) lor Oa) se shat en-oferatr moverent, een he tial eonditions on A-movement bes the empty ctezory Op the {Csin of he bracketed claus (1 the sper postion of CP, ke sng re in he Stace epresetaion GE) ‘ecnony of Deraton at Repeeaion (G8) Jobim soo deer oy PRO to catch Op} John isto clever fee Op [PRO to cae} But variates are subject othe propeny scenes led “trong Kn ing”: rable mst have renge determi by ts resid quarter (lenovo permining v0 unrestricted quamieation, a6 distinc fom Iypca formal systems), ora elie Fixed by an anccedent tha ets ‘rai tral proper: thus Jen but not Afar in (37) Te ater ‘nition applies when the operator isn empty category (3s), for example, cannot mean that Job isso clever that be ens atch seth, oF hat he cannot each something (srncone} or eer, sah ‘posto John ee, meaning that John ate something or ete. In short, language docs not permit free sanclor: the rong binding prone) dermins the euious sensntie proper of these contractions. We "ah fink oF his conton mm pst spain othe UG en 1 thee terms, we would iotrpret the empty operator binding. an | gp pronominal inthe sense of Huang: (1984) werk on Chines 8 “ret” hat ts recs renuratedThre aeem abs sich 5 PRO and om. wih horse says oar be Ss oper seca, haan oan al ena hil oso werk) Than evar ton is disallowed. a 262. Leghimate LF Ekmeats A Ture sharpering of the coodion Fi sggesed by consideration ‘of what counts a prope element at the LF level. The question here ogous to the question of what counts au phen hen a the DF level, Each relevant clement atthe LF kel na chain (9), perhaps 2 ‘ne membered chan, ©) Oh.) sem thatthe following ekanents are permite at LF each etn he er os 1 Argun ah cet in an Aston, Cems nd ‘Smarked, in accordance with the Chain Condition. * 2 Aan nh demet in an pon _ 3 Lesa laments each clement is in an X° poston as chapter? 4, Predicates possibly predicate cai if heres pia sng, VP- ‘overent nove span.” another cascs. 5 Operator-vansble constructions. ech a chain (2. whee the op Craton in an A-pontion and he variable isin an A-poton “Teese the nly eens at oo ave an eran a LF Suppon, the, tha thee ate the ony elements permitted at LF. ina Sordane wit FL Then the rie Affect ay apy (and mst apply) ‘ely to lsc an eet, pena hate ot, We conde thet Agrirace (and perhaps the trace of (fn) st be chit fit Varo nay ot be efrinated require for he prope fncion- Ing ofthe ECP the argument ketene is coe. ‘Consider scesive- cycle A-novement trom an Apion. Thi sill siaes chin tht not legitimate object: it 8 “hserogenecus hain” contig of an adjunct cain aed an (A.A) fai (en operator Vurinbe consoction, where the A-goition i occupied by a wate. “This hetropencous chain can Become a timate ojeenamey. 3 ine epraorsnibleconstrction—onl by ehminting intermedi Sie A traces, We conctde, then tha thse rust be deleted st the poi ‘tere ach LF reprctaton® cna ete Aes Tormed by sucessveeyele movement from an poston need not (eked. see the chin fred ir aeady «lepine objes~-naney. Sciancts since they ae no be dlted, they may not be eed by te least efor prise for derttions already discussed The sane i ti for Ain amet) an Xchais enel enert) On ese tarathough ofcourse not Igealynecessary—asumptions we de Tein eect the base pnp fr trace deletion stipulated in Lasik nd Sites theory ofthe BCP pow consequence ofthe peer cond Son Fl wih "ay dete satel 10 “must de Thre are firmer comeures a resin gutons are wih rep 10 ‘pear of NPs which shares some progerts of Aposton end od ‘repetesof A-postions but I wl pt purse these matters ee 263 Flan Eapetives ‘Consider finly he sates of epee elements, such a Enh sere Thalan¢ or ther various cutters nl ver caer lmaunees “scent cvs espettion ae herr ot etd 2 Iegimate LF bec Mt therefore sonshow be removed, Ese ane seested that sere i eisiated by LF substitution But hee tony of Devan and Repciaion ass ‘ss spn eur. and we might suppor on thee grounds thai undeetble bythe condition oo recoverability of dktion yet to be precise formulated. Then we must eat sere aan LF afi someting ust adjoin vit “The enltve cere has thee sem properties Fist, an NP mst 9p- a ina cetin formal ration te ser the construction: tw cll ths elementhe asocie ofthe expletive sr ake the expletive to Be Icensed by its presence, Second, number areren ot with ere bu ‘athe withthe associate. Third, thre is a aera orm with the ss0 ‘Se actaly i the subject potion afer ever ring, Thus. we have (40) wih te sssocinte in aes, but not (4. (40) 2 theres mn inthe coon 1. there are min the oom emanissn the room ay 2 ier was decided to tase by pans b.sthere is ult ha anyone wil pce “These poprts ae rather naturally expired onthe assumption d= ‘ving rom Fl tha the expletive is an LF af, with ts assoc aden fing t0 Since there Inks inherent ¢etures (lading mabe wd Sategey, these features wil "percolate rom its ascites ‘anpions. I agrement is checked at LF, then i wil eid have to hve een esablisied at § Sirotuse between Ast, en the asocae of thee sn (40-bit he observed overt agreement This analy fis ready into the fmevork already oulined. parcial i agree ment snd Cose are ete isthe manner sags: th esgned by SSuvctre sine they may appear overly, both check at LE since "hy have LF conseguercs having 1 do wi visit the Case ae) 4nd the Chsin Condon." If we assunse futher tht the spe of P (haf he peuttions of seton 28 ate corres) must bean NP with ‘features matching Apr tent wl alo fellow hat the axel mas ‘ean NP, end ish NP hat ass in overt syntana in (A). rao (1986) anes Farther that the expletive ea cei wl ve {osatiayadetitonal conditions hoking genealy between a cic nd the Postion asecaed with it, specifielly. avery retitive leat con tion that, he argues, hoks at D-Sirrure; om this further essurpton, he erives an ierestngeatge of phenomena that diferente Eas, Ita, French, and Piemontese expletive conetctins On the peter sumptions of the PAP approsch, we expec t0 al that expe eis 6 co constructions of his ype have the sae hai ropes ars anges, ‘ith dereres expeabe interns oF the lexical properties of the ee rensinvoved For sich reser then, iis phusble to assume that shee and its counterparts) eat LF afi e eed by Tr (as) LF adjunction of the associate tothe exleve yi the rae (2) ns suet the complex cenitting NP by peeolton. (42) Ug thee Lea ra ‘Otter wethetabiched eis conspire vo guarabie thatthe ony ee ‘ent that can ain 10 the expletive he asec With He apprope ae properties ‘Gie tha he mat have an NPessocite flows tha some ther cxltve in English ie asoited with ates, as in (3), comtasing tha (43) a, twas dct to rave by plane is unlbly hat anyone wil ace 1 shout therefore not be neces to stipaate Csbutlorl condtons tm thre ard expletives hr eouterars notes languages when Theis popes ar considered s* alo flows tht at -Srvtore a expe ad its assoite A nt sty al LF chin condoning hee ci AA-E 4D UL Given the Chain Condition halting at LF, st ust be that‘ Stratos the eapleive Eien a Case-arhed potion and the ass0- fiat A ina Oponiion*” Furthermore, we asiume thatthe bing theory holds at LF, then at $Survcture A and E must Be in a rekon {hat sats Corton sine at LF am antecedent race ation hol oftheir §Stucare postions Sinialy, the ECP, a chain conan a LL wilhaye to held ofthe expeve-asccinte pit § Structure. These consequent largely descriptively acerate, susteated (4). (48) a. ‘there seme hat a man isn the rom (ECP viltion) tere ste tat Jn saw a nan (Cordon A woltin) Sine, eter conditions om movement must be satis. Compare te examples ie (5). (45), here ws thought ha peruse man were om sale ‘we though shat pcre each cher wer om ae aman wae thought that pctoreo Were on sa) cosy Destton a Reproenbtion 1s ‘Toe italicized elements are properly eae in (5, hut noo (450 oe (450). The pram with (25a) sn the Bnd teor. as AS show but rater contin en movers (ihe ECP as we se er Se), Sich properties of expletives tow fellow frm FF, without farther sptlaton, Note tha tno follows that the binding theory inst apy at LE, whether ont it also apps chewbereGnchaingS Stuer) i ‘Another cotsequence hs o do with Condition ofthe binding e= oe, which requires tat an rexpresion, such a the sssonate of an ex let. be unbound. A lone standing question has been hy thee no ‘Contion C volton in he ease ef am espletve and ts felted asso. ne Bot we now assume tht the to simply have diferent indies? ‘Tec, therfore, no reed to comple the binding they to ache thine as ina numberof propose over the pest ears, ‘Cerin probiems of cope ofthe kind cused particularly by Edwin Wilms so are overcome. Comir the sentences (6) (46) 2 Vhayen mt many linguistics students 3 there arent nny ngs stants here (G6) has scope smbigiy. btn 466) many unabiousy has nae Fou scope: The LF representation of (482). (47) bother, mary linguistics stdeos] aren hese many figs sade were erally to replace there, is would be apectd to have scape ovr ot, bu in (47) no festion is eaiaked twee the two and he ape of mony a he assed tobe ato im plearesof many students ret ere ** 264 Farther Questions concerning LF Raising ‘ere i one major exception to the penerbation that the expletive E and i woocinte A are a banding theory (Condhien A) raion th __ SSirecture— rarely sing consrction such a (4) (iy *shere ees fo mon to bein the oot Here the expletveasscite ps satisfies all etn condition. but the | pmsion rung [A rare explanation ofthese Its provided ty Belles (1988) theory of pactive Case sient. Taking paritive Case tobe oblique, ‘erfore Grelsted in accord wth the uniformity condlton on Cass a ‘chaper? rssgnment ee Chomsky 1960), parve Case wi not be assign to ‘he associate in 48) bu wll be property asin at Sree tothe sociale of the exleve after unaccsatves an we mes assim, tl, sin thee erie man here i man inde roan. Asse ws efor that Case ust heaped 3 Strtre, pan that apse DF and is elevant at LF. Ten 48) *, since an SStucturecorition| 4s slated. Note that ever with hese assumption. till flows tha thee mus bein a Casearhd postion, by the Chain Conetion, wich ‘equi that an LF cin te beaded by a Cse-marhed psition this ine of argument score. there can be 2 proeess of Case transmission, fr tht process woul allow (8) satisfy the Case Fier athe, Case mist be signed at $Siructure daclly by some Case rmarker or otter device*® Lass (1989) ebseres that stir cone ‘son fliw fom such example (9) (49) consider there Bea sttion) 1 sTeonsider here sfution] (analogous fone John salient) (9) mst be that ass Case ety too Sli: re ak recive Cast (rom consider, so that the Chain Condition 6 said ier LF meg. There i i seems, no Sitar prooes transiting Case from the expletive tere to ie sence, the pene soi these exampis. Sif (185) notes the exitonce of pir it (508-8) (60) [yg ow many men iJon say that there wer in the oor) bs “fa how many een id John sy that [a ware ithe oom) (sob is standard ECP volations the tae tna postion that is ot ‘emmarked sn Lasik and Ssit's (1964 sense. The question then arises Thy th ot ho tre of (0s) the tee yy, the asia of the alive there, fered by LE movement tothe positon of thers Lasik and Seite theory provides an explanation. wheter we assume LE stitution or. as above, LF adjunction Tether ease th ate fon Js pmarhe by the proces of whemenemet in ever stan and eis this property hep aie to the postion ofthe expletive so teri Bo ECP voltion. Smitr observations hold with regard to Riza (182) analysis of wh-extracton of sujet nln: the sb fst extraposs, leaving capleve pro subj, a then undergoes neal w-moverent ay of Deaton i Reprotation 1» = teasing trace ark in overt syntek and the raising at LF to the penton of the egetne ‘The notion of LF edjuncion eininates much of he motivation for ‘Case transmision theones of expletve-asodate reitons ad the ap proche sll more dubious in the ligt of the obiervations jus te ‘viewed (ce also Pollock 198 and Kayne 1989), Neverthe. there idence supporting Case ransmision. ‘An niet though pub agent for Case cransmsion is d- eloped by Koopman (1987) in a comparative sty ofthe Went Aca Lurgunpe Barbara and langiapes ofthe French-English type Koopman potulites © parametric dierene between languages thet have Case ‘ins (+ CC) ant thou that do not (CC) Barbara [CC] and Engl Frenh,[ + CC} Koopman considers tne kinds of Cove ci, (59 2 (whee Vie a Cac assigner bh (Op... here Op isan operator andr the varie binds © (Ey. NP) where Eisan cape and NPs asocite Case (St) resus rom Veasing. ln [4 CC angus, the ate of V il asipn the Case“iranstied” fm V thro then. In (CC) __nyseye, lacking Case cai se race wl Be unable to asin Case, fo Secreta remo anc Spicmmteaeu mares ee rs pcos Epo ce ope aen pyc Eovraeceacis mennsenee phe masumtetate ree See eer Se rs ESTE Sea toon mene ee ote a age Sears dort etieamerttee a Se een Sect iri ternlerel Sn otemeeneamees oer vo capi? spose oa tame verb hich tows ction theo “ec ruent supe ni pr pusrten Hd ake Flee ‘Thse popes low nth spin tthe rac of Guniie ‘er haot aig Ce sex the come verb apn i se Cao UNeehgnon age. he wet met pea ‘vane te por (To) amar aot to x sta Afr (te) ew sve mo ve len aber, asst Stacorry we gion pin pret ve fave an nt spomet nor of Cae arson, ab scant eve jt hen ec ey a7 901 erie “Can ee von he ata ao roe thei Bete he aseoment fer Ce annason athe vere gist uch pecs Super ne net Koopmr's prams sn the Towing = 0 Seco if te ped geet ape piel parame {Se a ea ta ig saben tern o€ XP dees and Xe ovis ey. We shen om he roe (Ch whch ap N° ere ‘Ry evap mot hve A [9] seem a ict Cas ans Ther mpg er exc Case (-C) deme cmt Spee Slerthat Recess cone se shay (¥C} Bt ate fan eter vith ope ower ther X° coment are +l PEG the prancter ocd fo anal ements eorex tthe ti ono dace cir Prec Pgh ae +h Ion thar a eens hay vr eto se etios Bae [cq mone at oy ese Xero ech ees “ago the thee poets (Sn) flaws diem Babe he trac of¥ bang Chemo sus Cam A for 31). here othe Sprtor cant ie Ce Barbar, bong ~Ch sth we ate Peat olen tte Ca ker (rte iy regarenet fom Shen dene, wth «vale heding» pays enerenbet Sint tat ins he Cin Con, se aks Case. Not ha Se et nov enue tht the eertn regres Cie an tere ‘ovat ent, arr wnt or emp) epatns “Ime peer at con dey (le) Se Bar ~C snc! eC i gs a tc ha ‘tat (whch Koopman sos) wel fr 9c ade y 2° ‘Siw in ton Chena psi, vin he Chan Condon Concent rec be no epter on ve aig eGo “Tas mateo be sone sre for Case rams. ch sa ramen le We owe, Reeve or 2 zany of Deaton snd Reqestation wr © raromly specie paramere ditrence involving Case theory, wih a ange of interesting consequenes. arm not aware of eter convincing ‘dence for Case tarsmision so it may be thatthe property ci be lrizatd from UG. avr of LF movernnt, driven by FL 227. Some Concnions on Language Design Sensing, we hive found evento porte bs asumpins fn ngae doin skied incon 2 the mer pic sun Sens cong the sepa ys st ef Fone nd Aten Gerona those of sabeqat dceon. Ther sured conc | _stvetng at tsk cervaons and repeomations ar ste oer || i form of ia ele ceecion so wr reed tobe rina fy wel died sense with 9 superDoos sep n ena we _spcrtoousstok in reprint Prosedng a the wa nese, wemay hope to rae thas con” eel rns | FUG. Neos ta ato the prncis havea Lin ohare | ad erry cing nthe pce pit of UG such the FCP, _ the itn they and on eels he fom ead spa oe ngage xl ps domed coer we Chesly 1991) he prope of UG, | indo hy ae ae or suring mabe of wap or Gis, snes srt! Goons have to meet “ba ono | Beanie pit of ew of pram poe ht we hve proce Sov en $target re te PF pret Then to trie the stn of we hate to cay cat's moter of peas, We et ems hers aed ow a ere feed Siete rpmasia hed by he tc a we __ he deraton rom ote Lepsetin inn he eed ____ ef cot thn any th dean ro 8. Feo ne ee tr et on ef nol oe __ Sed ther rope of Ln pel, hse computes uy ese Lal Tne ree, lage din aps to fe pee _ fem ning terete pope. tough Cant ere Sl _ ter fom consideration fe‘ tase nsureon at the fon “tala hos at sy the enteral cng enon at ens” with eter xen aay Mesa the proper _t8 he “lest eo” enon hugh tral ad pause ers of enpiriclconsaquenons, provide further stration. The discrepancies teeneennatorabangusge deg and the siracre ef formal stems fonsrued for cumputtonaleicecy may also be Felevan here. 5 elles oter prperite of tur! Tamgunge, such a tbe extence of [Sunt estegorsy whith might alto be expected toyed parsing prob- Tens. Note that che cannot easy motivate the conditions on reonomy representation in terms of procesng considerations, sae they bold ULF. and only derivative at Straetre- Nor des there appa tobe fy argument thatthe arcu proper of language design are neces Sry for fnguaeike sts, Thee ar coatings ropes of rata langage, “Thee ae “comnpuatona tricks” that permit easy determination of the grammatical properties ofan $ Serre sepesentation in a Large ‘hs of cues, broad enough to alow fr language tobe ule in prac tie Bu language design as Sch appears to e io many espeis “ys Functional yielding properties that are nat well apt to the fare ‘ons ngage scale! opon to reform. There sno rea para ere, there f no reson 0 ruppome, a pri. that the pera design of a [Emge i conducive efient ue Rates, what we ser odcover at sre iniguing ond urerpected tars of language desir, oot ei those tht have bean discover thoughout the guy ito the nate ‘cf languige, though umauslartong biologie stems of the natal wei, 1 bk ence Garmce T. se s Dict Sore et tae pt ee hem Stas fata we crearage coment ce ae event at hee en See ome sree ra reed goons ae ceteris ert Bees aera ieee are oe oe ae ee a te ees teeta Ca na Eocesagtieaeemcarecrs Some ioan ee sencgmrane nsec ‘iva to ope apne nee sata Sinise abt Seer cioehe erm! er eee Sn SS EAS es “Finn of Beaton sd Represetion 1 | ere ne orp tas bene ae Ten | SL cuet rene ss |i Lomond tne tera eee | oo Stag aos oe ee | | Seco [owns Cet) tt, cy inn of eo | _Riecont ty etn pct gma Chay 5 Te eal ve ar sorting cae ea ee he psi of ew of | thecsmpueona een. Taking taper ry. wea conchae that rola of «kes! ha can te redid a eee ened a | Seang esac ating cnc in sours pone _ ieee se Wh ese ws og; just how ix hoki ia tbeory-internal question hat I wil pot aside. | © On eeson totus anes ae tnd Pe 0 EE 51 ox teveater somone ae || 8 Ass oer of eta or Xr tay. 1 wil prensa of a, X® fete twee ea. and NP or Xr eh 1 Sof pape cn cong on eo __ sisi codsies op cr sii oer ase pat ‘eis veer Rat th apache arson SP |__ Brvvln Sacre aR papas ose ts ie W | Speimone tires (Soar aetna mes Sree san Sars rae =e ae Soh annisientewtthas are | 12 0» Lor hee onion st $text Vab Res 989. to “Seog bide Sarre eT OW hy sk ui aon lop ia it sR 13. aune hte he trl ramenor of Chea 186, sd esinly fn oso Sao 18H, ough Rs mocone ae cde hat ‘ovconder bet 16, Noe ht he ako wih bes prt redaction fr x ombon tite ECP iat xyes pelosi bakng of onan ee ‘Gsesthat wold te he tp fs prope! by Raz (1990), We nl als et inte her poi ere of serene Xanga Fer nen Sup ort i tne Var Renan 13S Nema 9 15. Se Pooch 1 loo oe ew of th gos tha Plck ‘esa See Ends 198 ae fr mre mee dvelemen Ms AP PONA, 16. Onder seas, here new: for abr Frans, 17 Piso for aan and weak re prt and ape pete fare hat ce ea eth 18. Peek esteem the reper constrsion 2 tel ed of NEP, aig toahigher posinen We ah dk of asad of ope muri, 19. Mor ett he es Vag ABD 12 Thema hom ata dd elie othe etn fe ea io be gent dong ean be spon to Rea wc of parmesan {coon of weik Ape me are atuang). tan sme, SY sean. & [Mga psf ol, hugh continuo eh ten fr epoory es phate deve of cplyng doy cents th rer fur ee (TE, sc plas cone oe cniget on partes son 2. Nett re acpi consequent, Ty a ‘hat the sted ate ned olnpge veg, the UG pina nin et trem Inguage pater propre Sepesve work. y Fb (Son secon guage aequston appr ths coms 12 Thre real in fact emir saluion oth pra eon ‘term fun ana toch oe train ection 2. bt Yl at a se une wl or eben he ter usin sa, 22. Noe ta repo ea a aula of ena pecan ta acon tral etporal fees nt t be ced ih Se ‘Bony nar sya vel rep ae a lees consctin The ramet Chena Se 2. Rel tat wear ening, eel, Lasik nd i's (980 ey fhe ECP, me eoed m Charly Ica Under te hery.y (1D ‘th ae roan of Vt Ag. nd suse don ef Agree Posen ves ECP wlio 2S, Onoshr cass fain srt. Chosty 1987 26. Seman proper eines, then, wou be wien ws proper Themen, et head [te etc hrgiops whee te ech po oon sed pee Gonconte nce “= 2A acu ei the eerreton wos ree ight moda in || he caption wd he reine Saye cet Th eno te | fetch nex om toe) rosin |B. More ec sea hls ecen he wht aA ich | fhe pao roc 0 ti tapes nae sagas de sone ae ot Bikers atone, | _M. Nac at we mit 9 he Garo ey cea,” eh | ane sia y stcraregeie movenc Ti sane we a {Sefuerrnometo i ono fot |BLTis cee of ine out et ong heme | et citnson but no mw pote pest ante aye arent Bhi cite tometer || 32 Koco scoring be oui ote raising ops VP ler Ate went npn een aon crn ope AP 33, Thre fre een obec For cal kd exe fy be roe Sure or ened oe vane dao? Wht ‘Svante clener Andee | 34 Aw sop se speech 1 NP cae a Chor 19, ten sesso igh hci (9) frme ty oven en tbe ne . (Bue aed hina he prints post of yeni 3 An alerenne possi, sugested cena facto ining nce | rrp, tar Veron eae to the PF compen fo | seal “apc rand poly so o(obgoory} LE momen ong Uh ines of mitrretion f te tures answer (Chora THE) ae Oot ctetis fener ae ee, flow trom the emiaions dc above sont opal, win thepracn renew secre 36 Nose at tnt reson esesy 1 me pt ja en and fm 7) thscondinen apts “They moh be prt at caer aps, he sing comin do not et | 2h. ste Norbert Homan eco opr te ape tp “Sits fr te intergeatin of porn diced poss of so po scien Se Ss re ce emp _«ESerk a rnmsatea tlt ears tone pees see hard coe oes ner rant aterm 16 capi? 14, Ssh prope 1 be tpt on the asennad in Cy Sa Ua 17, bt pets hey are eat lore ees Sheed. For hw reton hm, toe Goblet what ins to the exes ‘allo of wih is esate ths amet wat as man the al conf mon theron ia (Ds). The ae cer mars Fr “Soong hs tote tyre (19) eh Get oe) ht he myn rege fr hr ataation of th ck of pareilech gree et wih ce aig i cpl cerns Cone ferme Sch ‘resem tore ans Be aera me ek ell yk agree ewten real mand sm Ba the pea fuera mo 8 on ep Sst sel rh gs sha te wae peed a» rng of ‘Str ns uty Se (198, hgh many ques reas ete. ‘nike pow roger ive the eb and asa so Bure 6 152130. Nee that mtn eqns ta he Wo Kind epee te ohobeealh ent £1, Wess hat Cine ius fo» ategy tt eine ces ‘nm proce a fen ere ove th fe te ene othe Spit cent (AEs he acne A, ening cha (A). Rea dat Andel Edo het sha, erber ses 4 Noe ta tee eps cou be acct forty ution on he di Unb ol epee in Chery ad Laon 197, bt we are now expr Inpibe por, cones pre tat hse te eae {2 Oro iking. io Higentth’s (HE ame. Not that we ems mame ‘he capline wo he eet, might Hevea eine ide 1 To aout oro pp sey. nae ene mari {Femae ine ea plan fhe ha he Siaiovtn ani can) tren man ck re ‘ey wae mas gis ae we a Scepter fa nh of oper dm to on ‘Step Lt patna oh one ct! Meena te pase err mcem pen a he teu wercinte triton Snes hte 1 SStiatg oom erase t ore Se pre 42 Str eos ho of chy Ce" redo Sie eee Samy cen bent ee pontine {6 se Pik fo mest Cn ammn Fe iors 4 Fev tea te cn ep Sey 197 {Kenan oie hr pole Ca chs bt he on 31 Some GenealComieations ‘Language and its we have heen sua fom vated pots of view. One spprcach, assumed ber, takes language to be pert of herr wal ‘Te human bain provides an arty of eopacies that ere nt the tse et wndersaaing of language the guage fry: these est be in fed part spite for ha function and a common human enon Inent ovr avery wide rang of creutenc ane condos. One ponent of the language faculty is» generative procedure fen Hlonguep aneefrthfngwage) that generous sina desrptin (SDs), cach 4 Semple ef proper. incating tone commonly ils) “emai ad phones” These SDs ae the expresont ofthe lnguage The theory fa particular language ists grammar. The theory f languages end he ‘presions they generate is Unters! Graronar (UG). UG tether) of {hel state Sof the evans component ofthe language facely. We "Gan cisinguish the language fem x conceptual ptr and «system of Pate competence. Evsence has been arsnlating tha these inc ‘cing systems canbe sete mpited and dewtomentlly disoe™ sted (Cons 1981, Yaad 1980, Smith and Timp 1991), and the ropes are quite diferent, ‘A standard assumption i that UG speies certin gic lve ‘ech a symbole yaem, often calle a “representations ye.” Each lingusic Hvel provides the means for presenting certain sjtematc ‘This chapel appl The Ve ram Big 20; ys ing 15 Hea of sin Brant eed ty Ket Had Sat ‘eyiec (Combid, Mast: MIF res 193), od pb ae wh vs cpr inforation tot nite resins. Each ings xeon SD} itSEgane of worn on et enh gate fra the Etna Sind They (ES) ck SD soe (8.2). ‘termina the Str, SStomire, Phonic For OP. ‘i Lapel Foo 029 Ine este ‘Some a rye linge re mia mong lege tome slaty te open of dct fy. A werking pte ‘Encrave pment ngage re eet om ser ie Fest tert to frm oe inate ties an hat Ce ve fc esos. n ht pair pheno are rt Scere” ty piles of anne Testo re went {ctr compen Klogra sve eTie wht ene Po fod (reared eon) Pte say he morn er Te fpprcach han neve prov oe tel oe using ‘Toc mehypotnane mae tan pt a aif sting = mee ether recurrent theme has been the role of “principles of econoery” inden te compton and sh Sr hey gee, Sch cor osmumns nse tenn vo fos ie tert ‘pectne tae Sng The inks go enon to eve hat {Teac onde she py flag if rer wdesod “fe unguge ted pesforanee ems ht ea coger ob as forming, ng. ‘Chen. end tbe eta, We con thnk of he SD na compen rcs forse poormans stems, proving iframe Sone the inion. Whe te no cr ses oe ie thet Tinuce epee orate once” do tape find conte teen te proper of lenges tteamerofisise “Te pevormance sens appeal i two seal art pra ond oneaetentons) 1a egos freon cote scons ocak sesame Two the fat tel, ten, ethene lesan C4, pong te Semcon rte erator and concent Spur roy Each nguge teins bet eps Gromite APand C1 lel Te level AP es ner bn be © {cPr testinal rac ef Ca he bern more coves "Arta sna ron ha + ng eos we co ee eae tn companion Teen pb —= ie tht ener in th spent, wih hese _peries The computational sem ae th nents ener deans and SDs. The certo opel uz agen | tenets chi of tens om hese compton a ___ gest pif tra: epoenation |S fr we are win the coin of ira conc eeeiy, st "aif gem oulook snp? UG mus deerme as ot _ psi gages Ima pny the proper of SDs and ete er Bale epreeaes tht eter nto em rp, nu Spey | the nr lols (A, Cth dene et ent te tae _ aL Sin by th ey aco pest | Sinple sign or nese we alee cmmepy nea ne, face tevels 10 be the only levels. That assumption will be part i the ria prosram T wie tcp __ Inet wth n gerne rar ws ned ht he ere _Clistie lod of Fates eect acompante al eee “tec repseiaion n cence EST aprnchen C4 ery _ ie LE On te aon hin vl ore et Ei (2) xn rm Pad rom yo forma pe fons of ound ard meaning, nner tse are deen 0 te Agia Pts of hecompaonal em as evn a Ym, "2 te PF compro Other pr ae vt oly fo a Eb corer Te parts othe cpa es at i ea “16 tosh ae the ener moa tr tht at bean tat th ars may isle empty cages seignd sponte tne, The rte of hee mons aerial tc ere sold ot be ___Bied byereed connections ch cms ep fe aad | “eps apd rom eh wen et hint ny lies sat motel opto hs ea AES, fb fircion marine egerenc (pm Ene dt FAD) os tnguage Uc incon the saan pie |__dth ange fpr vraion, Vaan mat dese shat“ fo tec aogring langue tt ye PLD. Sot orig then, td dpe of aon the PF ean. _ et and nae te esr Seesran arbres (soon _aconsps with pono! mats roger of psn for ‘ves (ftom, etd ety deca rope that ld oh ‘ais rca og he ha pre) Van Be vce a oF LE component woul! ke more proba, sine erent ners ould onl be ite init A nro onc that her. such ‘nation: teynd PF opine ad eee arteries (eh 1 ane {etn ace vriaon ind to ena prt of he on nd ere popei aia es Ise the nly ne compu em gue bone hon. sper om hi te Kno ae. [eras ena ale seine prep bat ems totimploble—es aor cloner of be Minit Pepan.* Tay acneaive prensa! tae guns 2 een says tins sate yg nvr eve reife Wi her uct peers ond noreons UG oie arat Torermoute ie stn nytt his omat consis ‘pif impuge Exchange ih arte se rks thera, ply. controw price lnpuge tcl the te Torn ve sso aes rr east Engi for Sami respec ete onsen os angi Siarics Slo comtanion an neg dre rm ropes te forest ‘Sraeonens “Fe ere rr pris andere (PAP) approach. sane tere, beaks tay tao aking pe toward he ron trot design steed. UG ponder aed jem of pil and ‘ntact parameter. The nggeartelrles Telnet cho of als forthe preter The nn of amma “Steorverto simiated nd witht costuion patel rls Contraco sch arth pe ave ise and pase emi cnt texan afc, cletos ot ykesnens xpi ioudh te ntaon of he pce of UG wt be vaso aan ft With rp ote cpa yer the we ass at Si consist of vari pnp wth ein eset Ferctional Stone and genre oper of te feo. A selon Zong thse opons deri anpage. A gop r,t a ino gute sre (Sa et a pi (2) dra fom Ate mace cle. LF pete ngage aniston woes ‘in Ete sumtarof he laguage sen noting tore exc [bmrine ard PF component ut) I thee pore sen tat ‘Sean on unre (fe i), a i+ Strucure pees in some tis ssc with as SD. Condos ot Seperation tse ering tory. Cae Bor, Oty tn ve eee = Misia Progam m ‘on hold only atthe interface, and ae motivated by prope of the iter, perhaps properly understood ae anodes of ineretation by priformance tens. The lust expresons ae the optimal ela tows of the imerface conditions, where “optimal” is Setrmined 85 the economy contions of UG. Let us take thee ssumptions toot be fit ofthe Mists Progra. Tn arly work, economy consdeations cee spar of the evs ion meric. which, it was sumed, seed u partir iestontiton of ‘he permite format for ale syste, gen PLD. AX iy as p= frst the presumed ole of an evasion matric hav dln, nd ‘hin the PAP approach, tis generally assumed to be comply di pnsabl: the principe are suet reactive xo that PLD sufi in the oral case tose the paranieer ales thot determines tnguage® [Nevertheless i seems that economy principe of the Kind explore in ‘arly work play sigifeat role in aesumting for propeies ef lan [page Wie proper frmulstion of sch primal rey be posite to move toward he minimalist design a theory of language ha takes & Tiguisi expression to be noting other then. «formal eet that at lis the interface conditions in the optimal way. A sil orth ep ‘oul be fo show tha the base principe of language are formulated In tems of notions deawe from the doesn of (viru) concepts Invariant principles detenine what counts as & posible derivation and 8 posible derived ob ingusic expression, SD). Given la ‘Baye. these principles determine a spe me of desvatons and gener ‘ate SDs. cath ap ( 2) Let wy that» derivation D comerges i Dir epiimae SD and easher sft does not Dsenverge at PFs lesan and erates o PF sno: D comrgs at LF 2 let ‘male and cases of LF ifs ot In an EST Uatiencrk, with SD = .e,x.1) (6a DStructre repeesetatio,o an Structure yepresnt on), there axe oterpossbiles: 8 o , oF relation mons (6.7) Ing be defcive. Within the Minimalist Progra ll posites are ‘chided apart from the satus of x and 2. A sil sharper veron would ‘clude the posi that rand have each leptin but cannot be [aired for UG eeaons. Let us adopt this mareoner condition as well “Ths, we assume that dration converge converges m PF ard LE: convergence determined by independent petion of the nee ‘ace evels—not ar empirically nnceuots wsunpton” m caper “The picilescutned are sims restive, so hat he empiri ‘burden considerable; an fii inte argument ray benecessary to Support it exatly the desire oxtome, for whatever Ultimately proves to be the right aperosch “Thess toies have been sted nd elaborated over the post several eae with rls suggesting thatthe meat conception outline’ Tray not be a fom the mar, I had hoped to present sm exposition in| ths per, bu that plan prove oo ambtious.T wl therefore Keep 0 fan informal kech, coy nding some of the pobles that must be des with 32 Pendameoal Relations Kar Theory “The computations! sytem tes representations of a given form and motes them, According), UG must prosde meas to present an aay ‘ota fom the locon in form sessile the computations = tem We may tke tht form to be some version of X-bar theory. The ‘conceps of X-bar theory sre therefore fundamental narnia the- fry the erin proper ad vltons il be tated in he sme and derentary terms of Xba they 'An Xbor rare composed of projections of heads seta fom te vinnie lations, then, il iveve the eal x one term. Fur ‘hermeve, the bei relations are typically "Vocal" In ster of the {ert (1), 0 lon rations ee presen: the Speer) elation of {Ze to, and he nad cwplemen etio of Xo VP (order inva ‘he ual conventions a9) oO oe x OY “The tea complement elation snot oly "mor oe” but sso more Ssdemenal—peiy, associated with thetic ©) relations. The Specthead reaion, Fil sug below, als into an “eewhere™ exe ‘ory Pting aside adjunction fer the moment, the naronest plausible Fpotiess shat Xba structures re resid tothe frm in 0. om eat elations are considered (hence no relation Detween X and a phrase ‘Miia Program, al ped within ¥P oc ZP and Red coment the oe al thn Another adie lal ain It red ed for eae re "fon af aver oh head fit Now Prase commen ken). “hooters ein nk toch we wil turn, The verso of ‘sist pera exe re eu tht arp elie ae cs, dpe with eh von govern yaa ad go rh the enn ye cl a i ot | marr: eos al te ms rer, ts prope ‘be pursued. pees “ade Cushy. stanarly sued thatthe Spee ed ean __ eestor Ce fr he ut postion, te set poss | Son angel Cv nde goverment yV. aig contrac |i the objet Casemaced by vb ts epee ey feel Gre ark The namonsr aprons we ar conser ues tht a hoe odes of strata Case game be ea thie tortheoree term remay unre Specs Felton, {As asaved in caper 2, a Gaberation of Paes (1969 theory of indon proves a natu eechanom, wire we tke the ese sc turf e dau obo) o Che a se ey an ~ fF Age sme Ag’ Ne YP ited haw ao eb spitrof TP UST) ad a pwc cy enol Mn ot ey nrc vs ‘copes agony tht nls an afemation mae ab otis 8 wel (Polok {Si Lae 1990) aed Ag vor ners tis Ahetwofantina os of Agr Apr caecon of ents (gas, rurier.peson tee oni o they tet nd bit meme otgh Ay tn Ap may of cure be ile sons, Sra wo verso NP) myer ‘We now pd toh apse nd ral Cate manesatns ofthe Spetead aaton (NP Age. Bue Coxe propeizs depend on horas ef andthe V hen VP. We thro atmo T ‘Mics to Ape foming (a aod V rss o Agi forming (Icon vm teers o Age the se ee pened) ea 0 eT Aa be vaet “The te mmurption tha he sty ween he ie andthe aise in both postin he lain of NP ‘ev'p mained ty Agr calcin of ees i bth Peis Stosmem dtermied ty the etre of te Ar bead ofthe ‘Sop and Cae an clan tha jin to Age (F or). ADP {he Spc ron to he Apr eompls bese msec Casa cre eure The Speen an en head ret etree decor cotguraons for infecional mero}. Tecra Case nrhng by sta sng 8 ang of NP 6 she Sper te ap Coating VI ou 0 (pee Aol tral fair eing wo [SP Ags I the Veal met Tppotes icone (rt hncfonh sue). he qusion ares why {he ee (Greco i he complement) at to Spec, Al a te mer To [Spex ABal sling unexpected croing uber than the “Eu nee tie We wi rn to his pevomenon too. Fielg tha los en passenger pean Srvc epee bes aly “dee oer of tanpuge.H paramees fre morpblopcayresisen he mansetsecbed ext. et Stout ten ngage arian in hse. “te sane hypothesis ees natal 4 peta dines. wth the undehing tr show nr, a2 ton fr #2 tson ef feu in theca ated than ete ‘meligee Ring of NP to Spec and A 10 Ast, creates the structive for NP- | jective aprerment intra tthe pedicle phrase. Te reling st turer 2 plausible candidate for the sal cane complemen of omer, ‘end 40 on. fn te former contruction (cmplenet of tender). NP | es further to [Spec Ame) at LF to recive acciative Cas: in the | later (complement of), NP rises evry to eve nominative Case and verb aprecment. yelling the ove orm Jolt inelige ith Jobe ‘tering into three relations: (I) a Case elation wih T Age] (ence it ‘ally the verbl complex [7 Ase] VD an agrerent relation sith ‘Ags (ence the verbal compen, ad () an agreerent elation with Ast fstuetre (4 (hen the ejeivalcompies) In both constrains. he INP abject is ouside of» full AP in the small classe constuction, Feud, and ue sitar i of ype that appears epee ‘An NP. then, sy enter ino two kinds of structural relations with a reicate (ver, ajeive}:sgreemn,invoing Features shared by NP ‘nd predicate er Case, manested onthe NP alone. Subject of er ot | Sects and object of verb ene no these relations fut not objet of lajeive if that an instanceof inherent et structural, Cate. Both __ltons involve Ags: Agr aloe, fr serecment eatios he eetent T FV alone (ising to Ag, for Case tions. “Te structure of CP tn 2 lng ford by other properties of UG, _ sbming the nimakst approach with Agr abstracted we comme roperty of adjectival agreement and the eubjer-obes inftionl 9 fe,» reasnableassumpsion, given that apeerent appears without __ Goes in NP“AP agreement) and Case appears without prec! (8 ‘ronsve expletives, wth he expletive presumably nthe (Spe. AB) f : 6 capers poston and the subject in [pes recning Case; se ne 1. Any [Povoprat version of te Cave Fier wil require two coctences of ‘Agr if two NPs in VP require structural Cae; conitions om Mowe reauie the arrangement given in (2) if structural Case i conse as Culined. Suppose tht VP contains only one NP. Then ave ofthe #0 [Agr elements wl be “acne” (he ether bang inert or pets isin). ‘ich one? Two options are possible Ary or Agi I the cove Is [Rav then the spe NP il have the props of the subject of an sitive clause if che chic & Aa ten wil ve the prope of the objet of # tranatve cle (nomirtieacesative ad ergtive- ahucutve languages, eapetvel), These are the only twe possibile. Tontues spat. The dintion Between the two lnguege types reduces toa trivial queaton ofmorpholony, 8 we expel "Note that fom this pnt of ie, the terms nominate, abst, 50 on, have no subtantve meaning apart frm what i eterined by the chece of “ctv” versus “inert Ags there seo real question ss ‘how thee terms comepond toro nee pe. “The “acne” cement (Ag im tomicatineorcesave Ingusyes and ‘ao in erptiveotolutie languages) piclly mags = lsat’ ‘Case toss Spe, whichis also higher on the earatabity earch, mong other proper. 1 natural to expect less marked Case fo be ‘compensated apn, asa tendens) by morematked agrconent (het ver apcement wih nowinatie and sbsltive than wth accusve tnd egative) Tae C Command Ceiton or anaphora lead us 10 ect nominative and eatin tnd in tramive cesrucons* ‘Simigr comsierations apply to Hcensing of pro. Assuming Riss theory (1982, 19863, pro iets ia Spec hod eation to "stone ‘Aes or when ovetied by certain verbs V, To reas these proposs ina untary Xeheorete fora pro i lense only inthe Spec head Felton Uo [ug "Apt, where 2 [tens oF V, Agr strong or V~ V" [Ucersing of pro thos fls under Case theory ia broad sense. Smior ‘onsierstine extend rather tural to PRO.* ‘Suppose hat ther propre of had poverament aio haven natal expression in tems ofthe more fundamental retions of Xar they Suppose furtcr that aneedent goverment a properly of chains, prenibie ines of ecormand ana tasers. Then the concep of 2" fmnment would be diaper, with pincplesoflngunge rested to Something else to coneepual recess local X-artheretie relations| to the head ofa projection andthe chain En rion. Msi Popa m Let us look more osly at che local X-har theoretic notions, aking Anse to be fundamental. Asume binry bunching only, thes ucts lite to (1). Turning 10 adjunction, on te astro ef Chomsky 1s8ea, thee is no sdjoneon vo complement anton (at lest. in ‘overt yest) bas & Kind of “atractre preserving” charter, and a —sepnent-estgory dsnetion holds This the aistures to be can sidered ae of the form shown in (3), where XP, ZP. and X each fave a higher and lower searen, ident by sbscriting (and ads). wo, ‘Let ut now consider the tions that enter isto ¢minimait program. ‘Tee basic sements of » representation ate cin. We conser fst the __ cee of one membered chars, consruieg notions stacy with an eye to the genera ene, The structure (5) can only have arisen by raising of 1110 adjoin 0 X (we put ase questions abou the posse rine of UP, WP) Thotfore. heads cin CH = (H,..n), and only hie hain, not Hin sltion, eres ito hed reais, The cxteporice| hat we esas ae deine for Hs wall as X, ba we they entra Frade ettions fr X, they donot do so for H ony forthe shin CH), snimponant mater, ‘Asse al notions to he iefexie uns otherwise india. Asseme the standard notion of drsnation fr the pie (.B)@ a pment. We ‘ya the category o dminces Bi every sepment of @ dominates B. “Te caegry a conaine Bef some sont of «donates. Ths the ‘Wosegment category XP dominses ZP, WP, X and whatever they Seminate: XP contains UP ard whatever UP and XP dominate: 2P ‘aiains WP but does tot dominate The twosspnent category X ‘ena H but doesnot dominate i a chapter For a bed tke Moxa) tobe the lest fullategory mania pro esto dominating 2. Ths. (5) Max(H) = Man(X) ~ [XP XP the two segment entegory XP, “Take the dman of «head 40 be the set f odes contained in Monte that are distinct froze and donot connin 2, Thus the domain of| Xin 3) ¢ (UP, 2P, WP. YP, H) and whatever thee eateries domi ate the domain of these, sinus H, “Asse, the fname X errors relation is bet-complanen, yp with an ascnated O-eation determined by properties of the Ted. Datine the component dain of he subst oF the domais Teflexively dominated by the complement ofthe consteoction: YP in (S. “The complement domes of (and H) is therefore YP ard whatever i Stine weer type ay arn) Sopp ntend of pug pt ad aed toe ol cnet ging te ope ener Xenon rant soma el ors acing dona cst wena tse Tee 0m ems sear een fon fr po lar ring cay fore Maina Program 1 hin CH= (pu) 18 i terms of these minimal sets tat the local endrrelaons are defied, the bend now beng the noneval choin cH. 1h then, the relevant dmsng are a intended afer V-rasing 10 Yj. Now that VPs notin the intemal domain of CH (= (put) be- use dominates (0, of (0) ‘The sume notions extend to an analysis of lexical strate along the lines proposed ty Hale and Keys (19933). In this case ah analogs of (@) woul be the underlying stature for John sted the hook. wth Vy being 2 “igh verb” and ZP an absactvexion of or the shelf (CP sty. Here shy rises to P the alga rest Va. ae the ements frm raises to V, in the manne ofp ® So for we have made no we of he notion “minimal down.” Bu this too has a natural nterpeaion wen we tur to Empty Category Pin | ple (ECP) phenomena. Iwill have to put ae a careful development | fee, bis ititively clear how certain tani aspects wil ene, Take the phenome of soperionty asin (9) and of reltived imal in the sense of Rizz (1980) (as in), (©) 2 i _wtom, dd Joka prsiae ost whom] fi. whom, aid Joba persnde whom, vis 'b. Superaising, the Head Movement Conta (HMC) (Spec, CP1 ‘lands Gud wslanis) Looking at these phenomena in tens of economy considertions i it Sle tht in all the bod cases, some element filed 1 minke “the Sortest move." In (ai) moverent ef har, to (Spee, CP i long in | petra sense (Geinable in terms of eomsmand) than movemeet of "ham to this pation. Til the cases ef (9) the moved semen his “kippes”a positon i cra have reached bya sorter move hd that prsition not been fil. Spelling out thse ations to acount forthe ‘ange of relevant cases ¢ not x iva mate. But dots sem possible in way tha aceords reasonably wel Sth the Minnis Program. Let ‘simply assure, for present parpotes, tht this tsk ean be earned out, ___ and tha phenomens of the kindest are acum orm this ea) in tems of corey considerations." ‘There pears o bea conic between a natural notions of es shortest move versus fewest steps in denvation a derivation __ Seep to shores mons wl have more eps ii edu te nue ry hoger ‘ot eps t wi have longer moves The paradox is esoled if ake the basi transformational pectin io be not More 2 but Frm Cha an ‘opertion that aris, S310 the stature (0s) 10 forms (108) in & ‘Sl step tdi the chain CH ef (100), (10) 4 esses fet be fel oka 1 wi John seems ob isy to Wi © CH= (od 0) Similarly, in our cas of sucesinecyebe movement. There i hen, to confit Hewexn reducing derivations to he shortest nube o ps ‘nd Leping inks nial ("Shortest Moverent™ Conon). Thee are independent reasons to sippos that thi the covet approach: ote, Forename, that sussssve flr sujet raising o (Spe, Agr] and abet ring to (Spe, Aso ‘tossing the race ofthe VPastemal subject. The picile (15) entails that this ia permis derivation, a in (12) with V-aiang to Agr. I emsins wo show thatthe died dervation not only peri bu ‘bkgntony: i isthe omy posuble derivation. The x srsghforwar, ‘Soppose thar in (12) the! VPantensl subiet in [Spee VP] ites to [Spec. Atl cher overly or covet, yielding (6), he rte ofthe sed subee So, / i ‘ He tapers 08 Age si Age oN as OW vy Suppo ure that Vis to Ag ier overt or overly om ing te chain (1) mh he adorn (13, yO). Now Saba ese equa fom Ob) 0st Oj ca a othe [Spr Ave pois. But the postion f ecpiedby Sj, Becing tho open. "Theeee, 1 cave Cae, OF) al move Gea 10 Sone Tages pen eosng (pee, Apo ert Spe T] of 10 {Spe Apa. Hut tha impos, even ar theme (¥.AP] fas to teh lions poston. Rabing of (. AB wt orm 8 tow ein with te the Ag poston of (16 ade few minal ‘ovain ML Put fu ta mer of M. Acodsgy. Oj cant {fo ny to etch poston nM apa rom the potion Spe. Al Siendy fd by the ube). Heroin of he VPSmerol st ite ISp, Atel oitn ck any hind of Cis signet tthe hermit tro np == Tealove that cong and nt msting the ony pei opin in any gute The radon of Ce hen oreo ‘aration at ener oa number fob ces 33 Deyond the Interac Leeks Strate Recall the (viru) concepts nerasitis within this genera approech UG determines passe syeboic representations and derivations A at- fuge consists oF @fenon and computtiona system. The compte tonal system draws from tbe lscon to form derivations. presenting ‘em fe thon he ena of Xt at eon (etcmines «linguini expression, an SD. which contains a pei (s Ineting the interface conditons. Hay, tbat would be the en ofthe Tory: exch fingste expres is am optimal Yeaheaion of imertace Miia Pogo ca conditions expresed in ekmentay serms (Chain tink, local Xbur theoretic elton. pair (2) sting these conditions and geerted in the mos economical way. Any ational susie or assumptions equie empire jstieaton ‘The EST framework adds aden) stun; for concreteness, tke -Lecrures om Gonerment ant Binding (LGB: Chem 198K). One era el axsumption bs o de with the way in which the computational sy tem presents Heal items for Further computation. The assimption that hiss done by an operation. call it San, which sets an eee) of ‘tems fren the lexicon and presents it in format sting the cond tion of X-ar theory. Saiyan “alate” operation a ters the fonction st LP are draw fom the fscon before compustion prose spd are pesented in the Xr Frat. ‘We this postlate aration level, D-Structur, yond the 60 eternal eface levels PF and LF. D-Stutte te ermal itestce Iween th lesson and the computational system, formed by Sat. {Certain piles of UG ae then ted to apply to D-Strtute, speci fly tbe Projection Pisiple and the O-Crterion. The computation procedure maps D-Siuctue to stother love, Structure and tht tranches” to PF and LF. independently. UG penipe of he various ‘modules of grasmar (binding theory, Case theory the pro medi te) rly atthe level ofS Srutre (peapselseere sx well, in ee a). “The empirical osifieation for this approach, with is departures om oncptual neresiy. substantial. Neverthe, we may ask wheter {he evidene wil bear the weight. or WREther 18 possibe Yo move 1. ard a mint program. "Nois tht the operation Satisfy and the assomptions tht wnt ‘a not unproblematic. We have Scribe Seis an operation tat Selects an aroy. nota se; diferent arengemen of xe! items wll Yd diferent expressions. Etacly what aray ix woul! ave to be ‘tui, Furthermore, this picture quires endiion to ensore that Suture tas basi propers of F-At LF the conditions are vil IF they are not met, the expression redves some deviant nterpreation at the inerface there is nesing more wo say. The Prjection Principe and the @Crinrion have no independent sgniiance at LF * But at D- Sucre the to principles are needed to make the pictre coherent i (he pitre i abandoned, they wl lore thr pimary role Tew pen les ae therefore dubious on conceptual grounds though treme to oun for thee empirical consequent athe constr apret ve contr substitution into 8 O-postion. fhe empical consequences ean be ex ‘laid in some eter way aed DStrctre eliminated, then the Pre ‘ion Principe and dhe 6-Crierion can be dispensed wth ‘Watt mors, posulaion of D-Strucae raises empiial problems, ss nged st onse when EST was reformulated a the more resteive PRP frmewoek. One probit, disused in LGB. is posed by complex activa onstruction sus (170) ith the S Strvtoe eepreenation| (195 (the wace ofthe empty opetor Op) (07) 4 Job is easy to please John sey Le OP fe PRO t plese “The evidence forthe S Structure rprescration (17) i cergeing, bat Jake cece oor 8-pston and hence cannot appenr wD Sine Sot therefore violated ty 48 proposed that Satly be weak ene: noo postions a eal itm, sch on, cae be isnt i ‘he courte ofthe derivation and sph okey at LF (and ee vant, $ Strostre). That content wil te printipes, though ot ‘vith er spit, one eight argue ‘We need not ry om tht mater, however, bucuse the technical de vice dots nt fel. AS nla by Howard Lasik the ZGB sltin as csv an NP of arbizary complety may occur in place of Jobe (Ge, an NP incorporating sre stch (1) inter). Within snyhing ike the LGB fraework, thea, we are driven tox version of eerazed transformations, a8 inthe very eles workin generative fpramnar. The problem war ecognied at once, bs lef as ar ones rrndox: More rot werk has bought forth athe ass of expressions Interpreble nt LF bet notin tie DStuctue postions (Rehan 1991) along wth ether reasons to suspect hat Were ae generalized tcsnsformations, ce devies like them (Kroch and Joshi 1985, Kroch 198, Leben 98, Epstein 1991-50, tbe special asumptions under ying the postulation of Steet lose rei Since hese assy tins lacked independent conceptual sopprt, we are ed opens ith the evel of DSiactut nd the “allatence™ property of Sai, relying its place ona theory of ened ansforeations or lexis! a> thot the erp consequeness ofthe D-Suructre conditions re rain tobe faced 'A theory of he prefered soe eailycorsrocted snd tas os 12 hve many desirable pepe. Let us replace the EST assumptions of {LGB ed rated work by an spprsch along the Fong Kes. The is simi, Mia Program 1 | computational sytem sets an tem X fom the lenicon an projets it toan X-bar strat of one ofthe fornsin (18), where X =X" ~ Ly XI aya x b tex! © bebe XI) “Tis willbe the sol reside ofthe Projection Prinsiple “We now adopt (oreo les the assumptions of LSLT, wih a singe eralized transformation GT tha takes phase marker Kandi Se it ino designated ely postion 6 in «pase marker K, forming ‘he new phrase maser K*, which ses X-borthsry. Computation proceats in parallel, sleing fromthe lexicon ely st eny point. At ach pont in the devivation, then, we have strate 3. which we may | bn: of 36a et of phrase marker. At any point, we ry apy the | operation Spel-Out, whic sitces to dhe PF component. Ws nota ingle phrase marr, the derivation erates a PF. nce PF us cant __ ply toa set of phrase racers ad no lpiimate PF reptsenation m5 secre. ea sogle phrase marker, be DF eles apply ti yi __ nam, wlich either is leptimate (so the devationconvergs at PF) oF ‘a (he deivationapuin crates at PP), ‘ier Spe-Out, the computational procs cosines. with the sole ‘ovsirain that it has no futher acest to te Hic (we mest ene er example, that J lft doesnot mean they woodered whether Soh |e before finishing his work), The PF and LF outpats st satis he (extemal) incertace condoms. DSusture disappears robes trie, GT iss subsituon operation I argetsK and sbstiuesK? for 9 in ut 0 isnot drown fom the lencos teefre, mist have een ong withthe Sten by GT ill, GT, then, args Ka @ and subsites Kor forming K*, which must sale X-bar theory. Note that this a de ‘pon of the inner workings of single operation, GT. Ion 8 pa deih some particular algorit for Move mor forth operation of mo- ‘es ponens ina proot Thus, jv to the ee that seas ly the desvaion its etcing only te rocecve nee. We never se 8. the “ist hall” of the raising of an NP to set aston, ‘Alongs the binary seston operation GT, which maps (K, K) | 10K*. we also have the singly substitution operstion Move which ‘maps K to K*. Suppase tu this operation works jst as GT do i 60 coaster targets Kad and satiates for 6, whee in di case pase arr within tbe tageed phrase marker K asl. We assis fetes that the operation eves Belinda ace fd Forms he cai (4.2) [Agnin, i le when we rom the eseton; it part the ince (rocking an operation caring the deeation oar one step Suppose we rest sufstitutionepertons til further requiing that 0 be enema! to the targeed phrase marker K. Thus, GT end Move 2 textend K to K*, which includes Kas & proper part For example, we an target K= Vad @ 1 form [p 8 Vi. ad theo ether rae rom ‘tin ¥ to replace @ or inser anotter phrase marker K for 04m eiber ‘se the rest st satsy Xar theory, which means that the erent ‘replacing @ must be a maximal peozction VP, the specifier ofthe new ‘prose marker VP = “The roqiremen thst substitution operations almays exten! thei ta fetes number of eoaguincs. Fist, ils ersio ofthe ste ‘leone that siete By the most detentrycmptcl conser ons: without it we would lose the efits of those cass of the ECP ‘hat fl under Relatvized Minimal (ee (6). Ths, soppese a in the coure of derivation we ave resthed the stage (9) (49) a: fy seems is cei ot be ee [Clq tbe cal) © [CC ohn wondered fe Ce Mary fed whet Row} Violtng no “Shortest Movement” Condon, we ean rae Job directly to the matrix Spec in (19) in single se, ter inserting i Tom the lexicon o form Jo sees ete e here (sperasing: we can raise fx wo adi vo Cin (198, nee isting can ftom the lexicon to Form fi br com the cr (cating the HNIC) a we ca rie ho to the mates (Spec. CP] postion n (19) ler wing wher to thee tedde Spec, CP] position to form Ae Jn wonder what Mers fie (tng the We la Costa)?” “The Cenension” version of the srt eee is therefore not only stright but jeied empirically without subtle empiisl Sreurent ‘Ascend consequence of the extension condition shat given @ ue ture ofthe form [XP], we ean insert ZP nto XGieting, © [eX YPZP), where ZP x drawn fom wba VP (ising) oF inserted from outsde by GT. Sinilaty pen [xX] = camnot mer Zt orm lg X 2P, There canbe no raise to complement pstion. We thee Misra oe wr fore derive one major consuence of the Projection Prine and 6 (Criterion at D-Stncture, ths lending support to the belt that these ions ringed supertoous More gerry snoed by Akira Wat ‘abs the binarty of GT comes close 10 entaling that X-ber siattores are retited 10 Binary branching (Kayee's “umnbiguovs pets), tough abit more work ego, ‘The operations jst dscns re substitution tenwformatons, bul we mus consider adjunction 2s wall. We thie continu tallow the X-or Structure (5) as we as (1) specie (20) (G0) 8 YX) Be ¥P XP) 1a (ato) eroeet category i ain to the zero eve eaepory X. and in (2h) 9 maximal peoecton YP is adjoine to he maximal poe. tion XP. GT and Mee must form stroctures stisying Xba thee. now inchung (20). Neve thatthe very strong ential mtivacon for ‘he act ce jst pen doesnot apply in these eases. Lat us asm, then, that adjunction need not extend is tage. For eoneetnes lel {mune thar tbe extension rguirement hols only Zor sbttion| in ovr syn, the only ease rege by the rl argumect for the oc 34 Beyond the lcrfae Level: S Strate ‘Seppose tha DSirucure i eliminable along shee nes. What about ‘SStrocture, athe eve hat has oly thor interna msi? The ase isue is wheter dere are $ Swuctire conditions He ot we can pense wit the concep of SSteture allowing Spel-Out 10 Sony eel inthe manner indicated eater. Plainly this would be the optial concen, Thece ae two kinds oF evidence for $-Ststureconitions. 21). Lengunges der with respect o whee Spel-Ou apis inthe ‘oure ofthe derivation 1 LE. (Are wh praze mowed rin sis the lage French sige with overt Vrasing oF Englshstyle wth EF Versiing) by Injust about every mode of prema there is extensive "idence thatthe condtons apy at SSucure. we chapters “To show tha S-Siructure is reverts sopefvous, we must show tha theendenee of both hinds though substan not compeling To the ene of evidence of type 22), we must show that dhe postion of SpelLutin he derivation ie determined by aher PF or LF popes, these being the only kvl, om minialist essumptions. Fonhermt, paranetriedillenes must be reduced to morphological properties the Minimal Progr famed nthe ers so far assumed. There a SHrong reasons to supe tha LF condions ae not relevant. We expt Tamguages to be very smite a the LF level, iflerig only ata reflex of roperies detectable at PF she reasons basally rede 1 considera ‘os of lara Ths, we expec that eth LF eve these wl eno relevant diffrence between languages with phrases overtly eed on ‘Siu (€x. wepraer or vets). Hence. we ar ed 0 sek morphologic! propre tht are reflected at PF. Let us Rep the oncson in min, tuning iter. ‘Wit regard to evidence of type (218), an argument against S Stee twee condos could be of varying strength, asshown in (2) (22) a, The condition in question cur aply st LF aloe. ', Forherore, the condition sometsnes must apy aL ©. Funherore, the condition mast nr apply at Sirsa. yen (22), the weakest of the Ue, slices: LF asindependent mo sation, but §Strtore does at. Argument (24 x tomger on thea ‘pion tht, optimal, conditions ae unary: they aply ata sing ley Hence st LF i possite Argument (22) woul be ds. "To sample the problems shat aise, conser binding theory. There are ua ergurents showing that the binding theory conditions mist apy at S-Struture, not LE Thus, consier 2). (23) a. yow it eke the pores hat Jn wok) by howe many pers that John tok} i you sy be ied 12 who rsd eked [how many pers that Jb 10k). Io (238) he eoommands Jn and cannot take Jobn as anecdent i {Q30) there sno ecommand relation ad Job canbe the amscdent of Te In (2) John gain cannot be the amecedent of he Since te binds ‘properties of 2%) ate thaw of (238), not (3b) we conclude that he {eommands Jon atthe level of sepreseottion at which Condition C ppl. But LF movement sits to who in 23), Consivon © st apply at Since. Hinges, ith wh init Miniatit Progra wa “The argument isnot concuve, however. Following the line of amt | ment in sion 1.3.3 (ee (15), we might reject the last sumption thal LF movement mijoins eof 232) wh, Forming 24, the ace of the LEmoved phrase {24 {how many pctres that John ook who] [sid he ke) | __ We might asume hat the only permissible option is extraction of how rat ro te Tal NP, yedng a9 LF form song the lines of 23). the trace of fe many. @5) how many] who} {sid be ka [¢pctres thae Job took] ‘The answer, then, could be the pai Bl, 7), sang tha Rl side {is 7 pictures that John took, But in (25) he comands Jo 0 that ‘Condon C applies asin 23s) We ave therefore nt corseled to as. ‘sume that Condition Capps a $Structate; we a ke othe peers fable option hat conditions invoking interpretation apply ony et the Interface lcs This an argument of the ype (228), eck but sufi Gent We wil turn to the passlity of stronger arguments the types 22) and G20) “The overt analog of (25) equites “td -piping” ofthe entre NP ow many pictures thot John tok, bt itis pot ear thatthe same 5 tne in the LP component We might, n ft, pred futher. The _ LF rule hat ascites the test k-phrate with the sephase in [Spe CP ned not be const as an instanoe of Move a. We might = think of 35 the symatc basis for absorption in the wenee of Higa ‘otham and May (1981, an operation hat associates two phrases vm generalized quant Iso, ten the LP role need sisty none | ofthe conditions on movement “There has long been evidence that conditions on movement do not fold for multiple questions. Nevertheless, the approach jst proposed Sppeart to be blocked by the properties of Chneso- and Japancse ype ronghout but observing at as some ofthe | ‘onditions on moverent (Huang 1982). Watnabe (181) bs engud, owes, hat oven in these Innings three overt movement in ths exe movement of an emply operate, siding the eflets of the ‘ovement constraints. Watanabe corrc, we could seme that & trhepeator always ries over. that Move subject tothe seme me chapter vostions in Japanese is fie of these conditions. What seman i he {ostion why overt movement of the operate is always requie, a ‘question ofthe eaepory (21a. We wl turn ota. ete real sin she miimalinassrapiions tha Lam conetuing ‘on be ual all conditions ae Sela conditions. and a input ‘ress she optimal ezaon of uch str: condtons. Let us ‘cider these sation mere cosy ‘Consider a representation at PF-PF is aepresentation in uves ponte, wih no indation of syractie cents of relations among them €or strecur,binding.sonernment, et) Tobe inert by the performance syle A-, % must be consttated ently of lei more PF objets, tot, Senents that have © nif language fPeolent inerpeton ste nee, In tht ase we wily thal Stes the omit of Fl Iteyprerron (FI Cals Ft doesnot [povde npropit intrusion {othe performance ssc, We take FL fo be the convergence condition: fy sneer F, the dawation D that Termed it converges at PF: others st eases at PF. For example it iain reat consonant ora [+g + ow] vowel, then D eases: Siler i wcontins sme morphological clemeat that “survives” to PF. lckng any interpretation atthe interface If D converges at PF is futpat ceive an aicolatory-percepual interpretation, perhaps 2+ bbe ‘All ofthis ruightorvard-indeed, hardy more than an expresion cof what stacy aramid, We expt excl the same tote te at LF. “Tomake teas cones, we rst spel ou exp what ae the Ip mate objects a PP and LF AU PF, Us she standard probe of un versal phones, ACL, we ass each eine objet 1 be ach (CH= Ga, --y ag atleast perhaps a ms) with CH hed, 0 ncaa fn eerste conn Woy ya the represclation A safes FL at LP it consists emily of fepitmate objects devivtion forming b converges at LF if. stisis FL, and ‘therwise crashes, A convergent derivation ay predvoe ute gibessh ‘actly a P. Linguistic expressions may be “event” aon ‘Tinconmersuabe dimensions, and we have pe notion of “wel-ormed Sentence ae noe 7), Expressions have the inerpetations signed 10 them by the performance syste i which the language i embed red “To develop thee ideas propel. we must preced 0 chartxrize 20 sion th te bse properties of A and Aposion, These nocons were eli defined in the £68 framework, bu i errs of assumptions hat are ho longer bed. in particu. te assumption that markings restricted {0 tered, with mulile-banching constructions With thew assump Sons abandoned the notions ae use only in an iouative sense. To te | pc them. et us consider more closely the morpbelogcal properties of Tene tems, which pay 2 major role nthe minimalist propa we wre fetching (ee section 13.2) ‘Gotsder the verbal system of (2). The min er typically “picks un? the features of T and Agr (infact, both Apr, and. Ap in the genera ‘taining to an infectional element 40 form [VTL There ae 60 ‘nny 0 interpret the process, fr alescal element x. One 1 cake 0 ea ars uiiected form, PF sues ae then dese to interpret the bse comples (a as «sep alte phoaoopial word. The ether | approach ist take to have infctonal fstre i the lesan 25 at Inti property (inthe spnt of lesa ponooay: these fates se then checked again he ifecional ean Uo the orp 2 He | the features of and I match, I esappenr a cles the PF compe ‘nt under Spel Out f they conic, remains and the derivation sashes at PF. The PF eles. then, are simple rewriting ules ofthe ust ype ot more elaborate us applying to complex Thane been tacitly assuming the cond option. Lets now make that _ chece explicit. Not that ented no longs adopt the Emende Pollock ssimpoion tha in Englsheype languages [lowers 0 V. V wll have | te initonal fests belo Spel-Ont im any even, and the checking ‘rosedure may tae pace anywhere, in puruela, afer LF movement | Frenc-ype and Enaish-ype largungs now look abe LP, wheat owing of ¥ in the later Would ave produced acjncion sutures ite elke those ofthe asinglangueges. ‘There are various ways to make a checking theory pect, and to aplure generalizations thet old acres morphology and yaa. Suppose, for example tha Bake’s Misr Principle is src aecure. ‘Thos we may take kcal cleent—say, the verb Vt bet 3 quence V= (6 If... If). Whee asthe morphol! complex [Retnd...tnf], Ra rot aod Ia en itessonl estine2? The PF ules only "see" 0. When V is odin to a Functional category F {RY Amo. the etre In seve fo V if tats Fad 50 ‘any Tf remains at UF, the derivation erases st LEThe PF form | always satisfies the Miror Princip in derivation that corerges at UF, Other technologies can readily be devine. tn thi ae, howene ii a Chapter ot sear that such mechanisms ate in order; the mst peramsive ev ‘dence forthe Mirror Pritipe es ouside the domain of infitional ‘morphology, which may be sbjet co diferent prnsipes. Suppose, sy, {hat richer morpholoey tends tobe tore “vise” thai closer tothe word boundary: ir, snd if the speaations ofthe paragraph ending With note 13 are om the neh Wack, we Bou expect nominative or “sti ageement (depending 0 language pe) to be mow periph ‘ral in the verbal morphology. “The funcional elements T ard Agr therefore incorporate features of the vei Let ut call thew fests farure: the funcion of the V- Fears of an sflection element (1 chek the morpbologeal prop ‘ei of the verb selected fom he Txion. More generally ts all Sich fences of = teal tum 1 L-fenaes. Keeping to the X-bur {hectic notions, my tht a postion i orld i ina foceh ‘elation to an Leste, that, nthe intel domain or checking do- Imain of « frad with an Letue, Ferermore, the checking domain an be suid into two categorie: nonadoined (Spe) and aie. et ue cll howe postions narouly and broad Lee, respective ‘Avaructaral postion thats marrowly L-rlatd has the basic properties Of Apostions: one tha not Lerelated has the Basie properties of A postions in parti, [Spex], not etelated 3 doesnot contin a ‘Wate Toe saat of Proaly Lore (adie) positions has bees dette, pariuly i the theory of sratling* For our lites pur poses, wera lave he matter ope. ‘Note that we evblly asin ai lau. that Veraising to C acivaly Fraising, wth V incorporated within 1 and is wowvated by properties ofthe (C.1) system, not morphological checking of V. Css ther properties tit tings i From the V-aute.as disse in seston 14 “The se considerations extend to nouns (ssrsing the D fad of DP to have Nefetres) and adetivs.Patng ths aside, we can continue ‘speak informally of A- and A-postions, understood in terms of ltednest as 9 fie approximation ony. with forex refinement sil recta, We cin proce, then, to define the kite LF objets CH= Geet) i something he the fair way: beads, with 2 fn arguments, with 3 8m A-postion: aunts, with a i a Feposivon. and operat¥-vtabe constructions. which we wil ie fecurn2” Ths approach seems rately anfeobemati. Let us ase oad proceed Mit Mogan w ___ The morpholopeal features of T and Agr have two fonctions they heck properties of the verb that raises 10 them, and they check prop ees ofthe NF (DP) that raises to their Spec: hi, they ensure that DP nd V ae propery paired. Genertzng the checking theory, lt us 3 Some tha. ike vests us are drawn from the con ith al of thei ‘moephologia ets, nding Cae ard features snd that hee too Imus be checkel inthe sppropriste postion® in tht cse (Spe. Ax = (which ty include T or Vi. This checking to ean take place an) age ofa derivation to LE [A standard argument for §-Strctur conditions in the Case me ie | at Case features appear at PF bat ust be "ible at LF nce, Cie | ust be resent bythe ine the derivation reaches S Snare Bu that | argument coTlapes under a checking theory. We may press, then, | th the asumption thatthe Case Fieri en interface eoncition—in fas, the condition that all morpoogial festres st be checked some ‘whee for convergence. Thee re many interesting snd abe prbles evbe addressed reluctantly, 1 wll put then aide here, merely snerting [Next consider subjectverb ascemert, asin Jos hus Bil. The @ "atures eppcaris three postions inthe cour ofthe ervation internal ‘Aas end th NP Joe Spee, ABR ch ceking is worphoogel __ ees 1 the esl es were pope ches he deen con- | Sse. Bu at PF and UF the tire opps al Toe, 40 hee nes i he NP and ver hat agree. Age lyse @medating role: |e it ns perored is uncon it dsappean Stace hs oon ak, Vgetated and NP led, Age mst in fat have fo ing furs: Wits that eek V eine 0 Ast, and NPs thst heck Nin ec, Aas The same sue Teh ces he ese of the veh and the Case of he sje. The Vata ofan fetal Gloeat Grappa wen hey check V. the NP estes wes hey heck _NP (or N. or DP, se note 35) Al thi automa and wan the _ Minima Popa Le snow rer othe fstype of S Sucre conan (28. the | otton of Spel.Ovt ater Vrain in Fresctiype lunging, Before Wisin in Enginvige tngnges (have now der with low ‘ing As we hve sen the Mist Program pms ly ne sh font he protien BP endions reine mero popertes —7~ 98 utes must force Versing in French but not in Engioh, What can see conditions be TRecl the undeyng inition of Polk's apposch, whith we ae scaly sswming: Tretch-ype languages tase stoma” Ags. wich ona over rang. and Engishigpe fangages fave “week” A. (Rte leks Let us adopt tat ies. rephrasing it in our er: he Viicsuncs of Aer ae suong in Freeh, weak in Engh, Real that hor the Vfcoutes have dove their work, checking ane. they an I does ot ie to Apr overdy. he Vfeatures sevive 10 ‘pre us aon make the naturel esumpton that "aong” etues ere tlc Fr ad "weak features insible at PF Ths fetes ate not TRetinete objets a PF they afe sot proper components of pon sae Therefor if stron toe ren afc Spel-Ous the d= ‘rerun ™ tn French overt ming i © prerequisite for comer ene in English ts aot “Teo ror questions remain: Why is over raising bare in Engi? ‘why dae English nuaanes have sad e ie overt a do verb in Frenc? “Fae fist quetion is answered By « natural coon contion: LF movement ie seheaper” than overt moveren (ail the principle Pro rena) See ween 12-3) The dttie idea stat LF operons sa kind of cones” rls, operating mechancly beyond any ey obarvable ets, They ae es costly than overt operations. The resto reach Pa fan oe posibl.” ming vert Tr Enatrype languages ove ring nt ford for converts (hertre tarred by economy pring. “To del withthe second question, consider asin the inition tht davies Pollok’ eecourt asin of the auxiaes refx thet 5 vrs waaty the ae pacholdet for ceiinconstrectons 8 me ‘Rly ight" srbn Adopting the stution (but not the accompa Taito, ee assume that ck mens Bick semantically ee warren, enol vile 0 LF rie. 1 they have not asd ove peter be seo rie by LF res a he deviation wi cash Tons conser the eiference Retween SVO (or SOV) langues Hike Engst Gapmese) and WSO langues ke lsh, On our ssumpucns 7 te rlkad overt to 1 AEE) Tesh. while Sand O rag in Oe UF Component to (Spee Ah and Spee, ABoh epee We bee career weap nee fleences i srs of he strength of OE raptor eaure One oss otha the NPfenture of fons aise Popes » | ssh and weak n Wh Hee, HP as a (Ste Age TH | Engh orto Sper Ost ve evan vox comere: The pom “gle Prcrsiate bry such raking i 10h, The Exened Proton Soci, wbich requires that Spe. 1] be alized (peraps by an empty ercpey) reduces to « morphologies property oT: strong. or weak NP- ‘Garsce, Nowe thatthe NPfeatoe of Agri weak in English st were Raton, Fnahsh would exhibit overt objet sit, We ae sil Keeping 10 Theranial esumption that Aas and Apr ae collections of tures ako relevent subec-joat dtinson. ence no diflrence in sent lrfeceres Note alo that language might low both weak and strong Inction, hence weik and strong NPeatures: Arabic is = SuExsive | Gn, with SVO venus VSO conelating withthe richness of vinble verb intecson Aloe tcc tnes, we can eliinate §Structreconitons on sing an towering in favor of morpholpial properties of fexal tems, “Arad with the Miia Program. Nete that a.ceraintpolony of an func is predicted; wheter correctly or not resins to be determine. Tt Watenate's (981) theory of shonoveret is correct, thee is no © gtacre vation with ard to wh nite tpungeerenns 5a, ‘Fresh Japanese) reduce fo morphology inthis eas, the terra mor _ phology ofthe whephrases. Sil the question arses why raising of the _Mvopcator isever overt, conzey to Procastinate, The bese conor btn assump tht operations riven hy ess: ey feast reson” apd if hey mus be ot etervise (Chany 1986, ba cutyecr 2, Our sssunption that operations ae driven by eo | pholowcal necessity: etsin features ust be checked in the checking ‘main of head, othe derivation wl erash. Therefore raising of treat to (Spec CP] must be ven by such a requreent, Theat ‘al asurption that Cmny have an operator feature (which we sip htapt in cin under cosrcsos ( ton Coie ch npn (-) Minion Prog ws |G) a Jom sua that be was loking fret, ad oi Bi |b Join sai that he was looking frac and 0 di Bil ay that he was ooking or et) “Te frst conjunc is everabways ambiguous. Suppose we rele the " ambiguts in one of the possible was, sy, by tang the pronoun to ‘efer to Tom sed inerpecing 2 car nonspexfly, that John sid ‘that Tom's quest would be said by any cat In the eps! case (200), © paral requirement of some Kind (ell it PRD regres that = the second conjunc must be iterprete the seme win this case, ‘ah fe rlerring to Tom and @ vat understood norspexticly (Lako 1970, Lasik 197, Sag 1976, sad 1993). The sei ru the fll sentees (30) noreevint linguistic expeeson with dative low fang intonation for: it too mist be asgned is properties by the theory of grammar. PR surely applies at LF. Since ie must apply to G06). the simplest assurpion wovtd be hat enly (30) reaches Li (Gt) beng derived from (305) by an operation of the PF component iting copies. Tere would be no nee. the, for special mechani fo sesoumt for the praleis propetes of (0a) leteresting questions aie when this path followed, but seems promising Ho, the trace ‘ction cpeaton mey wel bes obligatory variant of « more gen ‘pertion applyingis the PF comport, Assuring is aproech, (28) ptaionalabievistion for (1. 01) [ogi hich hoor] Jot ved gin which howe] | Tre LF component comet the phrase wh to either (32) or (26) by an ‘operation ain to OR. (62) a. [which howe Lain [hich fai [house] We may give thse dhe nstveintrestations of (34-) 09 2 futich x ra house] i 3) © Dwtich a x house} For comergence at LF, we must have an operatrvaiale strate _ Accordingly, inthe operator pasion Sper, CP, everything bt the ep- ‘air phrase must delet; therefore, te phrase wh of (2) dees. Tn he _ Ee positon, the copy of wat remains n the operator positon des, ust the phrase wh (an LF analogue tothe PF rl st deeb), Pe apie In she preset cise (perhaps seers). ese cites need note spe ‘Sid ther option wil east. We ths devive LF forms intrpeted as {@ia) or Gb), depending on which option we Dave selced. The LF forme now const of egimate ober, athe dvvations converge “Along he same tines, we wi trp whi bok id Jon real ter as hich xx a book] [lob fead af (answer War and Peace) or “fetch ] Dot ead Bok] (answer: hat (ok) “The assumpvions re ssghforward end minimal sn spit. They carry vs nly party tovatd an analy of recosruction ad inte retain, there are eorples sind obscore Hhenemena, meny scarcely ticerstcod, Inefaras thee siimpions are erable od propery fverable, we coe elmiate reconstrtion a8 separate prooss, ping the term ony port of informal deserve epraratts for «| corin range of phevaens Txtemling erations of Van Rien and Wiliams (1981), Friin (1986) pss ut that such consratons a (34-) beave quite differ ty under reoestuction | (64) 4 which claim that John was aslep] was ewig to discuss which claim tat Joha made was he wil o discuss {In 24a) roonnrseton takes place: the pronoun dors ot take of tment conta. (40) reonsrtion is not obigatory andthe naphorie conection san eption, While there are many complistions. {oa fist eppronmaion the Contras ses to rede toa diferene B= twcencomplomn ad adjunetthe bracketed case of 34) nd (8). respectively Leben (1988) proposed an analy of this dint in tern of generale refrains. Incase (a) the complete! rs fppenr at the level of DStrctre in esse (3) the adjunct coud be ‘ljined bya gerald transformation i the course of derivation, i {ac flr whatever proses are sponsible forthe rcestrcin fs “The apprench sapling if problemi. For one thing there the ‘question of the propity of resorting te peerlae! eansformations For anther, the sume reasoning ores reconstruction in the cite of ovement. Thx, (3) it anlogous to Ga); the complement preset efor esng and shoul therefore ore a Contion C violation. (65) the lim that tin was alps to in lt Be COME Unser the present interpretation, the ace fis speed out as demic tothe ni subject Wile dls PF, it eins at LF yitine | esction. while adjuncts ean be inieded ney Mesinait Popram ms te unvante recoesructin eet. Condition C of the Binding theory ‘requires that the pronoun im canna ake antecedent within the er ‘bedded TP (compare “7 sem 1 hw [ok Jo wth him sapere 4 Jol). But hin can take Jor as antecedent in 35), contrary te the ‘tition ‘The proposal now under investigation enerome these objestions, We faye moved 10 2 fulrblown theory of generalized tamsforations ther is no problem ere. The extension propery for subaiteion ents that complements cam only be introduced etal, hence Before wh ence of {gine to he wpe afer aking to [Spe CPL. Labo’ nasi of (G9 therefore cold be eared over. As for (35), i “reconstruction” stntly arte ofthe fomation of eperstorsaiable enatucions. it sll hold only for A-chais, ot for Achains That concuion seems oie over a corserabl range, and yields the mbt rests in this {et as return now 10 the problem of nding theoretic conditions at SSiructue. We found a weak but sunt argument (oF type (289) | tose the coelusioe that Condiion € appa at $Strutre. What shost Condon A? ‘Consider constretion cha those in (26) (96 a. 1. John wondered (ich itr of ie [Bil sa fi. thestodents eked [what atitdes bout each othe] the seacers had note 1b. i. dot wondered ho tse hic pctar of Rist) i. the stents asked fubo [34 oted (what attudes about cach ether ‘The sentences f (6 ae ambiguos, wth the snaphor taking ster the ‘matx or embedded sujet as amecsemt but tote of (265 ae unan= Digtows, with the tace of who asthe enlyamesedent for hin ach ‘ther. 1268) were formed by LF raising of the insite whophrase, we ‘wou have to concute that Condition Applies at § Struct, prio _ the operation. But we hae already seen tht the atumpion fo unwer. _ leds we fave again, a weak ut sient argument apna allow ‘bading theory to apply a $-Sinetore. Aloe ook shows tht we ean so stl be, m6 ‘Chapers ‘inder the comying thers the atl for of (6) are (372-D) (27), Jon wondered otic pictre of himself [Bf which picture of inset) by the seen asked, what utes about eachother} [the caer ad nce (what tits bout enc be) “The LF incipes map (278) to eter (83) oF 8b), depending on which option essed for analy the pase wh, (68), John wondered fle pete of hrs ay (BLS Ittich ptore of ese fe John wonsered ich [i ¢ctre of is] (Baw Ivhich a tue of se We the interpct a3) (3) a5 (396), as before. (09), Jom wondered [hich xx ptr of his] [Bl sa] 1 Sohn wondered ich 9] (Bil ane epetre Rise], Depending on which option we ave sled, hms wil be snaphorc to Job orto Bl “The sae analysis apps to (2TH). yckding the 10 options of (40) coresponting 0 09) (40), the sudens asked fubat atts about ech other) [the teachers had noted the det ask hat he teachers had noticed fr anides bout each ober 1 (44) the antes ofeach eter i the stds; in (40) it he techs ‘Suppose that we change the examples of (36a) to (418-b),eplcing say took and Fad noice by hed (1) John wondered ivtich pcre of him] i ook 1 1 the oderts asked what atv about cach ke] he eaters a Conse (48), As before if am ake ether hr or Bi 5 nts ‘eden, There i Crier ambiguity: the phrase tote. can Be interpreted either ional (nthe seas of "photograph oF Healy (ick up aad wath gwny with) But the interpretive options appear 10 Correlate with the che of antecedent for himself the ated Minna opa we ol, be idiomaticimerpetation is barred: the antecedent is Bilt is petted. Bilis replace by Ary, she icmatciterpettion i exch, ‘Te pattern is sir for (410, eacept that there iso Beraidiomaic ‘abies. The only interpretation is that the student asked what str tes cch of the teachers had abot the ote ences) he teachers ‘replace by Jones, betes nterpettion, ‘Why shoal th interpretations dstibue in this maser? Fis consider (4). The peiniples already cussed i the two LE | emonsin (42a-b), (42). John wondered [which xa pictur of ise [Bl tok 3) Jot wondered ich x] Bil took fx pete of hist] I we sl the option (42. then hinslf kes Jon 8 antecedent by Condition Ast LF: if we select he option (426), then hse! takes Bl recent by the ime principle. we eelae ill with My, then (Uda) & fered. Having abandoned D-Strctre, we most asin that iio inepretstion rakes place at LF, i natural in ony ever. Bet we Ive no operations of LE reconstruction. Ths, fake plore can be _imerrted os photograph only ifthe phrase pres asa wnt at LE _ that is in (420). no ($23). fos that in (428) we have only the | toniiomatic interpretation of tale in (425) we have she. In short | only the option (42) permits the idiomatic interpretation, aso Blocking Jol as acezdent of the reexive and barring replacement of Bil by “Mery |The sume analysis hols for (418). The two LF options are (428-6) (43) 9, the stcests asked fvbat x attitudes about ech othe] Ine teachers tad 9) the students ast wha» [the eachers had [x tits about cach ote Only A) yi an interpretation, with hove. ads given its uni _ tay see ‘Te conclusions follow on the erat assumption that Condition A ‘ot aly st Strate, porto the LF rues that form 42)" Coe. | Stion A were apply at Struct, John could be taken as antecedent | & hime in ta) ae the later LF procemts woul be fie 0 choose _ ier the imate or the tera iterretation, however the recon 200 ‘aners mtcedent of ea ofer in (4B), with reconstion providing. the Ioterprettion of hae. ates. Ths, we have he siongst kind of ieument agin an SStuctae coniion (ype (25) Condition A ‘amo apply mS Stree ‘Note aie thit we drive & song argument for LF represenation, “The facie are stmighiforenrdly explained in tems of eve! of represen tation with tw proper (1) pases with e unitary ntrpreaton such ts the ion fake plture Or hie aides appear 88 ris; (2) binding tory apis. tn stand EST approaches, LF ithe oy ca ite Tue argument i il larer inthis minimalist theory, ching [D-Siructure and we ae now arguing) Stricture ‘Combining thee sbuervation with the Fein bean examples. we scm 1o face @ problon, in foe a nearcomtracction. ta (A) ether option i allowed: ims may tke eer John oF Bilas snecednt, Tcomtast sm (4) reconrocion appear to be freed, baring To ss amterdcit of he (by Condon C) and Bilas antecedent of hin (6y Condon 8 (44) John wondered ivi ite of hiss 1) 1k ohn wondered fick picture of Tom] fhe iked 1 John wonder [vb pete o kin] [Bi ook fi, Job wondered nha tee about bi il had) ‘The Frsdin-cbent: theory regres reconstruction in ll these eases the of ae bing 8 complement of pce. Bt he fs seem to point tore conception shat distinguishes Condi of the binding ther: Which docs aot force reconstrrtion. from Conditions B and C, which nat be Ted onthe same lena eeices We can, the think of Cg map fing some array A oflesital choices to the pir (1). What it A? At fest, it must indicate wat the lal choke are and how my Ge ‘ch i scketd by Cu i forming. 2)- Let stake a muerte tobe {et of pis (LI. 2), were Ll a tem ofthe lexicon ad ke nde, ‘enerstond tobe the numberof tines tht Lis eet, Take A co be (Gt leas a numeration NC, maps N eo (8,2). The procure Co eles an item fom NY and odors its inden by 1 then performing ‘ermiibe computations. A computation contracted by Cy, des not ‘unt 36a derivation at let aloe @ convergent ne, unless all ides sreriduced to eo, ‘Viewing the lngape Las a deivaton-gnerating rosedre, we my ‘hink of applying to 8 numeration N and forming sequence § of ‘ymboe deents (6... 6) erriatng cay fa pate, 2) and Nfs redyced 10 zero (the computation may goon). § Formed inthis ‘vay i «derivation, which converges if the cements of 0, satisfy FL at DF and LF. respectively. Economy considertions set the admisible convergent deivations. ‘Given the mumeration N, the operations of Cy recursively construct sancti bjs from tems in N an syntactic ebjes already forme We have to decrmine what these objects are and how they are cor _srted. Insofar a the condition of inclusiveness Rok, the syntactic objets are rearrangements of properties ofthe lexical items of whch they ae ukimatly constituted. We consider now jst the cemipotaton| Noi. forremons sat will Besome elenrer as we proceed and tat tend to support the view that there is indeed someting extraneous” about ‘he conditions imposed on language athe AP (encormotor itefae ‘Seppe thatthe derivation hae reached the sage, which we may take to he st [50,,., $0} of smtatic objeas. One ofthe oper ‘on of Gq i prosdarethat ects esi tem LU frm the ner tio, redicng is tex by 1 and introduces it nto the derivation as $0, Call the operation Seer AL he LF intertse, 2 can be ior ‘rte only ist consists of single sytatic object Ctl, then. Cy ‘must include «second precede tat comes syracti objets already Termed. A derivation converges only if his operation has appied ofr nou to lave us with uta single otc, aso exhausting the iat "numeration. The sips Such operation aks pai of sytacie objets (60, 50,) snd replaces them by @tew combined sjacie abject SO, {Calls operation Aderge, We wil rezira to properties, merely noting ‘bere thatthe operations Sele and Mee, or Some close courterpars. areaccesnry components ofan) theory af atu! leguase ‘Not thar no question aries about the motivation for application of ‘Sekt or Mere inthe couse of «deviation I Select doesnot exhaust the nuntaton, no derivation egunerstd! and no questions of comer ‘ge1ce oF economy ase. escent application of Merge hes the sme property, since the derision then fis oye an LF representation My aeai, no derivation i generate, ad questions of eanvergence and Seonomy do ot se. The operaions Select and Merge ae “costs”: they do fot fll within the demain of discussion of convergence and ‘ccnomy.* Simi, we donot have to atk about the effet of ili ‘ate opertions, aay more than proof theory is conceal with as ‘ques of lines Wat oes not say the formal conditions chat deine | copie en Tartration, ™ proo.” or a cheseplayng algorithm with evaluation of improper ‘Within the framework jus oun thee i so 90 reasngt ques tion ao why one numeration formed rather than ane or ether than none, 50 that we have silence, That woud be lke king that a theory of some farmal operation om integes—siy,addition-expsin shy some integers are aed togter rather than eer, or none, Or that a theory ef the mechani of wson or motor coordination ena thy someone cheoss to look «sunset or reach for bana, The ecblem of chic af actions rea and Inrgly mysterious, bt does not se within the nartow study of mechani” ‘Suppose the leva ite Lt a the numeration N has index Wa dei ‘ation to be seers, Seat must sense LI / times, itrducng it int the derivation. But the syntactic objet formed by dine appl ions of Selet vo LT must be distinguished two eourrenees of the pro own efor example. may have enely ilrene properties at LF and ate ths marked as distin for Cy. i they reformed by eine p= plstions of Seles aecesing the sae Jez term of N. Not that hi 2 departure fom the incloweneeconiton. but ons that sem Pesabl: its rooted in the mature of tanguage, and paps reducible to be output conditions “We wart the nil array A, whether a nuneaton o somthing eli, ot ony to express the compatitlty relation Between mand A but aso 'o fis the reference st for determing wheter a derivation fom A 10 (6.3) optimal —tha isnot blocked bya mere economical deriveticn, Diterminaton ofthe eer et in dint protien, sae conse. ‘ions of econony gers. As x fist approximation, let us take the romertion wo deci the ference sett evaluating derivations for conomy, we consider only alematies th the sme numeain, Sicton ofan optimal dation i the reference st determined fom ‘he numeration NY poses problems of crepuations complet f00 vest ‘o be raise. We ean reduce the problem with « more "keal” inter taon of reference ss. At @panicuar singe Eo & devivation, we ide only continuations ofthe derivation already constr —in erica, ony the emaining parse the numertion N. Application of ‘he operation OP 10 Zis bared i this et contains # more optimal de ation in which OP doesnot apply to E The umber of dation 0 te considered for deermining whether OP may apply redocesradilly ne Caper 5s the derivation proces At east this much srectue seems 10 be ‘eguired, presumably more See section 49 fer some empiric! evidence ‘supporting this eonsrat of rerence st (whic, i any eben i 0 Be peter) in fic, a een ore stringent condition. ‘An elementary empinealconiion ce the they is tht enpresions “use” by the pesformance systems be asinine epreetations in a munner that dos nt induce yo mich compaationslcrelen, We want te formulate economy condions that avoid “exponential Dowup" in consnuction and evaluation of deviations. A fea tr pretation of reference sts © sep in this dvecion. Where “lat” Properits of devations have 10 be coniered a in dcterming the applicabliy ofthe principle Procrsiace of eater hapirs, we expect to find some ea algorithm to rede computational complet. Ta the case of Prerastinat, it peal fies ose if strong feat p= ‘ent wheh i srighforunrd~ and even easier under an interpretation of “arena” to which we tern diet Bot we are all atone wy from a comprehensive tery af economy. & topic that i om Bie ox plored forthe St ie wn content of ingry Utah oa “xpliatory adequacy om he rear aged Given the numeration N. Cy, computes unt it forms derivation that convergs at PF and LF wih the pe (9,2), afer reduces N (0 ero (fit does). A “pers tingusge” shoold meet the condition of in Chusivnes: any steer forme bythe computation (in paul nd 4) i comsttc of ements already pesent in the lesa tems Se lected for Ne new abets are added sn the couse of computation apart fom etrangemens of lea prepetss particular no indies. ar kel inthe sense of hae theory, ee note 7) Let us asm that tis condition elds (italy) ofthe computabon fom N to LE (08-2), stand eons tke 0 be rad fale fr the compat tion PE ‘A lieady noted, the inchsvenssconition not fly met. Dist using ststios fs single ea ite is 8 (ether narrow) departure ‘Avether iveloe the elton operation (Delete a) Let us assume the ‘tis operation marks sme object a5 “seat he iterace: wil sharpen it a6 we proceed, assuming forthe moment that mati de feed: hough ignced at the iterfe, i sti acesibe win Cy, The question ams ott hve interesting rmeatons ‘core property of Cy i feature checking. the operation that dss moverent under the Lat Resort condiuon. A ay par four eoncer eee on rasterstions 5 ‘sil be to examine thie toons. We ean begin ty reducing feature hocking to deletion: a checked Feature ig marked “iminble te ne {ace Even a cursory look shows thar ths cannot be the whole sory, tet ws take it asa starting poi, etutring toa ore erful nly imation 432, ‘Ostput condition show hat rand, are iret consid, El rents nerpreabl atthe A interac reno terete at C1, 0 ‘anvesly AL soe pein, then, the computation spits into we pas, fue fxmirg and the other forming. Te sips assumptions ae (1) | sha tbere 8 farther neracontevcen these computations snd (2) that computtional precedes ar uiform thooghowt: any operation ‘em apply at any pont. We udope (1) and ase (2) forthe compue- Hon fom W to though mat for he ermsputation fom N to the aie toes srctres ching he irl strate of filets) by prosesics very ferent from these shat take plage inthe N+ cor Dutton, Investigation of output conditions should suis to ahs thee asymmetric which ¥ nil simpy take fer eran ee ‘Wie assume, the, that at some’ pointe i the (uore) compuacon 1 LE, there ean operation Spel-Out that apis to te stare 2 luedy formed. Spel-Out stps away from E those elements eleva ‘nly to, leaving the reside By, which ie mapped 0 Fy operations ofthe hind wn to form EE itl then mapped tom by operations Unt thow of the N=) comptation, We call the suboyien of Cy ‘at maps X to x the plerologcl component andthe subse tha ratinos the computation Irom Z, 10 LF the cont component, The Je-Spel-Out computation we ea ort, Let us assume farther that Spelt Out delves tothe module Morphology, which constrcts werd tis nis that are then subjected to Frter phonological processes that ‘tapi finally to, an whieh einatesfentres no Tonge lean 10 ‘the computation. wil hae itl o say about the Phonological como foc here, except er some cozment bout morpholoiel rice tnd loca orden Tre special properties ofthe phonclogial component relate to the ed to rads instrucions for semonmover stems for prodiction fa perception. As noted, thi eguirement may he the scare of eter Inperectons of Cy. ad i hs seni “etraneous to Inge poss Talis we wal epiore Giver these fly elementary assumptions stout the sructure of Gas distinguish Iwo types of ei! featre hose tat resive an 20 caper Interpretation oaly atthe A‘ intsice (phonological) and those that ressite an itespetaton nly atthe CH itesace.Fssume Further that these ses are disjoint, gen te very special propenies ofthe phonoke- «al component and its PF output Tis reasonable to supose that overt operations do not delete phone: Jogi ears others, there would be hite reason for them 10 ap fear ina esc tem at all. Suppose this tobe so. By the assumption of ‘rioemity of Cs it low tat cover operations cennot do so er, ‘any phonolop festus enler ts eoven component (air SpeOu the derivation vil crash x1 LF, solating FI. We will ae the sil “sronger sumption shat overt operations cannot detect phonological fears ot all such fears cannot, for example, distinguish one over ‘operation from another!” Thus. the phonological ematrs of lee ier is sembly ato, as fr as overt operations ae concered. Tis the form in which the instructions for certain rls of the phonelogcl ‘omporent are “code” in helena tem For the N+ compsaton, ‘nothing would change if the phecological properties of book were coded i the loion a5 23, with a ue of the plonelopcal component inter Preting 2 the phonological matrix Te Book “Arsong the features thet appear im lic etris, we disinguish further between forma entures tht re accessible in the course f the ‘computation and others that ae ot: thus betwee the Formal atures [EENand plural andthe semantic entre (tet, The bass or he ‘Sstiction and its fet nse susan! questions among the many ‘hot I wil pt mide Ree Soc Fentress Felon eileen in he ‘onelopical component. Since we take computation to LF to be uni form, we eanot stipulate that evan Features are eliminate only after SpellOut; bt the maping to PF has comple cient properties nd eliminates fetus m ays not permite in the N+} computation in prtouly, teint forral and sera etre, “The lena enry for apne for extmpe, contin we cottons of fates: phonological ferture suchas [bepine wth vowel, sean eater such as [ats and formal features sich e fomial. The rhonolosicl Fares are sipped away by Spel-Out and are thus svaiebl only the phopolepcal component: the others are ket bind by Speti Out, sad the frmal ones may comin to be accessed bythe ‘over computation to LF. Within the phoocogicalcanponent, no [howolgka fetures re siminaed i the couse ofthe compuation “Cacgone ond Taforataes am “though they may be relevant 0 ts operation at se it eater pars tin the morphological subcomponent. “The collion of fornal features of the Ica item LU 1 will cal FFL. 1 sbcomple of L, Ths, FFlapine) ste eaten of et teres of alan tat function in tbe Nh computation, excluding he ‘hosclopcal and (purly semantic features. Some ofthe fates of FPF(LD are turin oi, exter empl nthe lexical ety or | sey determined by properties o ltd. Others are optional, Sed {enters the numeration. We wll turn wo ths mater in sso 42.2 Iisofaray wear considering propenies f the computation Irom mime sion to LF, we eric steam formal estures,thogh be ouput ‘eptions atthe A-P interface somaties force a epartire fom this eideratun: weston 44 Ie the cae of airplane, the insinse properties ince th categoria fexture [nomial he person feta (8 person), and the gender featre {E-humman} Is optional properties include the toncateponal features of umber and Cas. The icine properis of bul clude te eategosl feature [eb] ao the Case feature asgnsceurative bat ie fetes | stn are optional tral othe en). Chow esi! et LL ‘wh diferent optional Faure are dsinct members of the numeration |W Ginpane1) 6 the rumesaion, #8 Fes term must inl the ‘cpl Feature [nominal] andthe noneacegraesures [ peso. | {human aswell one or anette choke among number and Case fexures— perhaps pra and [eccstne i which ese may spent ina convergent dectation for wet olan Furies erly evs ‘ations disintons and complenty that donot Se to elt 1 he © sompuicna procedure Cy. teat those aspects af 10 which Twill, list etenton logs with «host of further question about onda, sbrucre, and interaction with semantic fentres, which ignore hee Torte sme reason~ perhaps tmpropery as ure inguiry ay eves ‘A guiding inuton of the Minmaks Program is that operations apply anyuhere, witout spl stipulation, the derivation essing i “rong choices made. Le ws asume this to be ue of Spe Out 25 0f ‘ter operations. After Spel-Ont, the phonological compote! canst sek from the surserstion any ter with emai fate, andthe co- et component cant eet any it with phonological feature, Tt {is requirement for any theory on the weitest empirical sesuapsions, ‘there, sound reaing relations would cllep = m Chapter nis unnecessary to add siplations to this effet. For the phonoloi- cal component the question does not ais. It has rules of a special ture, dic! fot those ofthe N+. computation, and these only ‘ody forms alieady presented to then. Accordingly, Seti iopers- ve inthe phonolopal component: items can be Selected rom the urmeraton sn he computation fom Spell Out to PF. “The opaton Sela is sya 0 he covert component, Howe assuming the uniformity codiion onthe N+}. computation. Bat an item with phonological entre is sete, the deration sil crash LE Selection of LI must he overt. ules LY has no phonological ea tues. In chs ease LT can be selected covertly and mere atthe 08. like overt merge, fr simple reasons to which we tetra). We wil se ‘hat this consul possiiy ray wel be reid. ‘One interesting ese concees song festures: cana (phonslpicaly esa item with sro estar he ease covert? "To cay the mcs we have to mie hentai he sent prop cy, Featrestengin i oe kamen of guage vite oral fe tre may oF may not Be rong, fering overt mewement that wiles roerastinne. A Took at eases gests that the tng meron i srry restricted, peshaps 10 something ke the ste options (0). (2) IFFisstrng, then Fie a feature ofa sonsubstansive category and F ‘sehecke by categoria ate. 10, nouns and msn verbs do not have song Features, and a stone Feature abe cals for conn cloory ins checking damn (0% sy, Case or ears). flows that oes nverent of ageing. forming (Spee 0 0, Bohs possible only when is nonstate sa calegnra festa of fs vaya in the operation. Ths, the Entended Projection Prcile (EPP) pausiy edu to a strong D> fenure of J and overt whraisng tox rong D-fentte of C (assuming seheto bea varamt of B (Determiner) Otter eases would incude Ov [ising to D (Longobard 1984, and source tal), and Ho-C ran row underwood as volving 901 Agr or T bat rue modal ar V a= Joined to lun ide ht will make mere sense as We proceed. Adjuncion ‘of mina to transitive vers wil target [FV] comple formed by ‘ising ofthe ain ver Veo aight verb, an! verbincorporation woul kote tom wenk verb, Lat ws asime tat Something ike this isthe case For esse of exposition, I sometimes speak of functional enesory 2 song when I meen, ore exp hat one of its Features strong | sen before Spelt Out. Second, tego ud Teansoreaione m 1am alo plosing over a possibly significant disinctionteteo D fentres and Nf ha i, among thee variants of the EPP (1) seuirng (DP as specifier. 2) requiring an NP, reiring & nominal estegory, | whether NP or DP. The difeences may be siniant | wil turn to {hem in sections 49 and 410. Un then, referees to the EPP wil be ‘expres in terms of strong D-feanres, but the tention fo renin ‘tera arson he choice (D2). 8) ‘A song eure has to properties. Fs, it wiger an overt oper duces eet asrong fate can ot be pass by Ut wool sai and tater checked by tha sould permit Relaivasd Minima violations (We Isand,stperas- fog) le chapter 3 the gre Spe Ont property was std in was Of eo sereoce at PF (a song feature ersten at PF and therfore mast be semoved efor Spell Oud, but that formulation was bas on stil tion thar we bave now droped: that leva! access tales place before Spe. Ost. The ectic property was fe nly paral resolved n chapter Stand in Chomsky 19540) ‘Apt from its protims nd Knsitations, formulation of strength in scems of PF convergese i a restatement ofthe fase property, pt 8 true eplenation. In fact here sexs to Beno way to impose up the ase datement ofthe properties af sength. Suppose, ten, tat we Pl ‘an eed to evason and simply deinen song feature as ne hat a der ‘ation “canner elerte i dervation Dm i canceled i contain | ‘trong feature in a sense we must ake precise A strong fete thus eee a rule that chniaues ic (sreng) is associate with a pair of ‘Operations one tha neues it into the derivation (actualy, weer in of Sekt al Mere, a second that (uit) etaates ‘Gytety follows at once.” We als aru derive the condusion Ut srg feature gers an ort operation to cing by check ‘ing. This conclusion flows witha single exception: covert meer (at the rot of xia tem that has strong fate but no phonological __fstun—an option noted emer, 0 which we eur, tis peraps worth mentonig inthis connection tha the Minimalist ‘rogram, right or wrong. hus a cera therapeutic valu. It ll 100 289 to suse tothe temptation to era purponted explanation for __ Some phenomenon en the bass of assimptons that are of roughly the order of complesty f what isto te explana’ Ifthe asunptions Ive broader scope, shat may bea step forward i understanding, But | Romecines they do not Miniaist demands at est hive the mei of ae taper ishing such moves, ths sharpening the question of wheter we have genuine expleation or estate problem sn oe ers. ‘We have to determine in wht pace sense 4 ston feature cannet te ind within lett derision, The inte idee thatthe frog fentre merged a he Foot mast be elminated before it becomes fn ofa lyger soctre by frtber operations. The notion “part fn be tndemtoed in varios ways: Thre are four possibile, based ‘on the two ways of Pung ew statues Gubstittion,adjneion * land the to options for preketon (either te euegory withthe strong Fen othe os ened with it ean projet. “To isa take the case OFT (Tease wth a strong V.feature and » rong Deestre 2s in rene) forcing overt V-raising tT (aumetion) and ovet DP raising 10 [Spe] Gubsizaion), We must to knew how ‘Tand some etegoy K eas be jin to form & large eatery L, an Sse wit the stergsh fT. ‘Suppose T and K are oie and T prea. Suppose the operation substitution forming L~ fy TK] th head T and complement K. the fHrong fare of T remain inthe larger Sucre, Pliny that 6 a- ‘missles Fact, the oly way’ Team enter convergent derivation Projection of sang, then permis tbe devation to continue when T tnd Kare oie: the projecion of Tean tolerate emedded sone T Suppos that T sod K ae joinal and K projets. Then T i ether the specifier or complement of K (ubsiuoa), oF an adunet of K {cuttin} For enone tht wil be cued as we proceed, subsguent Joining of Lo T (by adjunction or substitution) i breed. In gene then, se wil ty to stabs the pepe (2). (2) Nothing ca jon to a nonpeoeingcuesory “That i nothing can join to 0 aur, pier, or complement. Hen ve need not consider the ese of nopprejeting strong, for if a stone Feature doesnt projet, the operation eguzed oeiinate strength wD ever apply Te hi the corms inerpretation ofthe eptions then the esrptve property of strength is (). Suppone that the deviation D has formed E ‘containing witha tong feature FThen (8) Dis cance fis nx category not headed by 2 “The cases just reviewed follow, 8 do others. Note hit his nt pei Cine poering sent at «descipive obvarvtion abou 2 ol. | Bits those of UG, and thee of speci | ators on arsfertions a Suyose K adjoins to TP, forming the wo-seement category M = | op KTP. By the dvivaion tolerates he strong fenttes oF, which Gn bests by later operations. This isthe ight result. Suppose Kis ‘ay adverbial. Then TP can te extended 0 M = [rp K TP by sdlonction | ofthe aera Kad M ean be further extend to N = [y DP Bf] by _ insertion of DP a Spc, T} to satisfy the EPP yielding oh expressions | a8 Jol probaly has Wt aveady, sere probably wl be moe tomorow. “Waa the NP breaks typi posn for merger of avert | Suppose we tae formed TP with head T and complement K. and he __ strong D-eature oT (EPP) as not yet teen checked Suppose we nen pe TP wih C Forming CP with head Cand complement TP. That || exhed by the desivaton wil rash. Agi. hii the Fight eeu ‘vite severally, equiva to avoid Relative Minialiyviteons es “Thase cas are typical. We therefore assume (2) hol! for stone | fetus, “While Merg scot for principe reasons movement it not the | operation takes place only when fered (Last Resor) and it oven, | Solatiog Procrestite, only when thai reir for convergence If asa song feature F i wigges an operation OP that checks F before ‘on ina rather taonl sense as Hist of “exceptions” whatever does ‘ot follow from general principe. These principle fal ato to eater euage. The later cover | Bret of phonology and morphology, choice of paremere options, | sn watever cue may ener into langage variation. Assine further the he econ provides an “optimal eating” of sch osyereses ‘Take, say, th word bak in Eni. H tas colecion of properties, __Seme iionreratic some of varying depress of generaliy, The cal _nty fr Book species the Wsjnersies, abtacting fom the rings ‘FUG andthe special propre of Engst. I the epi eoding of _Sformauion that just suis to yield the LF represenstion and that los the phonological component to constr the PF repesetation: _ he asymmeny refs the diferene between the N+} computation 5 hoger sed the phonolosical component. the former (vl) stfyng end ‘he ter rial oatng the pipe wiforsity and inusvenes ‘One idkoeyneraic property of a coed inthe fecal entry isthe sound-eaing rion The lexi ene also iter ists, o ents, hat it has the eateprial feature [NE we overlook open questions of seme ingerest here” But the Isc entry should ot indicate that ook has Case and gears: that fllons frm being of category N (pre Surntly. bp princi of UG). It should alo net spay poate or semanic properties that aze universal ot Engishspecie: predicable interotows between vowel sed Gal consonant, othe fet that Book fam he sued to fete to something that is smulaneouy abstract and once in the expe the Poo tha writing will eh Spans. "That va property of road range of nomial express perhaps ll ‘one of the may restr why staaird theories of reference ae et sppliaeto satura language. ny opinion (ee Rte 2) For te word book, it seems the optimal coding should nce a pho- nolo matric ofthe Fria ind expressing exc what i Bt pre “abl, and a comparabe representation of semantic proper, abort Shick mach les fs hnoken. And shoud cde the Formal fetes [ot book instars they ne unpredictable fem oer properties cf he lescl entry priape itr eteporal feature N, and no thes. The fit that Case and fetaes have to be asi toa lle Fem 2h ral principles and nothing itis to the ical entry bok ells ws ha {partic ceirrence i singular ce plural, nominative or accaatve {hough is pero eater prenaly cerned by iia sean ti proper). tn some cases soc entre might be thyme ‘he purl Feature of sir, of grammatical gender), More pono in some langeaps the asta works very diferent: for example: Se tite wth rotvowel ptler suture. But lial ents are deter ‘ined onthe ie peer around. ‘Suppose that hk hose as part of he arta fam whigh ade tio procesds to form PF and LF representations, ¥ have ser the hoe of book a two-step procs: (0 form a terion that nla ‘on with nde ad 2) nedace Book imc the deren by te ‘operation Selec, which as ok wo the set of stati objets stor Stand reduces ts index by 1. The optional festores of «partite ‘cecurene of Bok (8. [octsaivel, (ral) are adie by eter se (or sep 2) presumably by step. decision that redues reference sets end ence computability problems. Suppose so. Then the numer Colegio Tefomitons 20 ion N wil inde [oh accuse fra. 2) ths axe esuning fedex 2). Theft that these fenures ate preset is determine’ (wea sue) by UG, but he ence mong then ne, Recall that'we are concemed here cals with mechani, not with | chokes and intentions of speakers. Over the cat of derivation, then, {he mapping em xine to marcato renders with rep to pet ation of ook for Case and features sd he index ofthat alton sf properts. hough UG resutes that there ie tlnays some choice ot (Go, estes, and inde, ‘Thc much seams ry char. 4s analy plait Case ond fenars of oak are determined by spoon in + claus coniguration the word is wed in isolation, these Fates wl be ned ove way of __awiber though thre is no stucire. One could sy tha thee & ie suppose? situs” some represctatin of he intentions ofthe sper ‘(possi shared assumption in some interchange. But that surely the wrong course to pursue. Is possible (ard his teen prope) tht ‘ous ate automaticly sete along with Broader nominal configu tors (inching Cas, perkups festue), That i a posit | ould require postive escent. wll ese here the tl hypotess | Cx and @fetures are added arbitaily at pou ele forthe ‘Ke sume concusions ae aproprini in other cass, Take sich con seructions 25 (8) (4) far as John is concerned fe doubt hat anyone lever wat to ‘peak 0 the foc agua) ‘ere some formal properties of J and a fet ate rested, bat not by the mechanisms of Cy, The example falls lst with uae of CP in Slatin on the tackgrouod assumption tht lon is under drewsion (osthaps provided by discourse content, perhips nl). The ste might bet of te more intersting ease ef nominal expresso in aguas that expres arguments as adjnet assciated with pronominal Semen _ ethin word (je Baker 1995 and section 4.10.3), The partie form fakca by such adjncis may depend onthe asedaton, But might Hot be _ Sirebe i ems of eel relations f the kind admted in Cy yp. _ aly, Het relations, where H ia head and isis checking dain). a the numeration, then Cae and ects of nouns ate specified. _ Whether by the lercal entry Geto Feature) or by the operation tht _forms the nuveration optional Raters). Larger srctrcs we cea 2 chopeet but only for cocking of fstres of the nou that are already presen i the nemeration ‘Comer verb sy, enlace too represents i the op- tinal way the instructions for de phonlogeal component and for iter pretation of the LF representations phonologe] matrix and seme ray of semantic properties. It must als contain whatete information {sprovided bythe ver isl for the operations of Cy. Te lea etry ‘est sie to determine that expla haste eatgoral property V. pe aps by expt sting What about its sectional futures! Insofar 36 thene af dtermied by semantic properties, wheter by UG oF by se ‘he rules of Engh, they wl ot bla in thence, Theft hat explain has tense and -eturs wil ot bens in helene etry. tcse that much i determined By eatery (resmably by UG) “The particular specfention of sich fentare, however, nt pat of the leical entry. The verb aso hs a Casoasigning propety. which is in tans: ether determined by properties of the lea entry i semen Features) oF edo hoeyrerati Feature that are asad withthe ‘ert but ot predictable rm theta! entry ave vo posse sources: ‘hey might Be chon ebay am the verb ene the numeration. oF they might be the sul of operations tat orm complex words by sociation with other element (eg. adjunction to T). These could be ‘operations othe overt syn oth ponelopical component neat ‘oriolees. If eer syntactic operations ate inoled, th exert involved il be marked in the Ion, oF the waraitin the mire thon) ae allowing or eguingaaation. “The acsions ae les eat thas for nouns. but the pips ae the same. Whatever information feeds the phosoloical rls rst be aval sie for she computation asthe ie fisrodued into the deivaion. Trt the specie charac ofthis ifareaton has to be dscovre. 8 sell as whether it is provided hy eponal die of eatre values asthe em eater the numeration of by overt syntactic operations. Take in fected explain. It tone and gfetares ight be chosen optionally and sesined to the word neers the urzraion. oF they igh rs from overt Vans to Ar and. Or the werd mipht each the phone- Topica! component unnfite, te PF for rsuling fr interaction swith fanctinalskments within the phonologial component. The a Ser could vary aren or within tngwages¥ The quastons have tbe nerd ese by ca Categorie and Trantoratns » For my specific purposes here, t does ot mater much which choices = tun out to be correc, until section 410, when the tatu of fneiona ‘categories eases. For ceneetenss (nd with am ese art fo that | tater reasessnen Twill sume tha ese ad entre of orbs | afected explain ae chosen optionally the ern ener the eames | ion, then matched by oer processes. But akernatives te compatible th meh owt fotos. A separate question f he fer in which the infertion shoud te | ceded in the eical ene. Ths. the ease of hook, the optima repre | senttion in the lexicon could inclde the standard phnologa mati PM, or some arbitrary coding (ay, 2) interpreted wth the phono Fon component as PM preunahy the former, unk thee fs | strong independent reason forthe Istria, he cave of past, | ens) the ft that itis peal dental could be represented! by a por ological ates eo), o by an era coding (st, 47) inept “the phorolgial cmponent as ental (vith whaterr compestions | ave to be coded in the entes of ieguae verte) aya, pet = tly the former. unless there i stang independent reason to the cone tray. Til put thes matters aside, asuming tha they ave 10 be = tiled case by case, though in 9 manacr that wil wot mater foe what flows. ‘On the simpler srsumpsions the sel enry provides, once and for all the formation teqied for frter computations in partly forthe opeatons ofthe phonological component Gicuting moral | ty. we aseune), There sees oe no compeling enon 0 depart fern | the optimal sumption ___Uhve kept tothe easier eases. Let uF move tothe opposite extreme ‘Sppse the PF form ofa lei! erity completely empraitae: the Enel copula, for example this ase the ec oti wll provide “whueverinfornation the phacclogia ues aed to asegn 8 fr fo the etre feopul, (F), where (Fi some st of Forma eure son et) H doesnt scr to mater for ot purposes be test) how this information is presente: as ast of aerant, cach with - fori featuns, or by sme coding tat allows the phonlopital compo: ‘ent opie the lena ate insertion). ‘hire the worst peasibe case Ply. it would be 2 methodological ‘ror to generate th wort case tolls tine erm the fc hat the worst ease exis that old fr al exc ians, 1 caper “Theeure many interne cts, Take the lene demen comers Some repulrity cam be exacted and presented as phonological ma tnx, perp (RV. Bet the cice of vowel ot eule governed and therfore ms he coded in oe or anotter way. Structural gis ‘devoted mach erergy to whl he information sould be code in the form of morpheme alterants,iten-and proces rules that change one form to nother, and so on. Asin the ese of pare suppetion, i 5 Hot ‘den that there even i an emprcl soe. Thee ee smiar problems, rea rc, concerning serantc Features (om steuture et). ‘Wilh regard 1 fenetional categories the same general considerations spp, though new problems uri, Tt elon tt the leicon contains fubsamive elarens (nouns, verbs.) wih thei Kosyerate rope tis. And it sat lest reasonably clea that i contin some Fncionl cnegores: complementer (©) for example. Bu the station i more ‘ica inthe ease of ether posible funciona tego, parca, “To Aas, spc features 4 Cane eatepory K. sod so on, which why theories about ese mntrs have vere 80 Over the years. Pestltion ‘ofa funcional category hs 1 Be sified, eter By output condions Iphonetc and semantic interpretation) or By theoryisteral arguments Ws bade of roo, which soften not Ses) 1 meet. “he fnetonal ateorcs that concern us paricuely bere are TC, sand Athi formal proper are the prary Focus Soughowt , and C have romantic properties: Agr does not Ths, Ts [fie wich turer subtivsions and impistions about event strcure and Perhaps oer properties may be he lous of what loosely alld referents." i basly a indicator of mood or free (i the regen sense dectaraine inlropaive snd 30 on. The choice avon the options of sven te i ares, part ofthe proces of foring ureration from the lic, 38 (tat tenatvely assuming) in the ese of features of vrs, and Case ae (come) features of tours Funsional catepeies may abo have phoneloical properties. Thus English Tis denial and deslaatve Ci ht (ith null pion) we will return to imerogative C~ Q: Japanese Cues ate phonological in ‘ariat: and soon. The fsa ent again. provides av opti coding fox what ot peaictble ‘We expt diferent desiions im efferent eases. depanding on desis ‘of the phonological component the Heal inventory forthe lngtng fn perhipe more. Suppose. for ample, that spec morphologt properis ofa langage consi the phonetic celle af Fra Cagis ed Tearsormaons a ‘es sy, that verbs inete person with pfs aa nner wth su fies o hat extn sots are avaiable fr speling out feral Feats “Ten the level eis wil absiat from thee properties, presen jst the information that they do nt determine. I appears to be the ase Tor some langues that “template” conditions are requted for mor | pho strctre, I othe option is univer aia, ear a junction operations slong mith whatever ee the emp faci may oo | Trallcass she principle i cer. though ite answers to spaiie ques “os ae ne the lexicon provide he opti coding fr “exceptions” | Thoveh important typological difereners doubs ext, tee sere |__ ie renson tc exer mich of any peat mans nga oF in | Within them Perhaps Jespenen was core in holding thst “no one evr | siemed of a unveal momaog.” to any ferreching depe, or phology being 3 primary expository of exceptional aspects ef particule oc am Keeping tothe optimal assusption: tat for each kx tun in | 4 articular language, the idiosyncratic coins are given in «id | Feet etry. There are more complex thors Hhat seater the pops fies One mht propose, for example, that fra features, insur for phioncloia us and instructions for LF septation sppea in | Gsonct subledcons, whic are neces at iferent pins in the com | uation procest. Sich eatorntions mig lo invlve new kel nt _ lions among various peri of the derivation, o ensure proper match | Ing oF PF, LE) pass. The arden of proof fe evays oo the proposal that the thcory as to be made more compe. andi a comierable 1 have sid moshing abou other major componcts ofthe tory of __ Nord Formation: compound forms. nelainativesrctres, nd moth __ More. This i only the bares sketch, intended for no sore thn wha os ‘he development of Xchr theory inthe 19 an cay stage in the foe to eslve the tension betwen eaplaratry in deste 20 caper ¢ age, A fit wep wa to separate the sion fre the computations Tae removing a serious sedundancy eweon leseal properties thd phrase sinictone rules and allowing the later to be reduced to the Simplest (content fe for, Xcbar theory sought to eines rks ‘ogeterlsvng nly the general Xbarsheoretie format of UG. The promary problem in subsequent work was vo determine that Format bat vas astumad that phrase suctre eles theres shouldbe eli Ae, we understood enough about the mater which, neds 10 sy ‘we never do, so (unsurprisingly) many open questions rein, incline See that are quit ceil to language.” 1 cilr papers on economy and inal (chapters 1-3}. Xbar cory is pressed. with specie sipulted properties. Let us 0% fuse thee atimptions i cial analy ashing whet the theory phrase strctre should lok ike on mininatst assumptions and what fe consequence af ot he theory af maven ‘Ate UF iri, i rast be poset ses a ec iem 1 0 ss nonphonolopcal properties LF(LI th semante properics and the Formal properties thet are otrpreted there. Accordingly, LT and UF(LD should be avaiable for Cy othe natal ai assumption, dis ‘used eater thar hare caput conditions determine the items th are “vile” for computations. tn addition, Cy, can access the formal fs tures FF(LD by definition, Ie also appsent that some leer eis consructed of lesieal ens are acces along wih hae types Hou prass and verb plnasesimterprete, but fre, in tems of thir fypesand So 0. OF the lage ui, sems that only masa proj ‘Gon are evan to LF interpretation, Assi 0. hare out cor- ions take the concepss“mnial aed maximal pejeton” avaiable {0Cy_- But Ca should beable to acess no the projections. Given the incisive condition, minimal ard rasa pjetons are not idemifed by any special muting 0 they mst e deeeined from the suucite im which they appese follow Muyskeo (1982) =| taking the to be relational proertis of exeperss, 20 propertics it een wo them. (See section 13.2, Below (64). There ae no such nities 1: XP(X™) or Xin he strctures formed by Cy. though ¥ contin fo tse the informal nowons for expsitory purposes, slong with X (Ober) for any ether eaegory. A eateory that does not projet xy furter fsa matinal roetion XP, and one that i ots proystion st {lt isa mini projection XO" ry other san X. visible tthe ine face a for computation As we proceed, wll us the contusion somewhat for X° catsgores ich havea very special ok ‘Afrier goa i 1 show that computtion keep t eal reation of 110 terminal hed. All pins of UG should be formulated in thee ters-—whis have to Be made peeie~and only sch elation souk) De eevan at the interface forthe moves hat operate here ‘Given the numeration N, Cy. may set an te font N (ein its inde) or perform some permite operation on the sync objets aleny formed. As Sscussed ear, one such operation ney 09 ‘onertusl grounds alone: an operation that fers large unis ot of ‘thee aleay contre, the operation Merge. Appia to two obacs ‘and fk Merge fers the new oct K. chniating# and Whats KP K rust be constituted somehow from the eo Hs and fhe aly ‘alber posibiies are that K i fed for all pits (2) oe that Hi Tarwomly sete, neither worth considering, The sxplet ob com ‘trot fom cand sth st (ei, 0 we cake K to ivole et ae ‘this set, where and are the cst of K. Does tht eulice? Out put conditions dictate xherws; thus, verbal snd nomial elements ae Inierpetd diferenly at LF ond Behave diferent in he phonolog ‘component. K mast therefore a ast (and we assume at moat) be ofthe ene 1 2%} where dents the type to which K belongs indicat ings relevant propertics. Cal the ae of K. For the moment the, the sync objets we are considering are of ‘he following type (9) tesattems 1b. R= fr (oP) wheres, Bare objets and ys the lak! of K Orcas f type (5) are compres of tue ied ia the lexicon, “The recive step is (Sb). Suppose a derivation bas reached sae E = {8.8}. Ten appetion ofan operation that forms K asia ($0 comers E to = {K, dy... 8) intding K but not 2. Bn # ‘convert erstion, iteration of opeations of Cyx Maps the ei ‘umeration N toa sngesymactc objet at LF ‘We assume further thatthe tbe of Ki determi derivations ‘sod once and for al kK i formed), rather than being evived rere- Sertatinally at some later stage of he dervaon (se, LF). Thi of ‘oune, not #lopeal neces: Marti coud be diferent. Rather i an eumpion about how hun language work, one that swell wih oo ape te general thesis hat the computational process are stl dive ‘ona guided by output conitions ony in that the properties avaable for computational purposes are those interpreted at the interac. The ‘oper sston inthis cas is whether the sumption (lors withthe ‘ore general perspective is empnaly corel, net whether slg ‘ly neces; ofcourse itis nt ‘Seppose thatthe fate! for (,B) happens o be determized wriguy foraz fin language L, mening tat only one oie yes an admissible ‘invergent derivation, We would then want to deduce that fac from rovers of Bor. i te for in leguogeseerally em [oper ofthe mpage fey. Sim if he labels eniquey ee termined for arbitrary a BL, orotber exes. To the exes that soch Unique determination i pombe. encgones are representable in the ‘more restricted form {2B}. withthe lbs unigely determine. 1 sill Soest below that labels a uiguely dense for categories formed by the operation Move a leaving the question open for Merge. and indicating bets throughout Tor eanity of exposiden, eve i hey are ‘termined. “The Ine Y mt be coatructed fom the to constituents and Suppoe thee are lexi iems, each # st of fees Then the simplest ssumption would be that iter (6 2. theinersection of wand 8 ‘the wrion of ad B © one orth other of. 8 ‘The option (6) and (6) ae inmadionely exch the iment of «Bl gneraly he ieeant to output condtions often al he ion wil te no only trent but contraietry” ia Bieri ‘ae forsome etue,the normal case. We are ef with (Ge the abel y J citer » or fr one othe other projects end is the head of K. APs projec then K = (a [28] Fr expostony convenience, we ean depict consructed objet of ype (Sb) as a tere complex configuration nviving ational ements sch fs node. bars, primes. XP, sutsepts and other indies, and 0 om Thus, we might represent K = fa. {29} informally as (7) (assume co order) where te diagrams 6 constructed from nes paired ith [Gls and pir of sch ateled odes, and intel are distinguished by ssbecrip. Cayo 2d Teaformtione os o AY cn ‘This, however. siformal notation oly: empiri eidence would be required to postulse the additional cements that exter ito (7) beyond lepeal Features, and the exten sets (Ser note 7) 1 know of ro such idence and will heefore keep 16 the minimal ssspion that hres sretore representation is “hae,” exchanganyting beyond tesa fetoes ad objets consirsted from them asin (5) and Ce), with seme minor emendations as we mos toward stil ore pine “The ers component specifier a be fined inthe sual waysin terns ofthe syntactic object K. The headcomplement relation #e the “most loca” relation of an XP to terminal Read, al ete relations ‘whin VP being head specier (aac from adjunction, to whi we ar ict). In principle, there might be a sve af specter 9 pouty wih many consequences to which we return. Te pons of UG, we assume, ruil vee the oa! eatin Forter propstions satis) (6) for the same reasoes. Any such cate ory wil ete Le a a projec of the ead rom which itiately Drees, resting the tem heat tera lets drawn fern Ie lexicon, and aking complemen and specie to be iltions to 2 tea ‘To review notations, we understand tena coment LI wo te an ter selected from the numeration, with 20 pars (her than features} felvant (© Cy A categoey Xs terminal clement with no cates a parts. We rst the tem head Yo termina cements, An X° (rere) entepory iw ead or eategory formed by adjunction tthe ead X, which projects. The head af the projation Kix HOR) HH HK) and K is maximal, then K = HP. We ae so commonly ier tin the maximal zero level projection ofthe head H (3 the head LT wih Vand esas ire dnd). Were hs Ota {consents 2, Bf K have buen formed in the cours of compe favion, one ofthe two must projet—sy, oA the LE otc, ma nal is nerpreed av a phrneof the tye 9 (eg a & nominal phrase i Hs roms: and t behaves in dhe sane manne io the couse of : ; Me capers computation. I tna, then, to take the bel OF K to Be not its fut rather HIK), a decison that ao leads to tec siplifation. ‘Assuming 50, We tke K~ [H(K)- (ao where H(K) i the bead of ‘rand it label aswell, the ces 0 far discuned We wil eep to the assumption tht the head determines the abel, though Dot aly rough sti deniy. “he operation Meets 6) is anymmeti.projetne either a or . the Read ofthe objet tht projets bcoing the label ofthe compen formed. Ifa projets we can rer 1 ta the tage of the operation. boeing the notion from the theory of movernent in the evs wa “There iso sich hing st nonbranekng pojaton. Tn partir there feo way to proj from esc tern 2 subsement HC) consisting ‘fie eatgory and whatever cle ener into Farther computation, HW) being the actu “head and othe leva element ie not ean sh “pata poze” be constricted from larger earens, We this ‘apes with uch struts as (2) with the amu interpretation: she, oak ther o be tena esc tems and D's, N+ staring for what- ‘et properics of ese ems ae relevant to ferier eompattion (= haps th categoria information D. Ne Cae ete). In place of (a we hase ony (2) oe ee [os KI}, which replaces K thin dhe strate E eossining K: Z ise if adjnsion ie tthe oot. Rec hat it the hed tha projets the Ren etre the abel Funder adjuttion. determines tly We this have the otis of a “bare phrase srctue™ theory tha eis fly ste fom nator! minimal pina ‘The Bare theory ‘departs fom conventional sumptions in ever respects: n prt, ‘legos are elementary consiretions fom properties of lec tems, ‘sisting the inclusiveness codon: there aren ar Ives and na ie tncionterwen kx tens and heads” projected rm ther (ae 8) ‘Aconsequenc isha a item an be both am X° and an XP. Bows this ‘hus poblons? Are there enampts that let this posit? Ys to particu problems and one ase comes to mind 2s psibe es ustin: cis. Under the DP hyped ets are Dk Amumefurer thar clerks from se penton and attaches to inflectional ead In its postion. the lies am XPs tach oa ead eqs at it be an X" (on fail standard assumptions). Furthermore, he ‘moverent ines the Head Movement Consiraint (HMC), ficting aptn that it sun XP, rang by XP-aduncion wl the al step of ‘i-adjunetion. Ces appear to share XP and X° propectiy, a¢ we ‘youll expect on mina arumptions 1 he reasoning seiched 50 fri core, phese structure Meory fs essen “piven” on grounds of vital concep nee in the ‘ers indicat The sucess sn eal seroos are her ising or eeemalated in elementsy terms sting mile ‘ondtions with no objec beyond lexical Ieatures, Stipulated conet- tions are derived. Substation ard adjunction are sraghforvaré. At least one goal ef the Minimalist Program sms to be within reac: Dre stricture tory canbe eiminated entirely seem. o te base ‘ofthe most detemtary assumptions so ates his aspect of human language is “pete (bats note 2), 44 The Operation Mone 44.1 Movement and Economy ‘The sete (11) will el the Sentence he mar une if when Fate ‘nina lements are added by Merge and the specifics of the VP 0 ager raise (suming this fom ofthe pediate internal sobs hypothesis) ‘The consiction so formed inves the Sond operation tht forme categories Mone (Move 2) What is ths eperaton? We bave so fer ‘ssumed hat it work ike ti. Suppose we have th cataeny E ith teams K and a. Then we may form by rating oo treet K. That operation replaces K in E by fy fa, K}. In the optimal they, acting ee wil change io sd y wl be prectabe, We take hunt farguge tobe optima in she Cormer sense: thre are no adionat mevhanisms to accormoda fr- ther changes in © As for pica of, we hope toes the ‘Mandar convertion thar the wget projets (within the class of comer ‘ent derivations. so tht y 8 HK) or CH(K). HKD, depending on ‘wheter the operation is sbetittion or atin. The question dose ot are For Merge, bul i door for Mone we wl cr 1 his ater “The onl cher aperation for the moment x Delete (Delete 3, which we have ested to lenve the stare enacted apart from an fdiion ‘har ae aot vibe" a the nerfae “The operation Mose forms the chain CH = (a 1(a)sa) the tice of a Assune Forter that CH ices several other contons (© ‘Command, Last Ret, and ites), co be speled out more caf 5 we proced Ie forming th Perv ren 11), te subjects ised 1 she oot of ‘the category E.tarpetng te proatin of Land becoming [Spec 1. Bt raising ofthe objet args an embed ifectinalctegry K tht ‘4 proper rbsracture of © We Rave ken this ro Be cover raking of the obit to [Spe Agr for Case and apremen. Pic this oper ‘ion we have Gin informal poration) the stra (123) ened in he larger srctre (12. Oe Ags am OW vO (Cocgorarand Trfomations a Hise Tis (7... where K (aamey, (12) i (Apr, (AerP)) YP = [V.[V. DP})> we tang! K, merging DP ad K and protng| ‘Agr a intended, we frm (13, ith the raised DP the apeiier of AgrP (ager (3) Age pe Res Here Agi is (Agr. |DP_K}] = L, ane th term TT inmistitly domi ‘ating i (7. (T, 13} not (T, {TK} ah wae before Move rine DP. {Under the copy theory of movenent section 3.8). woleent ean 8 psir BD. where a= Bi Since we disingish among dst sel ‘ons of a single ite from the lesion, we can Bese that sch pats tse only though movement. Suppose. for example, hat we have con ‘rcted the objet (16), Pa head, and we deine (15 rom it by sisi reeting Let KL be the objec represented informally in (12. (1) respective ‘Thus, K = (8 (8. 2}) and L = (8, (0, K)}. We ae now imerested in 0 ofthe terms of Leal them = ae, where, fhe frm of Lich hat = {R, f-K)) ane ris term OL uk that K = (Ba) beret fa. We wish 1 cowsract the ein CH tha wi sve es he LF object formed from these Wo terms of L, wich we cll xs he race of, respectively. The operation tht ris niroduces ta second time int te syracti objet that forma by the operatic the nly "ss in which wo terms canbe ienical But we wan to deingsh the 00 elemenis ofthe chain CH formed by this operon The natural en) 10 doso iby inspection ofthe etext in which he er appears. Given 22 coopers the way syntactic obsstsare formed sufces to consider the co-om- stunt (er af a tem, always dine! foro ae is trace. SAP ‘hen that oases to sept Mn Eso thatthe eit of the operation i formed by replacing M in by {Nac M}), Nhe lb. The element ‘room appears twice in Es is nil postion and in he eased pon Son. We ean deny the iit pein of =a the pai (a) (Bi the ‘content of ain 3), and the ruse potion a hep 2, K)( the eocorstitent ofthe re term ain), Actually, and K would sufi; the pair simply more perspicuous. Though and is ace re ‘ema, thet postions aye disnes. We ean take the chain CH that isthe objec imerpreed at LF tobe the pai f positions. Tn (14 and (15) the postion POS, of, is Ga. K) and the positon POS, of fs (o,f. POS, and POS, are divin objets cootating the ein {CH= (POS. POS, formed by the operation: the chain é steal (KB if we adopt the more suse verin, We fer to CH informally 38 il). omit a mere presi account, the poi! Bing lar enoush {or our puryeses. “The command relations are determined by the mane of constr: ion of Chai re ararbiguusty determined i hs wey "may, however, be coret to allow a cetin ambiguity. Ral the curber dscusion of elipl a a special ease of “opy intonation,” the ‘special itonation found in the bracketed phrase of (a) (= (324 ef ape Tse sections 15 and 3.) (16) John id hat he was looking fo the wt ooking for ath Jobe i that ews looking for aca and 0 i ill ere (16 is ern fea (16a) by detion of the ached has in ‘he phonclopal component. A some pot i the deat, ce rack cd eement mast be masked as “subject to parallelism interpeasion, ‘Assume that this takes place before SpeltOu.” The marking coud be removal of the distinctions indicated by nueraton, in whic ease He bracketed element i in a cen sense nondstine om the pase it opis” (he ter stil marked by the numeration). Such a cenfigr= tion might be iterpreted at PF as assigning copy intonation 10 the bracketed exresson, ad at LF as imposing te pra inrpret ns (a complex and intguing mater whih has only been very pt tials investigated). Suppose that umeration markings om the copy ate hanged to those of he Gt conjunc tad of eg deleted. Then the and 2.8 Bil fay ta Cures and Tranfomations 2 antecedent and its copy ave scl idemeal and constitute a cai, i ‘hai understood as (constr from) pt of tens (2,04) that 2c iertica in constztion.K wl foo, ten that the cory eee, by ‘whatever mechanism dees traces inthe phonological component. At LF the two Kings of consructons wl be very imi though nt uit “dena, Iwi ten be naesary to demorstate that lesitnate LP ‘objets. in she sense of ear chapier, can be smiguly Mee with hs that conta arguments) property dsinpuished fom hose {soled in pram structures ‘Without pursing inca Hee, there are srong sons to supose ‘thatthe sic eopy (th PF deletion generally tovoles the se kind ‘finterpreaion athe inerfaceas the nondistint copy (ith copy ito ation) but somewhat stergthened, an that the ar fle under fa ‘more general conditions tht hold fora wie range of ether sons ‘ons wel and that go far Beyond sentence gra or dicoure ‘Similar ides mish scxomiodate the notion of lied ches Gin the sense of setton 1.43) and chins forma by suecensiveoylc movetent ‘We wil tum to these questions, ‘A ein CH (a. 10) fommed by Mone eoets several conditions hich we take tobe pat ofthe dfntion of the epeation il One of {heii the C-Command Contin: mst comand tac, 50 thet there cannot be an operation that lowers or mover "udewayss overeat is asng. i he spe eos dened by ccomiand. A eo od requirement, which eas aural, ethene ot the uniformity ‘condition (17)? where the phrase autre stn of wn cles is “teat property of teng mais, mitina,o thes. (07, chains uiform with rgd to pase strate slates A third requirement i hat Mowe most mest the Last Resort cotion ‘on moverert, which expresses the iden that Mose is driven by fetare ‘cing, = morphological property, We wil cue to the proper inter Pretaton of Fast Resor ad to empiric eonsnqunce ofthe coment ‘hate take to deine Move. Note that i deletion forms has, s su ‘tl cave. thee ray not meet any ef the conditions tha ld ofthe Operon Move 's meaningless o ask whether the confons that conte the {sirtion of Move canbe “overridden for emergent, Oto wk how ‘sconomy consferatons apply to them. Tha stra whatever he proper seations tum aut tebe. However formule, thes cenaons oe pat ™ haptr ‘fie dinition ofthe algorithm Gye. Vitis them would be on & par wih making an ilegtimate move In game Of ces or aging & bine itegiimetly to proo. In such cases further Questions abou the obec eine const (eomverpecs economy, shortest game ex prot te) not arieIthe proper conditions ate CCompd, wnformty and ast Reso then there 6 no metrngfl question about the ees of ‘ating tee coring whatever ers may be inrodeed). “The computational stem Cy is based on two operations, Merge and ‘Move. We have assued farther that Merge always appli in tbe sn lest possible form: atthe root What about More? The simplest case "gun appliction a the oot ifthe devin hss reached the tage ‘hen Move sles and ages forming, {2}. But covet move ‘ner ically ered # and therfore tats more comple foe given selec K within ¥ and ase to tarpetK. forming (. (0K which Sbettutes for Kio © The more comple operation is somes st ced by economy considerations amy. Procrsiat, which requ thor some operations be cover, hence (Fey) operations that embet. Furthenore, even over Xauntion of to Bs within te category [Po hunded by Bz nee not sry speaking atthe oot even P™ the roa category For overt movement, agting of an embeded exegory is obvi sot ered by Proctstinate, or by ature seen (ce (3), One mere iment therefore alvays targets the root, wih the minor guaiestion noted in the case of head adjunction. and is invari eye woul be itresting vo svengten his contusion: thai. 10 show that overt targeting ofan embedded category (ence lowering and en tye raising) i ot posubl, hen fron nt necessary. Argues To ths ete: have been proposed (Kawashima and Kitshars 994, Esch Gear, personel communication, based on two assumption: (0) tat eummand plays # ears ele determining Brew order accord with Kayee's theory of ordering. 10 which we will return in section 48: @) thatthe only flaions that exist Tor Cy ate thse establiied by the serivtonal process ie (Epstein 199 Specialy, in Epstein the fry, eczmmand is jst the relation thet olds between ad lens of B when 2 attached to B by Merge or Move Is then i i uchod o an eed catepoy by an operation, wil eter into no ‘Seommand relation with any “higher” element 30 to ores es fished barwcen cand and the dereation rashes at PFU follows tha Cngoris aed Trfomatons 2s ‘overt operons are never of the mon complex tye tha noes ane tee extepory, hence must be jee a mit be ising (oa lowering) ‘pecans. Bat covert operations might ot meet these conditions. the ordering requirement big elvan foe LF convergence ‘There area miner of questions abou checking theory th rein open. which we wll hve 10 resolve a8 we proce. One has to do with thecking in Srveture (1), where F is funcional category wih a joie. Ww oF \ ve oy ‘Suppose Fis Age and as Tor V, and we ae interests in Cat check ing DP isin the checking domain f both 2 and F. is Cae coud therfore be checked either by features of or hy features of Fis ‘eatenabl to suppese that hs the Cnscasining feature a an inn ‘propery ihr Usted init lnc entry or determined By the ex would sfc then to conser only the checkin om of 1 de ermine whether the Case of DP has bon propery eld (asestive ia V. nominative i aie TA move complex alemaive is that F ‘and «nes har all tues eleva to checking and that DP ‘heckedajainst F. That would mean that i the numeration, has the ‘Case-assgnns property, which mates that of As wus wil ede the simpler assrption in the absence of eden othe contary. We else ltr tha the consi sls erry wel mia ko 8 usta an intersing fet) ‘Wher the sme concision halds for Fare is» matter of fc ‘ot decision. depending the aniwers 10 4 ution raed in ston 422: how are bfeauesasiciatad with verbs and ajetivs? Fave ‘eoatvely asuned thatthe assignment & option. nthe ransiion from teseon to numeration. 1 so, then Case and ¢festare fection aks they ae not properties of the funcional etepoey Ags (of (18), and ther is no mate reltion betwen F an Iwai ontnds to Sue the, thatthe features of T and V that check Cae ot ears of DP appear only im thee categories, ot in Ap, which leks such Features, The dacsion co itl revere ut seton #10 442 Projection of Target, ‘Consider spsin (UI), embedded in higher incon eategores, ain 09) = (Apr TD, 09) © ‘We usume tht the DP the man hen ies overtly, targeting Pw rae covery argting AfroP. etch of the raised elements Becoming the specier ofthe arsed category. As noted, theres another option the phase tha raised might sl” have projected. 1F ii anteraly ‘comple, hen it Bvores in X ith the Age opie ian them it besnmes the head ofthe new projection, with the AP is come plement. This. i the raised DP projets the targeed IP becomes the Specifier ofthe D ben of the nan, which now # D's a the areted ‘Acro brcomes the complement of i, which Reads new DP. Ate ai ing and pojsting, neither the mon nor i is 2 DP, 3 they were before the operation took ple. Ih pre-Miainalist Progra work, these abv ‘nly unwanted onions were excladed by conditions oe uraformatons si slates properties of Xr teary. Bat weno longer eat © Wish lo, make eno to thee, 0 we hope to sow thatthe conventional lmwinptions are in fac deriable om pscipled grounds that it Impossible for Move to rate o targeting K, then projecting rather ‘tan K [Note that these questions ate oly for Mose, nat Mets. for which the concuson i tue by defition Let us bein with button, cra into ajuntion later. Recall the guiding assumption: movement of targeting K is pert tel only ifthe epeaton is morphological driven by the nad 0 check some fester (Lat Resor). This ides can be Formate n numberof says, Consider he seterpetatons of Las Retert, adapted from the Cts nd Troon: 2 (20) acan target K ely if 43 ature of is hacked by the operation a ature of ether or K i checked y the operation {the operation ia necessary step toward see ater operation io Which feature of il be cached >> “There ae vavous andar here, which wl be esolved as we proces Recall hat Last Resor, however ti nly trp i oe nde Hood as part ofthe definiion of the operation Move thats a an attempt 10 capture preisely the inate Wea that movernent en ty morphological checking reiienen Suppose that res fo target K, forming L= (H(2. (0 KJ) La projection of with Ibe He) = Bead of Since te operation is sib lution, we have oe ces 0 conser. (21) a. ais the head of Land K its complement '. Kis thespeifer of Ha, Suppore case (21a). The operation Is ot permed ander versions (204) oF 205 of Last Reso, No property Pca be checked in the heat ‘complement structure that hes ten formed. The pettion mighe be allowed ony under interpretation 28) rang of 21 pert so as to llow ito reac some pesition fom which can then rie Torter to tger K where Pill esate. Hower any postion ace {un the newt formed peston of ead of K] would ao have been sso from the postion fis trace: the HMI cannot be overcome ‘nis way. aor can any oer condo, it wernt. Ths the st should ot ase. Since T wl Iter propose a version of Last Resort tha ex Projection of i afer it raises to (Sp, Agr] therefore bate in (05) wong with many ote eases thu i mpl to ruae Vt ting KV ptojting to orn yy V Kl with ead Vand complement K. “The only possi. then is ease (2b) afte sing K = (Spee HC) ‘2 therefore nontrivial 2° (otherwise, K would be i complet in {he chain CH = (formed by the opeetion, (iam N= Bk Does afer rasing, sot san X category. This new X ategoycan- ‘ol if be moved rte being invisible othe computations system ths, interpetation 0) 24 rectly elevnt and in ny event, Se may put the matter aside. antigpaing elimination of ths option Keeping to (ts) and OD, he question is wth he option forming ae cranes LH), [with K = [Spe, Hla] and beading the chain CH, Isteptimate. We hope to show that itis no ‘There are two lines of aqpument that could bar this operation. One approche to question the legitunacy ofthe chan CH (C0) that is Fone by the operation, In fat, CH vines the uniformity condition (07), whieh rogues dat end ra) have the same phrase strate satus sine Ha) cman and 28 not “Assuming this condition, then. we conclude hat nonmanna a cannot ‘aise by altton, arpeing Kx whichover demon project IK pro Secs then is maximal and ie tase i no, volatng the uniforsty dion. I projets, the uniformity condion i suisied but K i the complement of andthe operation feed bdr esse 2a). The ‘nant’ interpretations of (19) are tho riled out. by ease (21a) for DP D and by ease (218) for BP 2 D. Sima, the D Bead of ron ‘rviat DP cannot rats, targeting XP (x, am Ags? leaving the residue ofthe DP behind, whether D ofthe target projects: andthe V head of Sontrinal VP cannot raise tarptng K. whether V or K projets. "A diferent approach to esse 2) considers not thee CH but the structure that fora by postion ofthe ried lent o-Tn 1 he target K = [Spec (2 inthe checking drain of Hin, ut ads comnts are nti the checking domain of H(K) (he head of K) Returning 1 (19s the arises. ageing IP. are the raised DP prox jes then the mens of IP are inthe checking domain of the D head ‘the man, bel the man notin the checking domain ofthe bead of WP. ‘We might ash whether «proper checking ration i extabsed in this “To amuer thie quation, we have to role an ambiguity about heckng that has not ban sorted a caf. The nave eas tha operations involving Case and agreement are asymmetric. The tad tinal inition hat verb ais Case ois et, nt conver ‘That asyimetry i cared over only in part to some of the err pprvachs bed on government: 2 tasitve verb ass Case 0 the DP ic povers, and the hed agent With is specif for checking of & fextresbut nominative Case apna inthe Spec head ration With | snform ntrpeation of Cas Spec had reaton, ang together sith agreement the aymmetey tui is agin express, but not ‘tured, Theat scription tha the Vor T head checks the Case ‘ofthe DP in Spex st thatthe DP checks the head and the features Csgei and Transformations » ofthe head are determined by thse ofthe DP in Spe—the verb seers ith the sujet, not the subject with the verb. Simian thas bose andar to speak ofthe wh phrase raised to [Sp C] we bing cemasd bya Qentare ofthe complementarC, not ofthe ler tens enued ty the rated whphrase. The inttve bas or the distncons i ly ‘dear. Case an ncinse property of verbo ¥ element, not of the DP that “reeves Casein a certain poston: and fetes ae properties ofthe DP, not of the verb oF adjective that acunes then fates in 8 Speotead configuration. The question we ow fae i whether these Tncitions eetualy play 4 role im the computational proces Cig oF whether, ike others of ancient vintage (grammatical eoastaction, they isle ino taxon arias. IF the Specie elton relly seymmeti inthe tnaonersuppoced In informal description, hn the Tats hat Ki in the checking domain of (ain he constuciowL formed by rusng the moved eement would ot establish a checking relation, ice it x HC) that must check the Fenures of whi H(@) camot check the features of KI the intuition 1 not relevant to Cy than a checking edation ic estab and the eget coestacton Lis at barred on these rounds, ‘These wo approicies are not logically equal, but they ovelap for standard cases, forcing the tags to projet. The redundancy we {ets ha at eat one inner. The appronch iter o ulformity fas advantages i extends other caer andi concep much in ‘es. and we donot fave to introduce a nti of esyretry expresed iatey in terme of turin properties of heads —and aot Mn any simple way. as wll come even clearer Eeow There iso fa 1 ‘now, no reason to suppose thatthe property inti) pays any role In Cy For hese reason, wil sae the efor approach Sommariing. these are good reasons why te fart, ot he rd tfement shoul eect under substi 1 special css thee ate other argues that fad tothe same eor- ‘sion, Keeping to sutton, soppone that the tpt K is consti ofthe caepony N= (HK), {Ks M]} projected om K. Note that K {an X° category: oterwis, x an Xestegory, no 8 vb tage. ‘Thus. Kis ther H(A) or an X* proton of HK) formed by dicing sfements to H(K), with M its comment. Suppose that rises, targeting K and projecting to form L. (HC) (0. K)}. Low replces the teen Kio Ne forming, N= 0 Capers HO), (LM) The end oF i HAR), he Bea of Lis Mio) and the nud of M is HOM) al isnt the head of Nw dint eo the head of either ots consents Ni ot ite sytacte object of the sent permite bythe recursive procure () The derivation therfore ctl “The anpiment extends eiety to ajunaion of 210 K, so tha the Intel of Ee CHa), Ha). He sil the ease that replacement of the tees K in Ny Lik a hepinat sytatc objet. This ese of rut interest ithe Kp af the eae of X"adjuetion ip compu Hon, which becomes even moe ceil a we proved, “To surmnarze, we have the foling answers tthe question about projection afer Mose. Suppose tha 2. K are eaegois ie and ase to target K. If the operation i subs, K must project for ‘convergence. whher K is embeded in E (over raising) or K = E (over rang). Suppose the operation adjunction 10 HY within N = [HH M1) [yr H® M), with tend Hand complement M. Asin the target X° mus project when adjoins to HE. Furthermore, adjunction of ronmasinal 216 XP (inctoding root) i barred by the uniformity ‘snditon “The only cass sil ot covered are adjunction of @ 10 K ia an ad jnetionsrucrore N = IK, Mk. where M i oie to nonminimal K oF IM projets In such eter siter 2, K, and M ae all nmi! (XPS ‘or they ae ail nonmasiml (Xs); we wil tara to deta. We wil soe that ¥P-adjnction to XP hos dibs sits, snd ths marginal ese ‘ven moreso. We therefore ignore i, esrtingatenton fo N= [KM] Sth K, ang Mall X°% aed Mprojating to form L= [CHOM), HOk)),(K. M]). This case sole plusie morphological conditions. K "ch. The correation Follows he ebsenes, ithe resoning one here a apie Jost how broadly considerations of PF consergence might extend is cer, ponding beter enersanding of morptoleg) and the itera tHructre of pasa. Note hat sch comierations cou pert raising ‘thou pl piping even overtly, depending on morphologic siucue, sm the theory of overt rating of en) opuatrs in Japanese devel ‘oped by Watanabe (1992). Piping min principe depend x wel on factors that constsin overeat: ores, Empty Category Pnciple (ECP) consideration, the Minimal Link Condon (MLC) that requires “shortest moves.” oF ‘thatewer rns ou to beth ight try for this mest Bt il ‘murky area I the ese of all such principles, one open ston as ‘en wheter vcltion causa derivation to rash or allows itt con- ‘serge ws deviant (ya Subjcency Volaon van ECP violation). The ‘gestion aul Have an answer the terms now being considered. Ths. 1 pci Foren by he need Sats some prncneP. we eon ‘line that welation of Peas the derivation to era 0 tat it des not bar fo exomial derivations without pied-piping for example, the Principle P tat sometines bars prepesion straoding- ‘Any fer laoraton woul be # deprire fom minimalist os sumptions, Bee tobe adopted ony infer asta is forced on erp fat grounds neve, nthe Best east, A hos of probes arse tht look ‘aout The basic tsk ito determine how mach of«depstre (any) {sreuied fro these opti assumplion to account fr “generat ‘siping how PF and LF consdcatons enter into the ite: what these considevations imply about the setae of phrases apd the sts fn nstre of conditions on moverent an! how ngutge vation Setenrined ‘Arnot by Hi Kithars and Howaed Lasik the preps econ- my principle provides efter atonale fee the principle Procras- ute thing tle he hast hat an e cared slong for converge. i hac psste only if aking cover, no entering the poneles ‘al component ‘Consider now the case of covert movement. Questions of PF comer ence do ots, 0 eneraized piping cut only be required by ‘ondtions cs movement, Err Gcussion of Move assumed tha the princi tha govern the operation hl only or eaepois, nine oy ‘neperies mere assumed Yo move If tha happens tebe ts hen these ‘incipes old only of ort movement, which hast care lon whole [Seperis for PF convergence. The conesion could wall be tue fee Cuegorie sd Tranforatons 2s ‘ther reasons cven if tbe assumption is fle. AF the conan i rue (Gor whatever reason), then cover rising i este to entre ring. ‘The operation Move F eas along “es hagge” only whe iti “heard in the phonetic eutput. wil sssume that to be the case, The assumption accords welt th the general minal perspective. aa it ‘hes no obvious empire fw, ‘We senitively assume then, that only PF convergence forces) thing beyond feaeres to as. If that rns out oe theese, orto the extn thar it does, we have Gurher acon to suspect hat npuae*im- pecections” aise om the external regimen tat the cmputationl ‘nce must adap othe Sesorimetor appara, which win cor tain serie “extrancous” fo the core systems of langage 8 reveled in the N-» 2.computatin, ‘When the feature Fo theese item 1 aes wihow pist-piping of 1 or any Tnrgeeexegory 2,8 alvays i cover ening, does tery ‘aise alone or does automatically ake other forma fetares lone ith 1? Tere are strong erpircal easons for asuning that Move Faio- Imatcally eavis slong FF(LD, the set of form festres of Li, We therefore understand the opsraton Mose Fin acird with (2). where FFF(F]is FFOLD. Fa featre of the exit Lh. (28) Move F “cates slo” FFIFL “This much pie pin automat, rlting the fat that Move eats to checking of formal feures. Broader piping a tequird for convergence —"exteancou” insur PF convergence i the ding facior, which we tematvysseme to mca “always.” Applica to the feature F. the operation Mone ths creates at est one and perhaps two “derwative hein longi the chain CHy = (Fo) ‘constructed by the operation ssl One is CH, = (FFF om ‘ising of the se of formal features FFIF] and is ce; the other (Hear = (5). eatery eared along by sncaied ied ping Sand inca ut lest the exe item Lt eonsining F. CHyy asap ‘onstructed, CHexy ely when required for convergence. The computa: ‘ional psc Cy ely “looking at” CH, but eu ofthe corer ois ‘cl can “see he other tg a wl Each cates operations. Tos, (CHhogy determines the PF euiet, aed CH enters int cocking oper ‘ons in manner to which we wl retire A rot CH shoul be completely dispense, wee i ot forthe need to econ to the sensorimotor appara pan ‘The empirical questions that aris ae varied and comple, and it easy enough to come up ith apparent esuntereidence pu thes problem side fr now, ply mstmang the Best oot mel) th UG sites he materbardly an teow sep, mens 10 Sy. 1 ‘sume, then, thatthe operation Move rases F a derivative raises FFU # wal, corning along + phrase conasing F only when the ‘movement ove, reuie for eonvergence. The peel apprcach — ‘ratural fp ial ebistory on mamas grounds, and i onfers ‘purer of advantages, as ee wil "Note hat we continue to lyon the assumption that ool conserpent “derivations sre compared fr econeny that heeds erations Dy area subet ofthe converse ones De. Ths, ising without pe ‘Spine is more “econo In sone natal seme, tht ferent ‘tthe devon does at converge ‘Weave already comsidered a speci cose that resembles the eonemy riven (265 namely such operons ab wh movement. As disused In section 3.5, the emite wiprase need not raise covey for fexure checking sed scope determination, and peshaps does no ths, we ound fev etson to belee that nothing more than how many rast every From within the phase howe mary pictures of Join. A natural exenson oft aati that only the whfeatore raises inthe cvet operation {he reat ofthe nase enaiingin st “The revision of Move to Move F exter is casoning 0 al eases. 1 ate perms a way wo capture the esenoe of Last React (1 be «© ‘vader propery ofthe operation Mave F (29) Fis uochesked and enters it a checking elation, “Thus the vasa F in Move F anges over wehecke fntre, and she rel ofthe operation i hat emer into a checking eatin, ether checking 3 fentre of the ara r tng checked itl "We are now tentatively assuring hat al fears of some category suave ben checked, then as naccesible o movement, whether it 8| ead or some prjetion. But if ome festore F i as yt unchecked, «i fee to move Feonomy conditions eacide “entra” moves and anything ‘more than the mini id-iring requrd for converses. I cover movement, fetes rae alone Precasinate express the preference {er the covert epi. “This simple ad earl reinterpretation of Move a. akeady motivate for wihmoverent, allows us 16 ehrinat the completes oF interpre Cvegsin and Teformtons x “ig 0 Lat oo ei: once We see tit 20) as we the ase eae Fs ener sto 8 checking en chor sy gone fr wee Ks wages fee thn cherry moon ts wens cont |General dgtoneay tae ne We || sitcom trie nyrerens iy enor s oet een trices en eee coos oe ey. |__-be te pert mvs ar hoes — ee _ change. We therefore retain the conclusion thatthe target projects. Such ees 85 (2)-(24) abo fal fate place: though the torst i Feptimate, “faving an uncheck fete he extegoy to be cae rule to | _evement. having po unchecked festees (an analysis 10 fe reve dow), Ralsing under ECM is permed isome esa cock the fone festure oF the embedded I hat eds the EPP. inthis ee The _ inl of Grecd snes dispensable, excep isa it ncorporatad hie 2) Consider succnsvecylie whimenetet. Us allowe une his ap "proach oly where there ea morphological refer Semmes thi i Ws | Heat PF. asin trish and Eve (sce Collies 1993) serine am open “sion whebee Such vsily sony an accident of rong, Fe | eslng the workings fs procs that more general perhaps ves Sten ifmeipclogiet refers arent detested the PF output __Athnetion to nenmiinal XP is now barred wns some featere is _Werby checked (ee Oka 199, for development f ths pou see _ tzuive-qcc adjunction t xen more probemati. The condition could ‘ompes See Cols 1946 for fhe discussion; we wil ue og os abet XP-ajuncion ‘Asvong the mates sil be cai is the sats of the MLC. The efrre conclusion would be that the MLC is pat ofthe dition of ‘Move: Move F mest obsere this condition, making th “honest ve" ‘Bemisile Hrhat can te estas it wil sharply ade the comp _ tonal compen of dtrinng wither a patel operation OP _ the course ofa devation Is episale. tm contrat if the MLC i a Sonomy condition sting among derivations, OP wil be geriibe Ral ifn ether convergent derivation fs shot His aed tose 2 taper ven howto formulate ch «condition, kt alone apply iin some com Dimon esi way fer example, how do we compare derivations (i sorter ks inference? But the @estion dies not ae iF ‘olaton ofthe MLC sn eitite move in the fst place. Follow. {ng the ust minimal tun, fet ws assure that lation of the -MILC ican its move, exploring the se a8 we posed. ‘Soppoie that Frases, carrying along the vest of a EAEROY t= cng K. By version (29) of Lat Resort the operation speed oly iF i satis a checking ration, We therefore Have to have an denen: tary way to determine te fetes of 9 apd K that enter iat this ceck ington, no matter now deeply embedded these ae in @ and K. For the raed skment the question docs not ais. It isthe feature Fut that rust emer no the checking relation, by (29): ote eats ‘Of FFF aay ako eter imo checking ations as "eres." cred son inthe deiative chain CHyy = (FF, thats automaticly on fart, Bot that ey deta, gen F. I checking relation | ‘tale hy merger i the checking domain of, then the leat atures in he new checking domain are those ofthe head of, whic inv immediately termed bys label 2” Questions ars, then, oy with regard tothe eategory K that the expt of movent ining 2 checking detain (either an adjunct or «spe by rue ef te {La MI} te eget of moverent. The uf ture Fy of Kray eter nt checking relations 38 within the zro- level projection "ofthe ead H of K. Hand Hare contracted ‘eval from the Ibe! y, whi is immediatly determined by inspection (TK. By wl bea feature ther of H ie or of sme ment adjoin! fo H, and so on; thie wil be seid even further in seton 4.10 Recal that we ate keeping tothe optimal assumption that not ay 1 bt eo fetus adjoin to ican enter int achcking ration wih tthe cocking domain (end of section 4.4.1) For the wget, ten determination of he rekvant features i 0 i ink thew ae the fetes asocated with the lb, which we may el blael (60) A sublael of Kina fests of (KIM “That ist a feature ofthe zero projection of he bead 1K) of K: Wen Move F mises Fo tart K, sone sublbel of K oust ein ‘the operon by entering nto cece relation with F and etoes of er: ‘The feaues thal legitimize the operation rising to target K are ‘erefore determined sriptforardly, however deeply embed they sy bein and K: for example, tbe wis eaure im plas of whose wae did ou think were onthe mantlpeee. The computation “oaks | at only F anda sublet of K, though it aes” more “The centr procedure for etrsning the leant features ofthe rad ements “te veal cement adjined tt asuming al oer functional eat "pies tobe iret at this pot (ve wl eto otis upton) mw o§ « “Are cpeon Move F toring (I) nis the cate etre of he ‘sb V, caring FF] slows suomacaly ina ernie cui Whe | syrtion in over at in Engh thn noting upper in 1) | ifthe eget opt rechten 2 Vt piped ‘low comernence oP a ele it yma cot in a cecing | son ih te fal ete 1 (tke eb a) of Td ay | ete eae of Flan in principe ener noe eeking ron with fet fF (= Tere ang. ty espn. The sob of “Pao fore. ae ne eres ano Silay. when de Cae fear of Ls ated by Move F 9 ee | he Gears of and any of thse Sa ders aya te io 3 eras) checking ato wth vba fhe art. For eam | ing of DP for Case checking carves slong geste, which may __Pepgen to check agement stares fhe tre We wl re Surconeiences ___ Betis oer eae the flog theory ofthe o- | ron Move. Move rss tare Fort Kin ony 9 oly Mth 38) nd 36 as automatic cnsegoens, an) further om __ sven (ssmed, bt rly eased. £2 «Fan weteted eure 1 Peer in a decking lation ih eae of K a el ‘be petion caper (03) 5. FUP] ise long wih 1. Acategory t connning F moves alongwith Fon as regia For convergence. Cover operations are pare ei rising Other features of FF] may check sublabel of K as tse ana 32) incorporate Last Resort. {tu frm ow to vera us that come tthe foe when we for smote movenent theory as Move F. “The recurve Sep i the defiition (5) of adible objets permite ‘te construction of L = fy 2B} }- where Bate syntactic objects and ‘isthe bal of L.Inearierdscssion we kept 1 the eee where 3. are Tex tensor lrg phrases corstrted fom them, bt we hate om boo conning e more generals, ith the rable ows 1 range ove etures a well Specie. we allow an objet = (7. (F.K}).F 2 fate, formed by sing F to target K without pid-iping eae ory 2 Several questions ars, hiding the allowing: ers (20) (64) 8, Can the operation be substitution? 1. Mast the target projec? Can K beat feu her than a category? “The answers depend on how we interpre sch rations a8 "XO" and “pea” which 30 far have bre defined only for phrases consuced from lsc ems, no for features Rut these sions have 20 eer Sse for features, Suppos, then, thal the dentine sien sarereay ver thou change. Iso, he question of (3) ae std ‘Suppose the feature is ned aging Kae fering (1 (F:K)) ‘The anser o question (4a is negative. Feanot beeome a compe ‘nent for reason already dacomet. Hist thereto be a sper of K. lence an X"" by defniton. The statement i mesg athe notions Xe fue) ae ot dead for Xo Fear. Wf they ware defied, then F woud be ne kindof "xd the chain forma would woe he seformty condition (17), under eny satura interpretation. In sthet ‘se, then, the operation nut be adnetion of F 40 K. Move F eat te substation oly in overt movement. wih eatepory pied piped for comvernece, A For question (34). the target must pret: exo be or be cot surat fom) the bead of Fi he neton "head of Fis not defied ‘Qurstion (2) is ako anewerat Keannot bea Feature itis the obs ‘Somircted wil be no abel | exer Trains m Or pause assumptions, she cs of penis obit i extended ‘only very sigh by extending (5) 25), permiting the variables 2B to range over fenures in the renurave sep of the characterization of static objets K = [. {0B}. In ft. the only sew cts lowed |e those formed by cover adjuncion of fenurs to & bead which amounts to saying that a formal property fa lex en ean covert _ enter the checking domain of estegor, the question of PF convergence feng iar Furterore wea the aay jo thiscase as dese. From a more fundamenal pont of view, the cas of prise te {esis radically Konted by these revisions. ‘The only “rel” syntactic objects are exe dens and L~ (y.F. K} Fa feature. Ka peeing fateory nd coset rom HK). Ths view captures ater csely the concept of mevenent (taraormations) toward which work in gore ative grammar has en converging fr many years “Tax resort” oper: tions driven by moepolgiea equerens, whch vary win « naeow Fangs ying cra pops iferences. Other ebjessue oem ‘only a6 required for corvrgence—perhaps only PF comergence, hus Ining spun the “ewearcous” chracter of the lnk to semoritoter systems Suppose thatthe rgctK i nominal A reasonable conjecture is ‘ar the objet formed. wih a eatareadoined to a pure (Donita ‘animal projection, would ke uiterreable at LF independent. we il sc that there are ena rors 1 mappose tha an eemeot a Jered 1 nonmininal K foot ite etucing domain of head H(K). So hat the operation would be barra by Las Resor. Assuring tis we force hat pce Feature raising hence al covert rising is sane ton ofa feaure 1o ahead, which projes The only new bjets L= 4. [F.K)} alowed are those consruted by ajning the future F to ‘he ead K, which pects so that ries he “ype” of K. We have already found that in the ase of eateory movement, the tare always projets. The eaclsion s nom general covering cas ‘sfmovement, with questions remaiting only for YPadunction vo XP. ‘The picure is very simple and stighforward, and the agents Seto on asumpions thar sem concep natural nnd in accord it the Minimalist Prope. Its closet acim, then hus inp: spiny lose o “pee,” i the sone ert, Wher fe Sonchsons are enpisealy crs ater queen, tay = am Capers 445 Covert Rang ‘The sift of perpectve jan util hat broader consauuences. te ‘cas of wh movement, the person face wb unetcked, ies to an appropriate postion, covery sf posshle (by Procrasiate) and ‘thos withoot ied piping. I rain toe hen piping wil be de termined (we ope) by PF comsersnce and morphol propetis af the anne, Sma, the gamma objet Obj raises Tor checking (of Castor seme othr formal etre then the FestresFF(F] ofits head ‘aise deat snd the operation caries along a fl extegory only if the moverent overt I rakngis over. thea Obj bazomes Spe, Act Wiis coven, then the features FEF] rab nln, adning (© ABs, which has W (ors leven features) alieady ajo ti." “The same should bold for raising ef subject Sj. Is unchecked features re eligi fo raising. The operation fs subwitlion wal pie ping if overt 4, to ey the EPP), and i sanction tthe ap opiate head without pi ping if covert (pera in VSO lanes). ‘obj and Ob should furction sn mich the same way at LF whether they ave rused overly afl categories or covertly 3 fests. In ther cate there slement conse a est FF (LI, Li the head of Sy (0r Ob. FF(LI inclu the eatgorial fat of he nominal phrase ed houk! have argument (A-postion) proper, icing the abit 10 Serve ar contol a binder tate rents and extension ef ober ‘ations of Postal (197), Lani and Saito (1991) argue tat this ise {or obit rising: Obj rama covery 10 Spec, Aste] for Case checking as basicaly the sme proper 2a avert bjt in his egac 35| ‘strated for example (35), wih jadements somewhat ead 85) the DA [proved fhe dfendansto be guy) during each ther ial) the DA (proved [tha th defendants were uy uring each ‘thera ‘& theDA [occses the defendants during ach othr ia) For the consusons to erty ver to the Mowe F theory, it ust be tat the feturs adjoin to. Agrp ako have A-posivon prope. ‘ccomisanding end binding he standard ay. Thee every aso 1 ‘sue this to tet. ‘Consider ch expletive contraction (36a)? (06) a there iva [book missing from the she} 1 thee sem eno be rome Books oo the abe © categorie nt Tranforntons m ‘Aprement i with the associate ofthe expletive (namely, ool), _ which n our ers regis thatthe ventures ofthe associat ate 1 ‘he checking dora of avn I, Bot the operations cover. Therefore, | itis notte associate that rates tis unchecked featre evi the | st in site, The tural esumption, once agai, i tha these fears Tneepreations of (37) would terefore be roughly asin the prt ‘ves (8)—though only roughly: because on thi analysis, cy forma atures ofthe astocate raise, eaing is sani fetes Behind ‘here arent any ites on the wal thee ae pictures of many presidents fn the wall) liners icosiderobie in is woh) pire ret many (onthe wall) pictures are of many pesdents (onthe wat) _ Similarly in other cases The gezeralconcisons about expletive cone Sacto fellow. Spatially, the ascinte mu have ancheched fee eter Iocaty ets. The HMC i legly snoperive, however it Uunertoed 1 apply o feature movement * 1 abo follows thatthe expletive sere ctanot have checked all the | Features of if had, 1 would not bet etimate tang for be as ‘tus, Pity, sere checks the strong enue of I (EPP), oterwise = ble constructions sich a5 (37) would not ex in the fit place Bt there most lack Case ce features, or bik; her, all fetus of 1 __ Wil be chaked and the aecite wll ot ine. Ther wl be 0 way 9 expres recent of matrix verb ap associate: (39) wl hve the sme Maus 2839). 1b tereseams totes mas Covert ring 10 A places the features of the associat ina sae formal properties of Spee, Aas We | Mherefore expect the associat o hve the Binding sod contol properties the over subj, analogoinly tothe eae of covert bye asin to ‘Aso (See (3). Theses ke w sett asp erm in 8 wll 50> ject langunge, Here we expect that he coumerpat o such expressions 25 (40) should be mists contating wih (0), he oot oe chapers (40) 4 shore arived tre men st gh) without Meniying there av withthe ows Books thre men frm Engle (41) “Het hee men (igh without denying theses 1, “Found with hi oun books thee men for Elend “That appears to Be comer. Thus, we fd the ollowing contasts be tween aha (24-0) and French (@2e-d (2) a, sono entra tre vomin sera eins fre entered three men without ning themes tse mer entered without ening temsehes b ne sono entati tre fae dune orca fof then are entered three without yng anything oem trae entered without sping eyehng® cea enine toi homme sans uncer there i ener thre men without ening themscees Sch eo eked is vans Surmoncer there ofthe i erred hroe without ening themsees fp Kalan, with ull subject expletive sharing the evant properties of English tere, UF eis to ¥ of Subj (ctl. foal feats) a Signs A-position properties to Subj for binding ad conto, inline ‘te eae of mentation that make it cla thal ‘Subj is ovr the letemal domain of the verb (objet peston. basally). In French eh he fll NP expletive analogous to English ithe LF operation i tore all atures of the wan Fptvase. the penal target. Mane already buen chacked by the expleve. Accordingly. there i no over raising, hence no binding ox con ‘Consider the German analogue (4 (43) cs sind gestern isle Leute angekomnen, cine there ae yesterday many peoplearved without sich 2a ientifiiren theives 10 identity “many people ated yesterday without entfhng themsches' ere agreement swith he asccts, not the expletive, and the nding a eontrlproeris ae a in (2). 8 pried ** Cegoee and Fanon 5 Arerent with the ascite, then, apps to corre wid matte ‘subst binding and corto proper forthe asl, as expected on ‘the minimalist ssurapion that se an aprecmet ar local Spc head selations and that feats ese under Last Resort, covery if pos ‘We wll tun toe dose look a the factors involved [Note tat the entice dissin res onthe assmpion that Case ant ‘features of «noun N ae part fis interna onstiaion. eter ini ‘8 1 oF add optionally as Ns selected from the leon forthe ‘swmeration. Testor these fetuses form pat of FFIN] an fron ‘thin the “package” of formal Fexture that dive computation. rain, 8a unit We have seen tat the conchuson motivated on independent ‘pounds iis cones by the central ole plays within te emp tional system, which wl efter confide we prose, Abn: ‘nent ofthe consuson (sy, by tking Case or @eentres of No be separate Sexi catgories with ther own pesons in phrase marker) would use a sight complication, “Though eore predictions appar to be vet, many quesions vse ‘One imme problem is thatthe raised esate canna be inde i sch expesion a (8), whee 1 Is te race of thre (ae Lasik and Sato 191). 4) "there seom to each other [1 ave been many linguists given good bole] ‘We know tat this isan expletve-ssciteconirstion with essociate premen, as shown hy relicing of eur ele with That env ith an apparent direc eontragitio: the anoxia oth can and eaneot ‘eo. ‘Th sohation tothe paradox might be within binding theory. Suppose that an LE movement approseh of the kid mentioned in chapter 3 ‘and developed in deal eswhere inthe iteratre groves correct. Then the ed ofthe matin clause of (2), at LF, would have the sre (45a) or (4) depending on how cover operations ste ordered, whee ‘Ans the anaphor and ais the X® eemplex fone fom Fad the 69) 8 ADTEFUspur) ah BU FFUinguts) (Anat ‘On resonabe assumptions nitber of thew sructues Gufs a5 l= _Blmae bing theoretic configurtion, with An toking Figo) a ne copter is amxadent No sich problem would arse the examples (40) nd (42) orn such standard examples 5 (6). (46) hey seemed to each oer to have ben any] “These phenomena provide fue evidence that the features ofthe sosiate ie (oT rater that adjining othe exltve, over and above the fact that this operation is he oma one while adjunction fom the tssocse posicg to the expletive would be without any analog. TT ‘junction wee in fact tothe expletive hen there might Beno eet teen between (4 ad (26). The phenomena ao provide aon tsidcee for an LE moveren analysis of anaphore ‘Overt ring of Su sad OF to Species an Ahan, What aboot the cover analogue Is the postion of the adjinad features of Sebi land Ob nso an A-posiion? Ife pot cst the st matters Bow (or i) the question & decided; though A. end A postions die i the ws Properes, i not clot thi they Rave more than a txonomic woe Inthe Mininast Propram. But suppose that an answer is required “Then ve concliethat covert sanction of features of Sub) and Obj stables an Achsin the conept “A-positon™ should cover the Dosion cupid by the formal features of Subj and Obj both before tnd alr the adjunction operstie. We have taken A-positions to be ‘hove martonly Lelie tos head H. Adaping the terminology of tion 13.2, let us ad all sublatels of H™ tothe postions naromly Crete oH, ining Hise, fetes of Hand any fate soined 0 “Thee conclusions appear to accord with binding and control prop cater of eovety mised cbject and subject, on stndard assumptions. Th ‘iho foows that rated iy cs (ov Riel (1990) ses) thou be these of A-ha, though at may wel follow fom indepe ent casero. 45. Iterptbity and ts Comequencs We have now reached the pont where dstntions among the various ings of formal features of FF(LD are beaming important, Let us ake 2 coser lok a thee, continsng to sume that F automaticly cares tong FFTE] am ovr,» ful category es require for comverpenee (perhaps just PF converse). Coto and Tavlos ™ 481. Types ot Fates ‘Alona wih oes, the folowing distinctions among features ave worth sone F, where Fis cater ‘As sind car, there ae forte distinctions that roseeut hos of (4? some features are intinse ithr lated inte eel tern Loe Severin ty se tas ter a open ded bray Lt ‘Suppose we have a comers! derivation for (48) (48) we ald pias tsi tures the thc exice tts ice the eategori er {1 person] in Fw), [person] aed {hua Frans). apn fecusative Case] in FF (hl), and [assign sorsnative Case] in FFD, prion tenures inate (hr For the oouns and the ¢festes of a ‘As lend dicosed, these stntions ene ino informal descriptive ‘sage. The stiction lo eoeelate mor ols with oer fact. Ths, the dears ofa DP speciereammony show up both onthe DP and ‘he veal eae, But the Cse feature of DP dss not appear on the Dead. Threat feast a rendency For g-festts tote overtly mated tien raking tothe checking domain ir over rather than cover. 85 in verbal agrement wit sbjct versus objet in nomiatve-acustive Fomguagss with the EPP, or wsible patil apreenant in French a 4 ‘flex of overt asing In the Move F theory, the diferene reac’ 10 {Spec 1] versus [FH] consiracors, features tending to be overt on i the former hut not the tater. Let ux enatively sue this to be {he ese though a principe explanation is tacking, andthe empirical ects plainly tequire much lose sera overt fe broader range The intcnsi~opsioual dsinson plays virally no ele here, But here is much more important distnetin that has so fer been ovetoked, vide, certain features of FF(LD ener into iatepettion at LE ill thar are usnterprelable and must be eimiated for convergene We therefore hve a ruta distinction interpretable. Among the Ins rtaleatees are eaepva atures and the -eturs of soma | os hapers “The operations tat interpret (48 atthe LF neta will have to know that hud is 4 V apd plane an N with the fetus phrall {human [3 poson On the oer hand these cpeatons have no wy toner the Case of apne othe xpractient Features of Bul which ‘ros therfore be iit foe LF converse ners ot LF reais ony fosly 1 he intrinsic optional dis tition. Thus, the optional feature [plural] of nouns & Interpeabe lence not tina ut LF. The Case fetes of Vand T are itn bt lmerpetble, bene eininated at LF (assuring thatthe ae ie Sieuised from the semantic properns at they esl eet). Wf tows thatthe atures of the hand mst he checked, othe dation Cashes. The Vterpetae Fetes, then, ae etegonal femurs gener iy and @ feature of nouns” Other are — Interpret Tnceprtailty doce rate clon to she formal asynmetry of the checking plain, which Holds betwext «Faure Fo the checking do train ofthe target K ond a ublael F of KF iabgs ~Inereetabhe Seng of feature, oval features. the Cascasigning feature of T and ‘Vogfenurs of verb and ate. The age has Interpretble eaturcs, Sich api categoria enue, ut these never enter ito checking relations. F in the checking omuin, however, can ean Terrelle Feature, ictuting enteral and ¢festues Thee dereces between checker (within the tarp) and checked (iin the checking domain) Paya certain role in computation They ive seme meaning 0 the nt five asymmetry, tough wil cay weak cortion 19 ssthe nation “agrenent™sbows, “These descriptive olmervtios aise two obviows questions: (3) Why iss sublael ofthe taret that enters 2 checking relation invariably w toerpetable? (2) Beng —Interpreable, why is F preset at all? Question (2) is part of = more fundamental one: why does mgusg= tave the operation Move? Hit dus, and the operation ismorphology- ven se sure, hen there must e ete chakers inthe terested ‘aepory. The fet that these ae always ~ Tote gsi highs the speci ue the propery of displace of eateporis ha shar acters of hima language: te sole freon of tse etre checkers iS to free movement, sometes overly. These question bein to Fil Jit pace a we look tore coe athe dheory of mone “Cane cifers fom fetus i that Hs alnays —Toferpretable, for both toms ofthe caching relation. Case it therefore the formal nts Cason and Taner » pa exslence. and it int suepsiog tha this nie line ingiy has is origin in Verpnas Case Fite 452 Octing Theory inerpreabity at Ls doin by tre cup conn st i rly an importa propery fies. Aten to te Se at once tht tenn of checkin 0 fr pop ce frame ‘sian the ne ne of uit vero. Tse nee um shat the tot cere etre h i potent sre heen ‘We oe, Rouse hat there ae uc idreresdepentng one prebiiy neier ston se ne ook checig te Slea A ‘heck str the. sia the cmt ye, br Stile Tor inerpretion a LF. atta amet be cores Some fe {ues remain vl at LF even arth oe ich. rape ‘ene of nors wich eine And some pay se ne Scuile ihe conputaionl em sen eh he Coe ete euns. oreapl wheat fe coed afte cece ‘Weitere have to ge more uals ef he reation Be tween sy a LF apc tte compose ic {wo ropes are reed the stipe pon (99), Fetes tlt LI ate scr tthe compuston Gu Shoo whether ich or a Fetes vile Late ences to Gy. nc check in (4 ols ito cetin (9) on npr nan ines Te ak ad pr of te stration allows at onc fom sight mot featon othe tory oleh. acy a npc Wat as texdod anyay. The checking epraton taken on om eter work tas « numer fo ear, For one tng. roms reunden th ‘seve opts re dtrinie by letinfom te LF req talon it Bt we ow that he propo se Tee ‘yee dps ih he checking operation emi? Suppose thet we do 0, Kein js othe elton from wich the Aeration derive the aching ton ht ok berms eas of *e checking domein an of he tart ens tat oe ay detec. 2 ve hve een) We es for ssimed te ih operon Move dino ofthe Sonoma 2) pened er 5 capers (60) a Finan onhected feature B Fenterintoa checking relation with a sublabel of Kosa rest ofthe operation “The point of (50) as to prevent & nominal phase that has already satis the Case Flr (rom ris fur 1 do so agai in a higher postion, The concason score, butte formulation of the principe ust be rnd to yield the condtion (9). We now fave the means to do 0 quite sraighforwaty. “The ey 40 the probien isthe testo nepkeeted property Eater frtabie This property determined by hare output conditions hence (hailble “re of charge” We ean therefore make ase of it to restate (60) without cet. AS Uoughout we restit tention to formal fx tures his ogi to te computational sem. To begin wih, Kt ws ‘imply (5) by kiting (a) enti, sowing the arabe Fin Move F to rage ove formal fertres ely, We ten repace (30) by 3, the Final vento hereof Last Resor, (51) Las Resort (Move Fabes Fo target K only fF eaters into checking ‘elton with x blab of ut we stil ave to capture the intended fet of (0): ruil, that a {lovee feature is “fozen in ple” when i is checked, Case ‘ing the protons ‘Continuing to wnderstind “és” as “ivi at LF but sccesibe to the computation,” we now rformulted the operations of checking and deletion asin (3). (62) mA chckod fetes deleted when posible ', eked iscraed when possbe rasueis “stronger form’ of dteton iminating the ent extiey so that ie iacesuble 0 any operation, oot justo interpretaiy tr “Possiiny” in (5) i to be understod relative to other princes “Thus dltion is “impossible” if it voles prinspls of UG. Spec call a cocked estore cannot be deed if that operation would ontadtche ovetding prince of recoverability of deletion, which Should hol! to some fsb for any reasonable system: Interpetable Festre canne dete even if choke The questi of erasure, then. tres coly for a ~lntarpretable are F whch i eras by (520) Gegais oe Trecormtoe ™ nes that operation is tarred by Some property of FP should be ‘ely detected, to avoid excsive computational comple One sch rope is paramere variation: F could be marked a not erated when leted, x posiliy that wil be explered blow in connetion wth alte Spee constructions. Tenatively, tus same that thf the ‘only relevant property of F rate i abe bated if it erentes an ieiimate objec, $6 that no etivtion i penerated. That toi tril tried The crc ase tas todo wih easure of an emir tr a of syntactic objet Let (y (06}}- Erasure of avers N by N'~ {yf} which i ot leptimstesnactic objet (se (24), We cone hat (3) Ate of Zeannot eae. Erasure ofl ategory canes he derivation nthe parallelism cases ddscussed erie, for example, deletion snot flowed by erasure in he [N= computation under (52 what baprens inthe pencil om: foment, which has» wholly ferent character, is a separate mater, But Aegina objet arene formed by eave witha same ter note 12), Hone. suck errure na ated fo ths eason ‘We have 106 dispensed withthe checking operation. The problems Shout inerprtaiiy shined ie eater dncanion eis, and the Je- Senpivegneraliation (2) flows at nee infer a it vail Cae (49s) s tne without exception: Inerpeable fetes cannet be deed (4 fortior, eras) and therefore rain aeneabe to the eotpuation| nd vb at LE. Case (49) holds ness easre ofthe ~Inerpretabe fecked feature eases aterm or is hared by m parametied property ‘ofthe feature. Though examples ent, they ae fetus cite (296) ods quite genes. For expository purposes, I wil speak of delstion ss rasute excep when the sue aries. ‘The revision of checking theory is withot fet for —Iotespetale atures in the checking doreain suchas Case of an argue. eis thee atures that mst be nace ater checking the examples cused fare tical in this regard. Erasure of such features never creates an ‘ptiate objet, so checking i deletion and followed by erese ithou excepion. Features ofthe target are atvaye —Inerptale for ‘reasons yet to be explsined. The reed checking thiry ales them thou exception, and typically ere hem, One might ask what hap eur whe al fetes of FF(LI) are — Inert sd i cs he | checking domain of K: raising of th asoite ofan expletive or covert Seana ec ‘Sbjct arent for exemle, Not al of FF(LI could ers: iit di Sn ilepiinate object would be ored. Bu we nee pot seth Fro fem, beense it does not ane. FF(LE) alvays contains Tterpeuble features: the expo and entre ofthe argue “The only encopten 4 te conclusions ofthe Ist parspraph is pure cspletives to which we ilu “To irate th coiequeness tus return to sentene (48, bul irons, Wren the sult wei introduced nto the derivation within the vet phate, FF(we) includes D and specie choess of estar lind Cae, Sine I has srg D-festore (EPP), the estonia eanure ‘of we tases overtly (0 i cocking domain, pied-ipng he entire DP therefore the operation ssittion i fSpee Th. Tere ate wwe ways in tech we could aver in tis ae, depending em how Fis sted for the Move F operation IF =D then a ehacking elation is estab lished betwoen the ried categoria fstere of we ard the strong D- Fenture ef The Cat etre of we sechecked by Ta fee ir 05 16 the dessus, afer coven rising f the verb establishes the requied ‘checking relation, F could also be Cas, which would ean tha the EPP ‘ssid by the categoria Fearne a. fee sr, But F could tbe 8 ‘feature in this cave, Breause the vers ras ony covery 50 that the ‘heckng elation between fetus i nly estaished later Las Re for (SI) would be wokned sf features of be WEE noes in overt raising. Te Case feature of we is —Teterrecale, therefore eased when thecked. The gestures, however ate Interpret, hence acess 10 ‘ueter operations, whe etepora feature D. "Note that she EPP i voreed fre Case. Thus, we assume that ll ‘his of T inde the EPP in gla incogifitives. though ony ‘enol infeitves assign (nl Case, easing ifiives do not (Ge sc fon 142). ‘We eat now return ta 8 question that was open: why ae the textures of the target that enter ievo checking rations invariably “nurpetable? Suppor that &subabe Ff te target eaegory Kit Inuerptetable Suppose the feature F ha sees by the operation OP td raed to the chocking domain of Fis Inerpreable, entering ito 2 checking retain with F Both fetires sre Inerpreabe, ene Un ‘Changed By the operation, The operation OP is “ecally serious,” ot requied by the fetes that enter into the checking elation tht nes i, But OP might nonethlsscontbute wo convergence. Forex Spl, fee wider of FF[PT might exer ito checking elton wih (caegec an Testo oy ‘nother subabel ofthe target, one or the othe beng flected “sed or Gelete or OP might be a necessary step toward later operation that es delete and perhaps erase —Inerpreable feature, lowing con vergence. Such pssiitis abound, conserably extending the eas of Posie derivations and thus making it hawler to compute econo. Perhaps also allowing derivations too ily (ss might ete easy to ‘etermine). Preferably, OP should be exci. I Fis necessary ~Inerpreabe. hence avays ated by the operation. IF Frais to target K. then. she sblabel tht is cecke by F deletes and typically ‘Tis propery of feature checker eliminates the possi of “lay seperfvous” movement operation. I réiforss the mitt chara ‘er of the computatinal sytem, permitng ts eperans to be for sulted in avery elementary way withoo prolifratien of unwanted sian. To ptit ten. he “imperfection” of angus nce by the deplaement property restricted by language design 30a 10 Noid exceasve computational comple. Comer suceriv-olic raising ain (3). (54) we rely (1 eased to iliac) Overt rising of we fom r 10 f sees D 40 ct the EPP inthe ‘os deeply embeded clase the oly posi’ snce the sng nin tial dos not assign Case. D i Imterpretabl, therefore umafted ty ‘ccking Ii aces pun oes wt 1, styng the EPP i the ‘medial cause. Farther rising from ¢, to the matax abet can aes ny of the feat that erento a checking ration there. Conse a diferent case of successive eee ering. the simple c= ial eonsraction (SS. (59) Jo sages ABEL pet) ees rom te pedo inernal bec poston» (Spee, Ag () for agreement wth the ade, eased to Agr” By viet of Last Resort (3D), the operation must aces the festuee of Jom, which ‘heck apreemest. They ae Interpret therefore wnat, fan de raises (0 marx subject postion, sasyng the EPP, Cave ct ate ‘nen. Here ony of the relat etures may be sested, sine ll rr int checking relations (ne accu, the oes re rider), Jon has ‘ter ito doube agreement with ech ofthe wo Apr odes, hence ‘withthe copela and the adjective, What shows up at PF depends on ‘morpeleial pris mm caper ‘The example illusuates the fc that agreement cam be assigned with ‘or without Casein the bgher ard lower Spex, As] postions, respec Tels Sine the eteorial and §-extrs of DP romain accesible afer cccking wile the Cnee entre dows, x ingle DP eat emer itd ‘tiple satisfac of the EPP and nile present, but not ml ple Case retations. The later opin che core example that we wart to exchde under Last Reson ands scoters hack io Vergnaud's Case Fite. Bu he ober two deftly shoud be permite, as they now are. 1m (5) features of she suet ave Been checked and the Teter etale ones ease, Suppose thal (S58 embedded ain (56, (66) Met) sens ha Job isittige) “Thoush the Case featore of Job has been eased, its entegovil and ‘ploutures are orchanged. Therefore. Jol ean raise to matrix subject Spe. I, stshing the EPP ond mann agreement and yielding 23), repeated her. (59 obo io sss it siete ‘But Job ofers no Case feature tobe checked, s the derivation crashes rather han converging wit the iseypetaton “tems that Jk ine "eigen In (56) er efectnely “Rozen in plac.” as in he exaepes "tat oiily neat Greed (se ote 36, though not forthe reasoes fv incarc heoien, These reasons wae defesine i & fundamental ‘ey, Fling ota acount ofthe property lnterpretable and trl tion sccsliy to the computations est. ‘We conclude tht the rnc Case fstre of Tt deine not ony Soom its (paraneically varying) EPP festre, but also from the (ner ap inwaran) smn properties i vets. Being ~ Toteepetabl, he Case feature must be eecked forthe derivation fo converge: Sie i ot checked (5), the deviation rashes ‘Seppo that mates Hs (36) tense the ster ex cone Infinitive, the dvation crates again, for Case easns. its aang Infinitive censrtion is bavzed a i stands, presurably for sek "ional reasons: a noncnbesde nina cn bo cont srt with rbitrary PRO (some lnguages and constuction), bu nt ase initial with no relevant properties eter than the strong Dene EPP). Further embedding in Cling structures ritrodues he sme problem, so that Jom remains “rove i place in ($6) wih nia Svell! eter and Trforatons 2s ‘Suppose ha a language were 1 alow the eonsrction (56) but with ‘erly aprenent checked i the enbedded claw, not Case Thon rasng should be possible. The eaepoial and Qe of ln are Tne pretabl, hence accesible even when checked: and the Cae featue t ‘checked, Rese sll ania for checking, Rating Jo to matin sujet, we derive (87), ga with double greet an! double sac sion ofthe EPP, but wih ony one Care eatin in the rain case Such consietons have Ben eportad ins number of lngoages, ist (i his content) in wader Gresk Inga 1981). Assuming the deseipsons to be comet they are an have bec regard a) pri face ition ‘ofthe Case Fier. They fling ple the manner ust ered. We expect the matrix subject in (57) to ave the Case required inthis pos tion, wich could in principle be distin from morn: in am ECM enetrocn. for example that xx, it confi the contain that ise nas nt sine inthe lower ela, which would fave ented ‘Succesive-syec movement raises frter questions, to which we wil rerum afer the groundeek hasbeen ad ‘One consequence of ths reanalysis of the theory of move is hat IntrpretsleFevtres need not eter checking relatos, sine they su sive to LF in any event fo particular, capil and feature of NP feed ot be checked. The concinion rests an outstanding problem ‘concerning inherent Case it has sever been ler how the estes of ‘he nominal resiving ihren Case can te checked, inthe absence ‘ny plausible fcsons aepory; bat the question es ot are hey eed not be checked. The sne consideration overcomes the problem of ‘iccking of features dihwation or coordination, a in Joh and is rents ve ere * ‘Consider corporate into B (sy. noun ncorperat into verb), so that hes the morphological feature [afi] tat lows the operation Ie tis features ~Inrpretaie,eacorporation ofa wil be impossible because the fate wil have been ered ad wil this te unavailable for further checking: © wil be unable 10 sjin to second head, even ‘ough is other properties are inte. “The improved theory of movement has consequence for mine Spee consructons, which are permitted in prinple on minimalist asm ‘ions about plrase ste theory, noted erie. I hi option fe realized, we have the tues (3), with posse Frter polation of on or we ZY sm OX © conionee ee we may tectatvely assume Spec, an Spe to be equidistant st for movenent eng within these minal erin” Suppse a language permis (8) fr sore consttion. Soppos far ter thara Inept fentine F of His 9 neces ered whe checked and deleted, a parameterize propery. F can then check exch ‘Spee optionally easing at soe pot tense enversnce Fis 3 (Cae fete cou apn the sme Case repel such account ‘as teen propose for mile Case etecing Japanese and he lay ge eno 96). Watarbe's 1996) ljered Case hey as reste fn note #9 coulé aso be fermuked i tee terms. In section 410 we se tat Sina as have steresing consequences in ares of more ‘nirl concer bere. Spee abo allows am escape hutch for Relativzed Minzaliy vila ties ad srambiing with A-postion properties (binding. obviate weak crossover eflecs, ec), unke scrambling to an Apion, which Under carer assumpsion, inves fll econsirucon the idea Wis Intend by Reinat (1981) to seoust for Wieland vilaens in Hebrew Ura (194) hols tht superaiig, A-srambling, and muliple (Cae signmert correlate in many langues. Io, thet would lend fur ther ene support to the concson that (5) isan option tht tanguage may have” fom hs reached the construction (56) andthe mur ‘merton contin sn expeive se that ween deve ov example, 24). repeat bere. (59) “thee sem that [8 lot of people arene “The expletive shore checks the strong Fecue of 1 (EPP), but it i 0 heck soe fare of HI, ar that i ~Inerretable ed mast be raed for convergence The ~Inerreable fenures of W are its Case and features. Once agin, wes tha he expletive must lack Case or ‘features ofboth (se dacusion of). Seppove that here has Cae so that ony the features of H remin unchecked. But the @feanres of Subj ae Interpetabl. so Subj (ac ally he formal features of is head) ean rise covery hacking the (entre of 1 and allowing the drain to converge snore, with tn iterettion silr 10“ seems that oto people ae intl Wolo that ere ms ack Case, ‘Suppose thatthe expletive has features, Suppose that these do wot roach he features fs svi, a in (98). repeated het, with ere poral and ts essoate @ mar singular andthe raising verb poral ranching here 60) “toe sero be man in he ree “The Gentes of seem ar erased inthe Speceud relation with sere ‘The ents of here. beng —Ioterpcable for ah expletive are abo ‘rsd under this checking raion. The Case entre of sem i nse by easing of se associate a man Sivee the ¢featres of «man ae Inter Pretable they neod not bechocke, The desrtion a 6) therefore com eras incor "We corlide then, tat the expletive bas nether Case ae entre FF ere) contaies only D, which sss to say the EPP: the engl ve has no Feral Fete pe from i xg. Notice that agreement is overtly manifested on th verb tat hs here 1: subject Eater we considered the sgetion thet vert manifestation of features ea ection of the [Spec] aber tha fy FH] relation ‘The observation about sgrement with expletives eesitent with this Froposl but it would coat wi he strate Kes thatthe dane ‘on refs overt eather ta cower agree. The fv suggestions ae ‘pirical itn a this ie, perp ony hi cae, Seppo tha ser isa pure expletive lacking semantic fetus we es oral features apart fon ctepory D. We therefore capes it ote Invisible at LF, wo sty FL. We Rrow that thre cannot be trl ‘rasa when checks that would wolte he finda condition (3), forming sn epinate syeactic objet that would cancel the derivation By the peeral pint fdeetion-erasre, (52) i follows thatthe cate oil feature of here 1 only deleted when checked, not erase, long sv hes of us re ene 12). a coupe Since the expletive neces tacks Cae it must be the weit hat provides the Casein ordinary expe consrctions such 5 (618-€). (61), there a book onthe shel by tere rive yesterday a visor from England © Tempted [terete be book on the shel] “The ssa mst therefere have the Ces that would be bore by BP sn the consrutions (62-0) respoiey (02) DPis... (DP = nominative) DP erived... (DP nominative) © Lexpecd [DP tobe...) (DP = acusve) ‘We thrsfore canst axel the partiive Case theory of Hlth (198), conta tothe espn Sn chaptes 2 “Taree dsnetion Beween expletives that have Case ard fates sad the “pure expletives” tat ack these Features: English, if and there, respectively, The disiction is neither clear noe sharp, bu it deat for ou ited purposes” The former sus al properties of ‘he IV fd they check, erasing the elvan Features and therefore ba svocne ring The lar do not eras the ~tntrpvetable Features of the EV bend. Thetfore, using permite, treet this element: wo ‘tis equ for onserence “Two consequences flow. The direct prediction s that expletive eo srotions wil manifest verbal aprement sith the ssceate jst in se the epee lcs Case and Features Fgh ere, Geman x. ad alan pro but ot English f French which havea fll comple nent of features, Note that the cistincion only petal elated to (vert manifestation" A more inerating pediion. and ove that will ‘more df o confirm itor, that just incase the expletive lacks Case and dentures, he associate wil bin and contol asf were io the surface sje ponion, We hae wen some eeson to belie tht thse tre ‘We might mk why languages should have overt expletives Inching Case ard ett rther than pro, ln part, the reason roy rs 10| the ull sujet parame, tat more sem tobe volved. Thus ela tic and German both hve nll expletives, bu ekndic is «nul sah Tanguage and German not. these languages the exelent for he ‘aleve specifi two forms nll and ove. Thr sition ec 0 ‘e completa, determiad hy stra fectors. The opal rel ‘Cg so Tranomtons 8 would be that the overt wins is used ony when sisi eqived for onvergnc: PF convergece spe the to fos ee deel within the coven component. That could bear if the presence of the over ‘cape reduces to the Vaccond prope, which could belong tothe ‘phonological component iter no orden inthe core Nh eat uation, 8 we have assuned. Tat seems promising In both keguges ‘sams that the overt expletive weed only where the Vseccnd prop, sty otherwise hol’ TF at tus out to be core, then the expe ‘bay wel te mll— nothing beyond the cateprial etre (0) thoughout {he Nov} computation. The overt featere ae then st only i the course ofthe phonological aperaions ough code he eon “Toovgh a serious development of the theory of exletnes requires ‘much more careful examination. including & far bose rage of eas some concusons follow even fom fly weak considerations, 8 We we rend sen and wl eotnue to i swe posed 454 Chase Type Let us urn to the Forma etures ofthe functional xtegoryC (ome mene) tht determines eleuse pe for enamp he feature Q To inerropatve aes he onstruction (63, (©) Qf Jobo gave DP to May) Qs plainly tmterpretable; therefore, ke the atures Of « nomiea, it eed not be checked—usles ii sro, in which cae it must be seckod before Spel Our ideation eo be cersrcte AS swe own, languages lifer ia stengh of Q. The stoog Q fete is st ‘shed by 2 fnare F For Engl, Qs song. Therefore, when Q i nods int the erivation its tong fare must be eine by inseton of Finis ‘hseking domain before Qis embeded in any distinct confeuration (ce omay eet tng oi by Mero Mos ae __ Consider the Merge opin, Since ti vert fllcstepory 9 must be ‘nse inte checking domain of QI the operation ubsition, 2 ‘come Spc. Qf if adjunction is X° category. tn Enlh the to ‘ues related by (68. (64) & (1 wood fo wheter Qe fy] 1, wonder fr FQ) he ef ye often called te whence, wb we can tke to bea variant of D. see 0 Coun [Note that checingreton canbe esas by Merge, though the notions haves far teen dacssed only fer Move, We wil ern to the question afer proving the eo of toveren ‘Levu tum tothe second and more inscate posi F emes the checking demein of Q by musing. gai. the options ae substitution or aunt. The subsition option s realize by rong of Fy to {Spex 0) by overt wh movement, which pd pies al eategory for PF tcomversnce The adjunction opin elie by 1+ ring. Tiss infact sing f a eral estar s proposed carer (se ()). then Fin this eas is (V].There we generalizations and Irpiage specie proper ‘is fat any account ht depns fom rinse sumptions can be tomsidered explanatory cay insot at has independent jusication ‘der raising (63) de two lita output (65) and (65). de ening on whether the rong feature of Qi checked by adjunetion. Satstution (we abstract fom the contest between enbedded ard roe fons) (65). dp Joba given book to Mary) (gues) mics Book (Jol pave v0 Many] © (aoe) whieh 0 Book, Zo ave te Mary ta (58) DP of (6) ise Bak. dacs to Q. and the contraction inceprted ox 2 yon no qution. In (6b) DP oF (6) i which ook Sd the onstruction ke nerpreted a soetbing ike (Se), alone the Tg shechdi seton 35 Fis Interpreabie and Bice ned not he eheced. 1 bereor rss to the checking domain of Q ony if this option i sete 0 elininate the sen feature of Qi which case an enti wh-phese of complex is ped ape, sehen in Spe, Qe edjened to Q.reopextnely Suppote that DP i which book und the son feature of Qin (63) is sts ty adjunction of Talon. asin (58), so that whet surfaces 5166 (66) Sohn ive wick book 10 Mary ‘Covert sisng of the wisenure f unoermary and ence impose by econ eons (wo note 6). The interpretation of (6) net {6se) ap H would be ithe whfiatre ented covery to adjoin 1 @. (ee) converges, ith whatever inferprtton i hus~peaps bbe (0 pot made increas with focus and ecko questions. ese er Cnegois and Teron, ™ Svopos (62) is embeded and DP = which Pook, The 1-0 option is ew ursniable (eran. iti valle, and yes an embed Jercrto question inkrpreed as sbberis). The wh-phrase, however, ay raise overly 10 the embedded [Spec Qh. yeding (67 with (60) bed (67) they remeber which book Q Poin pve 110 Mary ‘Suppose thatthe ratix cause if alo inerogaie with the comple renter Q: Again tere ae two says to check iy sone featue, 1- ‘abing or whmovemen, ying either (6) oe (84 (opnin abstracting fiom the roocerntedded diinction, (68) do they eemerber which book Jb gave o Mary '. (ques) whieh book [key roweber(¢ @ Un pve 10 Mary “The secon option i value besa the wh etue i Ierprtabe in (67) bens ccs othe computation, (6a) is unpeobematec itis yes-or no questoe with 4 embadt Inet question. (68) converges wth the interpretation (9), ©) (ewes) which. book, they eemember [QJ gave to Muy) ere the embedded clause isimerpreted 8 yer--90 qustion, pres hy gierish, but inary even flere frm the iterpetsion of 0), __ ich resus if embedded Qi replaced by detrasie € (perhaps my evant esau of he fete characte ofthe embeded ce) 20) (gue) wick bok they remember that Jn gave to Mary Ween penton hfe) ese ge deo te fers qaston seat) Ten het {yo cpt or ne becomes Sees ats ung hr l'art sce) ‘rh PF mo eral chang MDP ohh he finn sist Pao at Lp em oe na fr Cas) ‘Te sete daw toto tthe Ta sed feted rman ie meebo ec in of ahh ese mp hes nl pan AN) moe or hana at rns cme "inst ncn a), Tey ma tet age | smnive nerve stacy fr hemsossen aka [_reinghs patos oonog acne ae Soa OE ee ™ chapter On diferent grounds, Reinhart (1983) proposes similar analysis “These bas conlsions ae weached by Tea (1988), in study of| ‘ypologealy diverse languages tht cates further the fb 10 use ‘morphological properties to aezouet fr some of the problems opened to investigation by Hua (182). The essence ofthis theory seams tobe ay det consguence of imal assumptions, ry pursed Tn diosing the operation Merge in section 421, we came to the corcluson tht it su be over, with single exception: corer insertion ‘fan itm Inching powolepicl features. neces at the rot. We an reste alention to a complemetier C. The option KA ope ‘hat phonological tall C maybe isried covertly a te roo, C cul in pencil be strong, in which case i rages a iamedate operation {o erase the engi festure. Sines the tapered operation i cover it ‘not be subsition i Spec], bu ust be feature adjunction oC. Dosuchenes exit” ‘Consider Gs dvartve C, whichis weak. Can it all variam be inserted covert ina rot sae! There ls good reason to belive tht itcen, Dedaratve Cis one ofthe fore indicstos and therefore mus be pret for nerpretation af the C1 iterfoce. Buti pever appears overtly athe root we have (71), not (716) understond as» dela ssrton) (7h a Toda es 1, tha John et ‘he natural coneasion that Cis inde troduced, hut covery. Fur ‘emore, covert iseton is necesary on grounds of onan. we sume that Preerastinate lds of Merge wll as Move * ‘Dscouiie properties confim these conusons. We do indeed Ong sch oot eats a3 (1b) or (72) with overt complemertize, but not ‘with dotaraive ores. (72 that John eave “Tous, (718) and (72) could be the avers to the questions (78) so (70), respectively, ur not (730), wich eal for = edarave asieron. (3) = what die et you 1 what would you prefer what happened yesterday Conse inerrogative C, ey, English Q, which we sll tke to be stron Supoveit ered covertly atthe rook 0 ye (8 Cogoris and Tratertens 5 44) Q1D Py, wll ee DP ‘We can rule out the possiblity that this the variant of @ sated by a= IML seMing a yevorno quston with I+ raising inthe overt, case, That variant has phonoloykal properties dat determine rising it tonation: fit is inserted covery, the deviation wil uerefoe eh 2 LF. The only pesibilty, then, i that Q regs interpretation #3 3 question, which has no phenclepical property, eiving intonation changed ‘Wie might ask why the variant of Q satis by adjunction of Fg —[V] ‘oes not have a rll eta, tke decarative C. There cou be sre: tural reson: paps seme bare gina «el cement with an fal feature (Agr isan exception, but the probiem wil be overcome in section 410; iFajunction to Q is « more snout counterexample. Tete sre so funciona! motivations. Thus, f hee were null akemant for Q wih song V-featre, the venteoce Join wl er Bill woul be amg sly interpreted as delat or interrogate ‘We test attention, he, to cover introduction of Q il asuning ‘4p te sro it Engl. Cover subeiaton is irspossible, the swan ‘entre has be satisfied by adjunction: the tome ete of Q mut be che by Fy = Int ‘The sueture must therefore contain whine with & whfeture hat adjoin covertly to Q. The wlphirase might be the suet the hit oF an aunet, asin (29) an TP lacking C at the pest of Spel-Out but imerpreted as a whuestion at LF (declarative intonation throughout) 5) & Qlewho wi she cal 1 Qe John wi x what © Qe Joh wil ite ear bow (ay) or (75), the concasom ncords reasonably well with the ats, which have alvays been pues: why should constivtan that seein 10 ave all the overt syntactic properties of IP be interpreted am wh ‘ston? Case (08) yields the interpretation what will John fi, That is allowed in some laguges (Freeh), bu is dubious a bs in Engh (Case (55) shouldbe interpreted “how (why) wil Joho fi the ar ‘That excladed generally. The rein res follow if we assume that Song festres cannot be inered covets that some vata of he insta steaeey has tobe employed (posse for (Se) and 38, Moched : bE 2 hae for (5). which allows no variable Formation i the whepase: se note 68. Let wt assim then, that cover eset of song ears i indent tuned. One mipht suspect tht the posibiy eau Rectse of sy of evdanee. To pu it cere the interface representations (2.2) are virally ideal wheter the operation takes place or not. ‘Te PF representations wit at enti. and the LE one ler ony svi oem, and nota all in interpretation. Suppose there isa econo prin (7, (16 renters te numeration ony Lhasa fet oF out Wh ep tothe PF level ft canbe dein in ers of teal ide ty: wo outputs re the sie they are decal i phone frm. a 12 seve ony if changes the phonetic orm. At the LF level the Condition is perhaps sighly weaker, allowing = narow end readily “computable or of loa equvulence 1 be ieerprete 5 ienaty*” Under (7) the referent sil deere by the sera, but ‘utp onions enter into determination ofthe numeration self they sie the operation that canst the rumeration from he lec ‘Covert iseron of complemenier has an LF ef and thereor ot bared by (6). The nat of strength is somenat dere. estar {5 ts presence i motivated only by PF manifsaion, i exnnot be in: tered covert, under 75), oF woud Rox have ben inthe mumeration {all There sa good del ore to say about thee questions. We will ‘return to some of fei aspects ina reader framework in sction 4 ‘With rgd fo covert insertion of eileen. we are ei close to whut sce wo be the base fs on only mcimaist assumptions and sith some apparent language tation thar seems rather peripheral ‘ough «good dal reine to be explained. 455 ‘The Minimal Link Conition ‘Seppo that nr replaces Meri (7, eng (7, (07) they remember hich Book @ Hob gave 10 whoo lt ‘Suppo tht (7) is arog wit te cowrplementizes QAP ats ‘a roor contruction, the strong feature of Q con be esate by 2 uation of to Q oF sbsition of whephrse in (Spec, Os i 8 be (8), ely the later ptien i eB. (98) s,s 10'they remem [bic book Q Lobe eve #10 whoa Cugorie and Trassem: ee Embed or not there ae wo wh phrases tha are candidates for ris sng to [Speco heck the strong featur: whi Book (dha, sien 9s) and (796), (09). (gue) fahich bok @ they remember Qo sve 40 eho (gues [to whom; Q'ty remember fich book), @ tose fell (90) 0 Whstnd otaton Irs barted straighfeewatly bythe natu fal condition that shor trewes are peered 40 long Of in this ‘2, by sing of which Ao to yield (98). This operation is pers ‘He, since he we earure of which Book terre, hence wes and the rasing operation places it in a checking selon wth Q erasing te song feature of Q. The option of forming (8) bars the “lone move eure for (96). Bt (98), though converges devi Sin the ease of Let us inerpre he Mira Lik Condon (MLC) 8 rung that 1» given sae of a dentin, longer ink fom a 19 K cans be Fovmed i thre sa shorter leptin for Bo Kn thee tems the A-movement exes of reltivied aieiahty can be accommodated (lo fst approximation; we wil stn to farther comment). Ii net That the iad ttn it devia ater thee ino such drvabon sn he aca fore derived by the MLC i deviant. What about the Anovement cases (superruig)?” Suppose we hve connected (60) (20) seems tht it wa od Job fe that 1) Raising of fn to mae subject position i « Relavized Miniatty ECE) voltion, ut tis tured bythe "shone mone” option that aes ‘110 his postion. Raising of 1s a legtimat operation: though is Case feature has been ensed in IP. its Dfeature and ¢featues, though kod, remain ase “Thera diferences between the A- and R-mowerent ct that have tobe deat with, but thee ssid, both Kinds of Retized Minna ‘ation fl loge oturaly under the MLC ‘Ces analysis of format features ths allows us to resect am iden soot isan vilaions that fs Bn inthe ai or sme yeu: they i wove longer that-necesiry move ad ths fill under as proach thet Jas somtimes been suggeed to acount for superionty henomena | 5 cape ‘Te ies ran into to problems. Suppose a derivation had reached the termediate stage” Ef (18) and (80), with iterate eatery (which ol, 1) loser to he intended age then the one we hope 16 ‘prevent fen sing. The fist problemi thatthe ntreeiate category as its features checked, so it should be ffozen in place. The second role fas fo Go with the enge of pense operations af stage 2: {here ave so many ofthese hat i Har to ee why easing ofthe inter metint ceepoy isthe “shores ove" That problem wasin fet more mera: tho was far fom clear why rising Of Jab wo [Specs Tn (I) these ove" (81) Ko was od John hat) Both problems are now overcome, he ist by auesion 0 iterpetabiiy ‘offenres, the acne by adil narrowing of the ls of permite Sento nr 3) at Res. rw rm cr ote fos betcen the K+ and Amovenet vison kandi In he former ee the eation “stan the MLE comers he Ice Sor no Resing | tlio main sss he EPP andthe fees {ico bu ot the Ca ture. But owe T has 4 Inert Cc ene hich ols ed dese, nse eration Gruhn tee of Acrovenent ee Averett sone” ds ys omer ation ert acount fhe mrtg ton to Bo oh, He mi take the MLC toh pr ofthe dobten of Moe, he la tan conry conden i coms aang crt eon “Shot move ate th enly ons there, Aunt arin, ht stone tan be pred nepcto fthe MLC forte coer tar eons an ep they coker npn fe P25 Weare no a pon fo apt ving sins many Pose nose weel appear ote non. Tee incre atl tthe denon of Move te conon (62 exyening the MLC, wie ce ena) Sos tes (onmand nd euitune sdocnen ae}. (2) ea aso ape onl eee etna operation Move free Ker Bel 0K ‘A egiiateoferation” sone satisfying (SD), Cuegect an Trantomatons ” Before proceeding et us eview the stats ofthe superringviodion (20 inthe light of economy coniéatens Soppese tht the cevaion| D with th inital numeration N has reached sage. The reference Set within which ative economy is evasted is deters by (Ni the set RON.) of convergent extensions ofthe denvation N= wing what rerains oF N. ALE, the operation OP y blocked if OF yield 8 ‘nore scomomicl derivation in RON. 3). CConsderaons of economy ase a stage of the derivation only if ‘there is convergent extesson, But inthe case of), ther is none. The aber is not with he inl mamraton N: there isa conernen er ‘ation that ake itferent path fom N eng to (8) with inserted in matrix suber postion, (8) itsexms hat Jobe wast ft LP) Superssing thm 8) i ot bara by ecnomy coiderations that re {tthe outcome in favor of (83), Reese (0) snot a stage onthe way to x convergent derivation ata, Unless th shortest move requeent is port ofthe definition of Move ther wil bea converge drtation Thom 0, namaly, the ome that nvavessperrising. Bat things wok ‘vt as desired if the MLC is part ofthe efiation of Move, a prefered for other rezone.” ‘As swell nown, the seperasing violation far more seer than the Whesand violation invohing arguments and there are mony related obems tat have been the ope of much iveiation.® The conch Sons ete shed no further ight on then. 456 AtracyMne “The formation ofthe MLC i ore nature eer the ope ion of movement as “atacion” instead of thinking of a ating to Target K. fet us dhink of K 3 atisting the closest appropriate «.™ We ‘fin Atroct Fin terms of the condition (4, incerporating the MLC sed Las Resor (urdersood 38 (51), (88) K atracts Fi F isthe closes feature hat can enter into checking elation with ube of K. EK aturacts F, then a merges with K and enters its checking domain, ere isthe miimat element icing FFF] tet llows conerence FFF alone i the operation i covert. Te operation form the ain eo. al Caner For expository puro, Hil sometimes ase the aie erminlogy ofthe theory of movement (ore, eas, though assizrig tha he Correct interpeaton fio tes of aration, refering othe operation ener ax Atiaet Mose “The notion “equiitance™ dial in chapter 3 caries over to Atact. F without een change, hough wecan simplify and generalize i to ruil eayes no coniered there, Let us consider the mater sep by ‘Sep oynning withthe ene nation Th chapter 3 me comidced sveral instances of the stare (85) (01 of eater 3. edited as requied inthe preset amework), ‘he rac of ¥ chs aoined to X to form [Y-X} 6 oxP Spec, an Specy are both inthe nina domain of the chain CHL (Verband re therefore eid fom 2 = ZP or within 2P. Move can therfore rns to age ether Sper, oe Spey. which are equal close tora, Reormulating the novom of equance in tera of Area We ‘sy tha Spey, being inthe sie minimal domsin a Spey. does not ‘prevent the category X°(= (%, (%, ¥P}]} om eating to Spe Bat st 6) nly a ci ct acter pos a ataces to the higher east. shipping Spey ot by sutton ssn (5) bu by adjunction, ether adjunction to X° of ead udjunction te IY-X} The case dd not arise in ehpter 3, butt does nom. parity ‘ith rad wo feature sng (hence al eovert movement), We want 10 ren the notin feesnes to nla his pos Let us review the baie notion of dora ae minimal domain of 2 ein are (ection 312) as pref 10 what wil be » considerable “Smplifestion The notions wee dein fo heads (elter alone oF hea ing chains). We now extend the ro fetes ns we, Real dat we bave ‘modifi them igh se aot 47 (Cezonee and Trafomations 29 ‘Suppose mis feature or an X* extegery, and CH isthe chin (aor | (he tit chainya Then (66) 2. Mana) he sas saxralpoetion int 1, The domain 8(CH) of CH ithe at of eaten inca in ‘Max(a) tht are distinc fom ad do ot contain ‘¢ The mina! deme Min(@(CH) of CH i the stale sabe K ‘F8(CH) such tat for any ye &(CH, some eK rtennly omit ‘Recall tha domain an mona dann ate underscd dervtionly fot represetatiosly. They are define “onsen fo als foreach CH the point of lca insertion for CH = a and when CH is formed by moveren for aomivial CH, “The domain (=) and the minimal domain Min((a) of ate as ened for CH = a, nom bein evant “Terng to “closeness,” we are concerned with the maxim peoetion HP headed by H with adjied form H™" (the new projection of Hy heading the coin CH = (1) ‘nim! domi of CH, ‘ynay bean X°eatepory oa fate In effet the minimal domain of CH determines» “asighborbood of HT that can be ignored when we ask whether a fenure Pi tracted by HP; within he neighborhood of H isnot coer wo HP than, Not thatthe neihbortiod is deters nly hy that is aime consent of "not by @ mere deeply embed sub his ‘eset, or virtually all categorie wl be egisan wth Hat LF ae | Weatsing Ths sve wll dsolve ler cn, x0 it nad notte anlar ‘The difiiton incorporates egidance in the former sme and __Asighforvardly extends ito the ese of junction, We wi ein we= on 410 tat the notions “dasenes and “eget” canbe Further Sinplies. Uncartes remain about 3 mrodeel proton of 1 wih ‘more than one 7 adjoins. I wil leave the problems unset for the cent they wl eredun!corsideraby x we pecced In the light of this more prscped approach to the theory of mane ea, Fetus return tothe phenomenon of successive ey. tat, fasing ofthe ead ofa chain CH ~ (1 form 2 rew chain CH (@P) A number of probes ar i tht ip 0 permisibe pees a0 ‘omper ‘Suppose that «isan argument that sass successively to forms (64), repeated ee (48) we ate aly (fy to be aed fc, build separ} Here the traces re ies in constitution toe, but the four dente! ‘ements ae dint terms, postonlly singled (ae dicuion of (1) and (15). Some tecical questions remsin open. This, When we raise © th eo-constiuentB) to target K, forming the chain CH (af and then rise a apn to wet L, forming the chain CH (Fi do we take F10 be the Wace inthe postion of ¢ ora of CH? fn the more prec version, 6o we take CH 10 be (4 L), (eK) oF (as. BO. Soppone the later, which tral, partir ise cesve yeti ising is necessary inorder to remove all —Interpretable Tears of (0 that the trae im te iil position wl then ave all sch fentres deleted). We therefore ase tha (8), he element in 1, ris to potion io fom the cain CH, of (9), then rae aga o form CH, hen agin vor) CH, (9) 4. CH, wt) B= tt) = Crest) more prec terns, tof (Pa) i and, 8 re ote bd plane] (0 ipl the eo-cnsituents theme and won But problem aes: only CH is» legitimate LF objet, with the ~tewrpetable Cese feature cininaed from 1. The ober two cits oat the Cain Condition, so the dertation thos ersh "The problem has ern scene for years, lon with oer coo- ‘ering seceiveyete movement to A-psiea, i whch medial ks right be expeced to have properties of aunct moveent (ee section 3.2). Verous propecls Pave boo pu forth It war sored ny haps "thatthe chains formed by suexesive- qe movement hecome 2 singe “inked chain” Tp chapter 3 we asumed that a single Form Chan operation yields 3 miinembered hain, fut that propel dos 2 ci into the curent famewerk, and the motivation has rely ds ppenred wih the ein of he theory ef moverent to incorporate the tc Tn the present framework, dhe natural peoposl isto eliinate the nine CH, aad CH, leaving nly he wel-ermed chain CH. Th Cvgnis ed Trnsomatont et esl would be acicwed fhe traces formed by sine ofthe head of chain ore ini t LF. Why mig dat be the ease? Tithe phonologeal component, ites delete. We have found no 13> son to extend that convention tothe N+ compotion, and indeed fannot were we odo se -pstons would be ingle st LF and rent chains would wolaie the Chain Condon (analogous considera tien old for other chairs). But we en extend the convention ply stipulating that rang of «beading the chan CH ~ (ay) deletes the trace formed by this operation that, marke it invisible tL, Sep pose we do 50. ALLE, then, ll hit is “een” the chain CH, whch SSisies the Chain Condon Can the deeed tres erase under the delaion erasure principle (52? We know that they cant fly eas: they are terms nd terme cach crise (ee (5). Bot the itermediate deed traces not enter into inerpettion. Therefore the economy condition (528), which erases d= eed even etues where posi, allows erasure of fra etures of the intermediate traces i someting remains. The phonolepcl Testes do not remain they have been sipped away by Spel-Out. But inthe ‘eof am argument, semantic atures remain. Thess re not sublet to the operations of checking thor) tncuding (52), which ae restric 0 formal antes. Therefore, «formal festreFof ap intermatace ‘ofan argument may erase, and inde must erate posible, We thre fore cancude that formal Furs ofthe intermediate trace of A-nane- ent erase, We can ow infrmay think of the eto cha so product ng the ines of chapter 1. with een” Filing in the dls of his outing, we havea theory of succesine- | otic movement that fs inte the broader framework. itemneite treo ate invisible at LF; the only cain subjeced eo inerretaion& the pir) 2 he gst piso fri nc in the postion of Jena insrtion ~ wich for comveriene I leon to ell he hte pasion, Sorowing the te fom EST. i an argtment chin, formal Tenures ofintermesite traces are easea® We derive the peer (8, (90 The termine tac oF an argument cannot be arated hee. oe not preven attraction ofan camer that te-comnnnds ‘The argument etends to wees of Asmotement general. Ths, she ad of wich a chain eu rey ase, bat the properties ofthe tae "if bythe operation wil depend on the fate composition of 2 x copter 1 x fs semamie entre, a formal features of # ere i = i a pore ele, sole formal ete [D) ean, though i is dled (Gabe at LF) ‘Lanne dian mst therfore be sich Ua he trace of a raised ex sive can vever be atacted improperly or bar attraction required for ‘convergence. That is ied the exte. The only relevant consition ‘ould be (1) which an expletive Has asd, leving race (71 theve seem [ce some books on he ble} “The Case and features of ak must tise 1 mati shough 1 dose ths postion. The problem i sereome if Taksrelevant fe tures that features th ean ener ot a checking flan with man [cleso, mats 1 atzacs the features of the ssc ook a regu But we already know thatthe trace ofthe expleve lack such fetus Ins ony formal entre its eatery [D), which sileant. he EPP ving aleady been satisfied by the expletive el. Ths, furthermore the only Kind of corsruton in wich the problem of atracting ‘tive trace could arse We therefore cone thatthe trace of 25 ‘xpetive doce not emer nt the operation Atte Mov: mobile fand canner bar raising. Once again. strict observance of minimalist assumptions yields the correct aay of Fats. without redundancy or ‘other imperfection ‘We have rete attention to interme Iaces of aunt sis But the etion “intenedte trie” imprecke Further questions arse she thet of a argument chain i asd 9 am A postion cs (92) 2 wat id John se beta [was sen) {© (uss) wha thre i the Foor Inet cases he race (= wha) beads an argument Ahan andi rid foeter by wiznoverea. In (924) features of # (ich could head 9 nontealargunent ein. as in (92) must sil be ecesible for Case ‘hehing wed associate ing, raped, These wre not wat we think of ntatvly a“itermadiat rcs” of ccssve-qytic nove Dt ‘te computational system does ne make the nui diction ules ‘modified te do so). We therefore have to ash what happens tothe f= res ofthe traces i these consratione: ar thir formal atures de ‘eed and erased as inte eae of sccesiveye aking? Congo ond Tarra x One posiility it sharpen the notion inert tae” to ex dude thes cass, bua mot attentive one 10 extend te Scssion To them, 150, best of questions arse. Thus in (2s), ot the move com lx strctre wha di Jon expect ¢ 10 Be seen € in with 1 heads nontrivial chain, ean the Case festure F of 1 raise covery For Cav ‘checking, or mast the whphease have passed though [Spee Ast] ‘verily? the former hen F mst not hase bee eed, or penny ‘een deleted, when wheoiement took place A convention i then reed requiring erasure of F svoughout the ary of chains contain Fs0 that no ~Interpeable feuture rere fn the operate postions ‘estions of vome potent interest also arise about the postion of ‘This ine of easing suggests 0 narow modicaton ofthe preoding account of feature deletion sider rover otal entre of sae eed (hence ers) they are nox peesary Tor the formation of kei maze LF objets that sntaty FT, There ar two Kinds of cts at ‘once here: argument chains satiying the Chain Ceaton and ‘opertorarable constructions with the varubl heading an argue doin, When Achans are formed, no Formal features in the ete po ‘ion are ncessory: the argunest en i wel Formed withou the Bt taken w-overent or se ether frm of operator resin eke ple, te trace let behind heads am arguent chain nd ist ave the fl ‘plement of fetus Interretable featres required for ierpt- tion ofthe anpurent et LF and —Inerpretae fates that have ot yt been checked (thers, the Case fest is reer chock, retin ing inthe operator and the derivation erases). We eet, hen ht In Armoverent the formal features of the trace ae dled ad erase, but in whmoverent (and other operator movement), these ears “The eles discussion i nae. As fo (52a, the reve forme ion permis the Case Fexare F inthe argument chain Headed 110 ‘ise covey for Case checking, which now dee nd eras tin bth Pesitons ofthe ha F.1,)formed by the operation fond in the oper or). (62) fs ou these way. Thee ate a vary of other eases (0 ‘onside. Uw leave de mater here, pening clowr soy. The general ‘ea then, shoud be that formal Fetes of a trace ae deleted and ‘rated i thy are unrecesay, and that some veri of (0) hos lor Inses oF Asmovement generally sot oper {eal seems natural o expect (90) t extend 10 (3) 7° (93) Trace iamobie “The operation Attn Move can “se” oly the he of ein, noite second of str member. Though i 0 fred he war me ge ‘al princi hat tre lo eat be tpt, 0 that we hae (38), ‘wth te quaifestions aeady noted. (4 Only the head ofa chain CH entersinto the operation ‘Are Move. 1 (5) nok, we see question tha was et umes fn te ease of Vesting: do the Teturs ofthe objet Obj din to the Bead of the Veehain oF to is wee? Suppose, sh, that Ving i overt, 848 French type guages. Do the feture FE(OR) adj tothe wee of IN, Ag. 3 copy ofthe rated V-Agr complex oo the ¥ complex of sich Vi par? The ater mse theese (4 comet, 1 wil sume: We wll tur o supporting evidence in seton 410 ‘Summing. minima assomptios fad 10 Gosshing Eke) the conditions (95) om Artacy Move. Where CH i (possibly ta) can aed bye, (65) can ais, leaving hese 1,2 copy of 1, Formal ena ohe trae of Amoverant are dled ad raed The head oF CH can aac o be atacted by K, bu aces cannot trict ad theirfentrercan he attracted ony der narrow conditions reviewed (an let peril op) AA prot is post, once again, by such contrutions (46), which appear in prerasing form a (9). (96) Ne sc Ip C1 “Theres good eidene thay esommands into the inital cause CL ‘Suppose that y= Aw and CL = thy oie on, 0 hat the perising ssracure i (7), eng ey se hin to tke Fb ar rasing. {97 Nd) se 0 Bite they ike Stn] “Then Condition C violation resus iy (= hin) thes Job a ted ‘ent It follows tary must also eeommand they ‘Wh then, does Vn (96) atric the subjet they of Cater than which commands it an apparent Relasvized Meaty lation? tego ed Tesomatone 30s In (9) seo bas two item arguments: PP and CL On present a sumptions, hat requires an analysis sa Larson hel wit seem ai ng to the ight web van subsequem operations ying (8) Gite Structure fT omit "ON oN é& iY Since PP isthe optional argument and Cl the obligatory one in (6) {stil that C1 is he complement YP and PP the apsiier XP, which {lds he observed order act.” ‘When ser rn oan 1 it fn the cain CH = em) FP isin the minal domain of CH, but this docs ot suis to place FP ‘hin the neighberbood ef ¥ that ean be gneve when we a whether ‘hey (81's lose enowgh o TP tobe atraced by TP. eso, bere othing his adjoined to Hat the point when dey ries.” Terre, tains closer 1P and has Fetes tat can etter ino checking Felton with (eg, its Dare. We expec hen tht they shoul ct ase in (7, conaty to at In some ngage the facts generally acord with these expectations. In French, far exazple, easing ss bared the counterpart 10 (97). tess Ps active, which aes preseraby leaving ace” (95) 4 Hean semble & Mae avoir nen) Jean sens (0 Marie wave talent bean ui seme, air du tale ean oir seme have talent “Jean sae to hero have ale |The results are predicted. Mori is loser to IP than the embedded subject Jeot in the position of (98) sti therfore bars raking. The (Cae of Jeo i not choked an erased, 50 the derivation ezshes. 1a soe coer (9%) the race ofthe sive eno be teed by (95), Therefore, ras Ings peri and he devon converges. I PP as such structures could be raised by A-movereat(topicaliza- Soo, whmovement i would tae the suture (10). (400) veer According 1 the pine (95), the lass should be a in BH). The ‘ence appears £0 be pts end somewhat obscure, however. The Stas of te English construction til reais explained along with anyother eld questions (96) recess herent Case and le internally 0 the con: struction sem 10 in teas of prepets of som, Could there bea \erb SEEM. ike sem except tha it sles DP instead of PP and asigns ‘role But no Care? We wou then be able to derive the stone (QOH, whieh would yet the ouput (1022-2) among others. (001) TYSEEM fe DP 4 (102) 2 Bil SEEMS 1 bats ning) by there SEEMS semeone iat it raining {there SEEM someone Jobs to be ely tht i ett Presumably, no sich verb as SEEM can ext, perp berase of some Jerse between faerent Case ad Grol of ima argument for ‘which we have ne natural expression and which, omy Kaowldee reais wren and trgly wneplored. There are many sila weston ‘Conder the ECM consructioe in (103). (103) a Texpacte fier to tea book onthe shel] 1. Fespeced [thereto ser [rt be Book en he she) In exer chapters se assured that the expletive sere rss (0 (Spee, Aer} to have Hs ecutive Case checked by the vanstve verb ‘spect and thatthe soe book then aj 10 The eased exe ‘We have now ce tht view The amine rases nto tere Bo IO ‘he matin veal element VO (= espe, Al) (Bt to is Spee FFthermore, the expe it pire and therfore cannot aie 10 [Spex Agro] at all" though ican rae every subject, sin N10 abst the EPP. Matin Vb atic an appropriate feature af he asso ‘ate book (Case eu), th bring te closet F that cx ete il0 ‘checking retin with Vb (L3H) is acsconted for alot same ies. Cleo Troms x ‘Auraction ofthe asocate i thee cas rally violates the HMC “The satus of this condom resins cole. Ie always ft rather nes into the theories of testy of movement, regan spel ss surpions. The enparcal motivation forthe HMC fala not eniely ‘dear, a there seem tobe fay ear couneresmples tos. Can the FHMC fal within he framework jst oti? That ses doubfal. ‘We can reset attention to adjnion of on X element o another X? element. Suppose we could establish hat Bate only the est ‘sch a meaning: 2 fenureF of 1 the closet fate tat eset» ‘heckng relation with x sublael off. ut that mot enough to yk the HIMC. Specialy theres meting 1 prevent rom siping see ead that ees no fetes tobe checked. Consider. for exarpt the eonstion (108, oy ye v Here = V, and y= Vy (perhaps causative contin). Suppor ‘onl Valls incorporation. Dots tirat a, olting the HMC? a= Vjso that Vy reguves erel afi, then we might argue hat ', atuac the slower verb Vy, erring atrtion of But thet seems implausible s€ 2 N and Vy reqies noun incorperaton. Iti ea snouph co sald further conden, im eft spuling the HMC: ut ‘ened song arguments for tat undesirable conch, The pen, ‘night be overcome if roun incorporation involves a elation, not | Stracturatccnfigration Se Dx Seal snd Willams 1989. Buse sl secs o bar unwanted cases tng hed singin oer cases The ‘uation resins nation. Given te cena ole of Feature checking inthe minimalist approach, ‘we want robe nr abou ut what it means Category Karat Fenly ‘CF enters ino a checking elation wth a sblael of K, But eater scusson sired an important probe: wlat happens inthe case of feature imate wt Caper “Two rated gestion aie (105) 4. tna confpwation fr feature checking re features checked i they fait matt? by Wa Fntoreir nt chaked dos the ernation crash ‘Seppo, for example, that DP has nosmominaive Case (accsatie, ral or something es) ad as been eased eo (Spec, TT finite, where ‘owintive Can i asiped. Then question (1088) aks which of (16a) (106 corte. (406) The features checked in this configurations that Case ean no longer be szssd by the computation nd the derivation srs by vite of featur incompaiit: bhi ote choking configuration a all 0 that Case tit sect. {As for (105), suppose that ver such as wan takes coneplement shat ray or may nt base Cas (want 2 Bok, want ~t ean. The (1036) ‘sks whether the verb mst have wo dst lic] entries, or may have ‘only oe, he Case feature spl nt being signed i there no be. “The tact sumption of earier work has beep that the snsver (© (905 is pesitve Ge. (106), not (HO) and that the answer 10 (136) Jenepnve. In he useav expe, DP isin Spe, T] but has ace sve or nll Case then it eznnet raise to check his still unmatched Fear: sid verb ke wan cam hve a sige exis! ent, withthe ‘Case fete ssighed or tot ‘We tow koow thatthe estmption abou (1S) was comet. W fextre that ~ Imerpreable is not checked, the derivation crashes [As dncsed all formal entre of ade that erate checking Soins (thas ll se for faures apart fom category) ae —Inerpreable (Cave ad features of verbs and adv, sre ature) Thee {oral st be chased, ad the answer to (10S) definitely positive “Assuring that suppose we lo replace (103) by its negation (esen- aly iin De 1908) (107) Featrescannot be cocked under feature miszach, “Then in the example of Cat oie, DP wil be ale ro mee further © receive Cove, bt the drain wil rah beeae fails 10 asin is CCaseassiging enue. And verbs with optional bes will have stint lene eis ote with and one without the Case ature Coeesies 208 Tranortons 5 Consertions of concept natarapes ee t favor (107 over the aernaive in part teen oF the answer to (056) We wil ree 19 ‘sone Further sopport for the same eoneusion. “To show that all unwanted consvusions are bared by the comin (107 combined withthe postive ansier to (1086), ese to ‘ya broad range of possibile: thus, de subject might have ts rorer acute Case checked in the checking demain of V aise fo 7, while the (improper) nominative Case fete ofthe oh resto Tor checking nominative Case. Both abject and T ight ate to igh pottons for Case chackieg. Opions pear, and is no sine ‘rater to show hata are Nocked for ate or anther easn, Kis tere fore resonable to refine (107) wo bat thse posit acres the board {eave smiesich cans the devation. The compen) ofthe com fulton required to determine wether derivation wil converge i (hereto sarpy reduced, 2 desidertum gute geerly. Further, 45 We wl een section 10 thi epi eset) n empiri) grounds {55 we refine he framework in accord withthe Minas Progam Tl hereto strengthen (17) 0108). (45) Mismatch of fentores cancels the derivation, A configurton wi mientcha fetus isnot &litinale yrnetic ‘beet We distinguish mamatch from nonnatctus the Cae fen tare accusative mismatches F”= [sien noriatve but ao atch Foa1 of e casing stitial, which assigns no Case V have kt he ‘seton “teh” somewhat impeeie pending cesr ans But to ‘ootent is clear enough for present purposes thus, the eatepvil estare Dot DP matches the D-tature of features rach i they ae wee Sea and son. [Notice that cancelation of a dervation ender mismatch shoul! be itingisbed fom sonconsersence. The ater penis a deen con Sergent derivation to be constructed, posable, But the pot fe Herat to bar sheratives. A cance derivation therfore fl nt "he eaepory of convergent dations i ha bloke any less opi ‘ection; mismatch cannot be evaded by viction of Procatinat or ‘tbr devies. Ifthe optimal derivation rete @mstch we ae Peed te pursue nonoperative Seppose, for example, that «eves of eppiations of Merge fis formed a verb phase wth DP, ae specifier and DP, as compere, Trang sccoaive and nominative Cae, respectively, We wil ce that a0 caper ‘he optial derivation from that poi leads 10 snatch. Since mis ‘mah is euislent to coavergenc fom an econen-theoeic past of Se, we eantot const est optimal deiaton from tat might onveres, with the then whject Baring accusative Case andthe the ‘matic objet nominative Case. Te nterpesation i otate’ on pure) concen grout sharply redass computational compe. Again Concept an¢ enpies consdertions comers. ‘We now dsngush beeen checking conan ad chesking ‘elton Suppose hat K sacs F which aes to foe (HIRI. (2K): fore = Fhe operaen i cove, and includes wha require for convergence the operation s vert. Each fest of FFTF] elon, yin inte checking domain ofeach sublael fof K "We row sy that, UG») Feature F of FFF isin checking cama with and is ina checking rea wit if, farhermore. F and f ach, IF ad fil 19 mateh 9 problem arses if ehey mismatch (confit, ‘he dervtion canceled with an illegitimate objet Tn the lstrve example, DP has nonnominatve Case and bs teen rue! to [Spa] he Cost future [CF] of DP isn a chesking configuration with tke Case enue of T But notin a checking relation tvth it: Hence the target TP di not attract [CF] becase no heck ‘lation is enabled I does, however, ata the categoria fest (D) of DP. to sus) the EPP. But the [CF] sina mismatching checking ‘ontgcation wif andthe derivation scanceled ‘Suppase that fe the Case auiring fete of Ko and have the nghecked Case features Fy and Fy Uespectwly, and P, but ot Fy ‘matches Suppose that 9s cer to K than 2: Does 8 prevent K from racing 2? Te Case feature Fy of doe ot do it ot atacted by K ant is therfore no more ewan than sone soni ature of P Suppo: howener, tht has sce oir tre Ft con ee toa ‘hocking acon wth subi of K, Then Baise by K. wih Cinnot "se" the more remote ment 2. A mismatching reson Js ‘ented and the dervatien cance cannot Beant. ‘Conse agin the supetaiing enampte (0) (~ yt) Seems [ha sos tl fn) The ermedite DP iis able to check the Dfestre bt ro the Case feature f mats he more emote DP Job ean hos fi features off. But eaneot "se beyond 0 only that inerening lene an rae, ean the derivation to erah. Had Jot been po titled to ruse, the derivation wold wrongly comer Congo aed Tenors m ‘The definition (82) ofthe MLC torte has te some sharper form (110) a5 a consequence ofthe refine of the concent “ehecng (10) Atv Link Condon K auracts only there 0 closer to K thane ste tht Kanraeis ‘The defition ofthe operation Aniracy/Move incorporates his property. ‘jeking the MLC in what sems to be the requted for. 1 Kea thie ere a he fina version of the MLC, apt from a referent ofthe ‘otion “doseness”in scion 4.10, ‘Consequeees rail 20d mer ferter ove ‘The notions of chocking shery have been defined only for Ata Move, but we have seen tht that ay be too narrow. Thus, checking domain are established by Merpe it (6) repeated ere a (11), Simple expletive contractions sich a (113), (IUD a(t wondes fy whether Q fhe fe ye) '. (Ewonden fe lif Ol he te ye} (112) theres a book om he ble To these cases wheter if cd sere romain in ther base poston, bat Sty the stone features of, 1 The operations ee Goel atogoos to rising of @ whphase, I, or DP to Boome [Spe I, checking the ‘stong fears in he seme way. “These cases of merge il al unr checking theory i we extend the tion “lose” to dsc sytaci objects wand K eng the catporia ‘eature CF ofthe head H(a) tobe cle to K. K witracts then PCF ers it 2 checking elation wa sblsbel of K. in which ease ‘eco Spe, HI) or ajins wo HK) for it ole projection, ements have already adjined to H(K)) The eesut a8 tended for (UI) and (12) bats 20 broad eevee. Thos i woud cover eget ‘fae sobet Subj with V satin any version of the VPinterna ube Iypothesis ond would ths fore the subj to ave the acastine Ce ad object agreement feats ofthe tarive main verb of the VP (er Jp raise the light ver rf which Subj i the pein. ‘The required distnsion can he made in various ways, One i to extend Attrac Mowe only to merger of senarguments. keine sty to tte conception of AtracyMote the formal eapresion of the fentrecheching rope of rural language That his ene pas an apne Arguments (and operator phrases constacied fom them) srt the ‘Geain Condition nontrily: an areunes is 2 nous chain CH (@., where has aed for feature eeking an sin &6-poston. to onirast the elements whether, fab thre of (11) and (112) 80 ot Uf the Chai Condition a "A second apprach is to sllow the caegorial feature CF of oto be airacted by « tie objet K ony it enters ite w checking retin ‘wih song subiabe OF K. The faonales that Merge creates chech- Ing domin nly when ver insertion is Fred, voding Procrsiaie ithe cas of Move hone a spacial case or both Mere and Movt. ill adopt the fest option, though without sitong fssons at this point. Some wil he sugested im section 4.10 though they rly on Frecic chowes about mechanisms tha, while eaonahle, are ot well ‘rounded. The cote f therfore highly etative 46 Movement and 0-Theary Under any apvouch tat takes Atmact/Move to be diven by merpho- logical features —whetber Move F Mone a, and Greed. or sre ater ‘ian there sould be wo meraction between Beary and the theory fof movenent. oles arent formal fests inthe leant sense yD ‘ally they are asp i the internal domain, notte checking domain. Sa they ifr fom he eres that ene into the theory of movement fn numerous ether respects. The condluson is immediate in Hale and eer’ (1938) configrational appecach 10 6 theory. implicit in some ‘thes (hough rejected theories hat pert percolation, rrensision, nd cher operations on entre). Le assur it tobe vl i fundarentl reps, Osher vial compleenary o she theory of checking oct ered in part asa descriptive peer tion in the Chun Condon: nth eit CH= (2-0, iv role and a enters ino a checking eation. Furthermore oly 2,3" ‘spn «Gol, so that onl the base pion i “elated,” able to as Sinn or receive ale ese moe 7) The peopets of a Flos fom {ast Resort aeverent. Consider the proper of ty, tht is the fa that movement takes place from a poi at i elated 10 one that fot fora argument, fiom # Boron to a non postn; for ahead (or pretieae rom x postion sm which a cle is asigned x one i wich tis ot Congo and Tendon, a0 Wit regard to asigment of oes. the conclusion is maul in Hale and Keyser theory. A Oro sisal nw certain strectra conten ‘on: ass that ole only im the sme hat i the head of tht ‘uiguation (hough the provers of or its zeroes! projection ‘ight mare. Suppose Brass, forming the ein CH = (= The trce 1 remains sn the structural configuration that determines le | and an therfore function as a ole assign; but the chin CH isnot al so cametastign a ole. Ins raised postion, 28 1 bas internal fonmal features aw Case asiger or binder. But in a eoafguationl theory of Creations, Ines itl sense to think ofthe ead ofa chain 25 wining a ole ‘With regard 1 eit of oles, siilrreascring apples. Wares to a Boston Th. Forming the chain CHa) the argument th runt tara role CH, aot a. BCH is no in any configuration, ‘is poten argument that can recane a Oo. Other conditions too are ‘oat ender earierasumptions or offers ke them. bit ¥ wil ot ‘gall out the problems fete. ‘We conclude, than, that x raised element cabot wei or asin #0 tole. eatenese “base rope.” complementary to feature check ing, which sa property of movement. Mote accurate, Olaedeess 8 property of the postion of merge and its (very let configuration, ‘The same coriderstions bar raing-o-oje. even ithe object Spe ‘ier na Larsonan she. We thus derive the PRP pict ha there no raising roa postion — acta in «somewhat stronger form, ice relatedness generally property of “base postions” "Thus, DP cannot ras to (Spc. VP] to esume an otherwise unas tigneé (le. Thre can be no words HIT of BELIEVE sharing the © structure of hic and ete bot cing Case Rares, with John eng ‘in (113) 0 pick wp the roe then moving om to [Spex] 0 eck ‘Case and agreement features. 13) a, etn fp THT A) 1, John yp [BELIEVE [tobe imetigen Surely no song feature of the tiger is checked by raising to the [Spee HIT positon, so overt rag ane in at no checking elo Son i essa” The only pasbiiy is aret easing to (Spe. 1) “Toe resoking snr Jn HET and John BELIEVES to he eller are therefore deviant lacking the extemal argument required by the rb | | me chaper¢ “The deviance of (113) sheds some lipht onthe question lft ope bout Bote assignent in confgution headed by untied 2 (0 ‘hich has adoind, yielding the comple (Ba see note BA Can ‘thick head chin, roriate i Ole asigament inthe configure ton heed by Suppore 0, Inthe tepiate Jon {HUT 1 the sg ‘ment chin CH ~ Jal) stisfis the Crain Condition: Jor reacves (Casein Spc. and reeves a Boe wikia he VP. HIT then ries © 0 that Jl falls into its checking domain. Jo ean reve the sub {zc Oo i the coniuration Spe. headed by the comple HIT M) Foe by V-taising wo I then al eepets ure satised ur the expres si shouldbe wel fred with Jr bearing a double ce. Assn that he expesuon te deviant, HIT cannot conte to airing = Perle when i acne 40 The principle shat Olstednes i 4 “bse prope.” retried to configumtiens of lec inserion, Mas 10 be Snderstcd in an ster orm, ‘Wat is she nature ofthe deviance in the desnation of (1137 More eerily, what the sats of olson of he 6 Cerin. whether i Involves an wnasigned Grote or an anpuent lacking & Gro? No fele- ‘et performance are wale, Bu economy conditions might pro ‘ide evidence. Thus if the derivation converges. i could Bleck ethers der ewonomsy' conditions, 50 the deviance woul have to Bea case of ‘nonconvergence i that eau is empirically wrong. “To alusate these posi, suppose that the theory of economy cous that shonerderiatons beck longer ones. The assupton ser flausibe*” Kitaher (194) observes thatthe coniion ean be invoked to deduce Precastinate, if we asure that appication of phonlepcl res counts in determing length of derivation: Wace letion is me ‘uied en if an operation is ener, so covert opertions are alvays Prefered. i hey yield convergence The economy condition resis Independent conferaion in Seton 410. Now consider any simple sentence 2. (1E4a) with strcture (118), before Spelt Out (014) a. Jobo tes it >. Goha VP) “There i derivation in which J is sere enim FSpec ot raining fom VP. Hf she VPAstermal subject hypoess scored, 3850 far asad hat derivation mat crash, ori wil Bock te less coo ‘omic! derivation with rag a Jol to [Spe the two beg with Caegees and Teanformaons as ‘te sume numeration, bu the dasited one has anes sep. Al fra Features re chected™ nsenion of Jo sates te EPP and he ee tures are checked as or by eres. The only dela of the umd eivation iin @ theory: the ergument Joby nck Oo, and like does ‘ot sig it external Ook ier ofthese properties eonsities 3 ‘olaton of F, the derivation erases, nd ke problem dppeace. ‘The “sonst desiation” condton tho, el tat 2 tan ofthe ‘©-Cresin causes the derivation to crh, by fie to say FI. Sul ‘pense question wheter the prober hereof am argues 0 tezske a Ge or the fale of O-signng head to sign te Oo ‘or both depen reasons forthe fs, les, re given in so 49: a argument without a Ge volts Fl, easing the derivation 10 ‘ash, Note that these gestions about veation of te ®-Crieion have 10 relation to the concn that Gree i mt «formal property that permits Las Resort movement as Cake and agreement fate 0. “The form of the VPsneral subject hypothesis has 20 far been Ie vague. We hive, however. generally assumed a version of the Hale: Keyser approach to theory that has eran consoqunes forthe By pothei In puticlar if» verb has several ternal pues he ‘ne have to postaate a Larsoian sl ei (58) oF (11S whet wie Sat verb ta which V overly ees. aoe The inerral arguments oxupy the poss of specifier and compte iment of V. Accordingly, the exter argument ctnnot be tower than ‘Spec. sb Tit is Spex. as 1 will assume, then the »-VP coats ‘atlon can be taken to expres the causative oF agetive fle ofthe ex ‘eral argument. ft would be natural to extend the same reasoning 10 ‘eanstve wer cosritions perl asigning thm a double VP ‘ore asin (11) he apent cle Feng wnéerstood as the interpretation asia the» VP confication, A Vc constrtion ethane ‘anal not V. Th concusion gin further spor if such ont ‘ions may assign a partalysdenyeratic semantic roe (See Maran 1986) riotranstveConergaie) vers ae iden trnsives, Hale a6 captc and Keyser sues. then only unscatves ching agents would be ‘Spe VPstructares. The ual, which sntural inthe present rame- ork, also hor weleome expel eonsequences, as we wil wee will ssumestin what Flows thes ers, ale of tranive vert asign an eternal cle cul he interpreta spy meaningless. The external role prop ‘ty ofthe »-VP cnguraton, ad «specie bering thie role is her Tere near ptf the combpraton: «trative verb signs ‘neal ele by defison The question ofthe nature ofthe devnce {nH would therfore no aie there are no sich objects. The enly Famaining question is whsther flue of an argent to reste & oe isa ese of nonconvergence or deviant convergence. Presumahy wis the ‘under the natal eterpeation of FI--the conlsion we oi reach insction 48, “The sp of ths ona equim that there be no AgeP intervening ween the Hight verb yard its YP complement in (HS), contrary to ‘whats amumed inthe most extensive detelopmcnt of the complex VP nays of unites (ce Kote) 1993, 1995). The fue takes sir fm for under refinement introduced in stion 410. We wl turn co the whole ne nections 49 a0 410 447, Properties of he Transfermational Component 474 Why Mowe We haves fr considered two operations, Meigs and Move, cach with vo cates, siteution and adjuneton. The opertion Mere is inezap hl onthe weakest interftce conditions bu why sould the compat ‘Soa syste Cyn human fags ot be resid oi Pll iis ot The most casual nspecion of exp conlns even tha es ‘cumasonly epee “cslaced” ram the postion i which he interpret tin they reve roterise represented atthe LF incerace®” Thee i sang controversy about the bai facts. The only questions are. whe are the mechanisms of displacement. and why do they exist? As for ‘ir nature, on minimalist assumptions we want pabing more tha 20 indication st LF of the postion im which the displaced item is iter: ett tht chains are logtimate objets at LF Since chins ae nt Intrdued by slim fom the lxcon oF by Merge thre must Be nother operation o Frm then: the operation Atrac Move. Conger ond Tenors a The second qustion~—why do natural angus have sich device? ares inthe ety days of generative grammar. Specutins about i i- ‘voked consideration of tnguage ws: Fitton of parsing on certain assumptions he separation of theme theme strane rom base-dter. ‘mined semantic (8) restos, 04 20 om"* Sush spectationsimole “extrencus" condions of the kind sta er, ctons imposed ‘on Gy. by the ways it iterate wih extra syste. That is where we ‘would hope the source of “imperfections” woeld ie, om misinalst assumptions ‘Our concer Here isto determi how spare an acount of the oper tion Atract/Move the facts of age low. The bert posible el that bare oupt conditions ate sified im optimal way “Thi question was wSoond for of the flr to tele the tension ‘betwuen descrpive ard explanatory adegusy,slonsde the sep hat Jed to Xshar theory in the 19. The ens conser was to show tha the operation Move art indopecent of another tori the variety of suctural condsions for tranfrmationa rls. These efforts were ‘motivated by the usual dua concerns the empl demands posed by ‘he problems of descriptive and explnatoy edeguay. a the concep {ual demands of simpliaty tnd raturahess Proposal msivted by these cones inevitably rie the new lading problem th epaces the ‘oft show that resscing the sours of Knguisie ery preserves (and we hope, even enhanees)desriptve ndequaey while explanation ‘eepens. The ests have met with 4 good del of sues.” though ‘minal assumptions woul! ld ws texpeet moe 447.2 Departure rm th Hest Care “The properties that motivate AuracUMove have to be expured some how in the theory of human lguage but we would Hike co show, if posible, het no further depart fem minima assumptions Fe ‘gird: Thais the probe that comes nto sherper fous a explanatory ‘Meguaey begs to ake space onthe revereh agen. ‘Conse Est ihe independeneeof Move a rom chic of . Although this curemly seems a femorable uppesition,# has been oeeesry to disiogush varour Hinde of movement” XPovement frm movement aed (among XP) A-movement from A-moverent. Various Kinds of “inproper movement” have been red out im various ways (eg. head rang tan A-postionfoimed by casing to Spee. One ‘oat isto climinete ary sich distinctions, demonstrating om generel a oars rounds thatthe “wong hinds” of movement erath—not an e35y probe fen thoughts by ov substantial eed Some of the general consrnnts intedaced to reduce the richness of deserptve appartus abo had problema aspects, An example End's influent score presersnefypotiet (SPH) for sobs tie operations. AS has been stesied purely by Jan Kose, the ‘SPH itoducts av unwanted redundancy thal te trat of mowener!| is somehow “there” before the operation thts place: that observsion Provides one motive for nondertaona theories hat consteut chins by computation om LF (or Structure representations, The minimalist “nproach overcomes the rlunarey by clinizaing the SPH: with D- Structure gone. ts orformuabl st consequences derived we ope to ‘ow bythe greal proper of Merge abd Artact Move IW has wo been propaed that something like the SPH hokls of junction: bar levels ate matched within adjunctons, This extended SPH introduces no redundancy and ie notated hy the Minima Program though we would Bke to dodo it from more elementary easiderstions. “The descriptive fats a ot entiely cea, but they might Be just dsuited: only YP ean jain to XP and only Y° can ajc to X though covert operations may fave the apparent tof adjining YP to nonmunimal X° (eq WP-djuetion to eatin Ve wich we no take o te adjrtion of fora ares of the rnin verb to V,). We ‘hen have several problems to dal with (016) & Why does the SPH hold a at? 1. Why there dilerence before a afte Spel Ot apparent oltng the (optima) usr assumption on Cy? Why doe the tpt K projet ater edjunction? Under the feature mover theory we are now assuming the questons ae nirrower sll A cover raising i adjnction of eatures so question (6b) dsoles and (1) arises only for overt adjunction In that case ichas already been answered for adjunction to X° and to preeting XP Furmermere. under the interpretation of Last Resort and caching do main that we are now assuming (2 (51) and note 7) itis welt thet ‘junction to nonin XP possible a al tus assure 80 never thes and se what problons arse. ‘What remain, then, is question {116} and eves eases of question (116 for vert moveren:adjurtion of YPto XP thats ‘Categories and Transformations a (119) there apc ucian adjunct Gonser question (1163). Inthe ese of overt movement, the asner stay le im pat in peopeics ofthe morphological component. At Spl Out. the structure aeady formed estes Morphooyy. 4 spstem that presumably deals only wath wordike elerts, uhich we may take to be X°-that ether an tem H sled from the lexicon or such an item wth elements adjoin to ta form H™.Suppone we esse the property (18), (018) Momphotogy Gas ont wth X° eatgores and their tues The morphological comporent gives no output (So the derivation sashes) presented with an clement hat isnot an X° o fest Or this natural assumption, the largest phrases entering Morphclogy are X's und if some layer uot appenrs within an X", the deation| ‘rashes Quston (116) renin. then. only for suction of wonmaninal (ster a feature oF» category) to oneal XP- XP with at leat & ead HIXP) an a complement. Both cass are hard nd the cond on (119) on checking comin note (U19)« ajoine to nonin K not othe checking desi of a), so, the operation we are tying to innate cannot take place, We will see that thee i easo to belive tat (119) hold. Then question (16a) i ally aeswered and we are Kl th the thee cass (147) | gestion (168) A casts of (117) are hare by Last Resort iPehecking domains ae ‘onsrued 2 in (19) A numberof spesaeases ae barred or nde ent eon Summarizing, the aspromety of potion fer movement has solid round it only the target that can prog, whether movement ss itwon or adjunction. The ony bvious pcble thet the constants appear tobe to sang, arvieg YP anton to XPetit ‘Wie lave founds number af sons te queson the sats of YP a {ueson o XP. Thee ar otters. Suppose we sper th Barres the ‘perton already found, thus alling agjenction of to nonminmal| Kasia (20, 0 caper my cy Suppose that projets so that L = aed the category formed is a ‘We have to determine what isthe head of he chain fered by the ‘njuetion operation: tia, or the twosegnent category [a0] The Inter chore fred owt by the een conn (17) But the for rer aves ut with 2 category [i] that bas no interpretation at LE, ‘oating FL Ic cannot be, then, thatthe eit element @ projets if ‘equites an interpretation. The same problem would have srsen, his tun for, ba! we fake the Head ofthe chin tobe a Siar Guts ‘ons aris about the age projets Once ean, adjurtion 0 00> ‘ninial XP leads to complications however we onset structs erro, sg further Questions about isa, ‘Sil assuming sch adjunction eb posible, conser the special ase of sefatachmen sn (121). y “Thus suppor we have the VP rad th oak and we ain tthe ead rea, fring the two segment eatgory ead [rhe Hoot. Under the ‘eed interpretation of (121), with he are prot, we ave formed the object (12D. where 7 The target VP-= [rc a the ook) Coming farther analy). (412) {Grea read, (ead 7) ‘Seppo, however that we Bad projected the sdjunet V (read) in (120, viding (123. (ayy Categories and Tenors x But this 100 is am informal representation of (122), jest as (120) ‘hough the mended iterpretadons dfn (121) we hve proeted the ‘acpi (123) the adjunet Further, he ater interpretation should beara ‘The sae qucstion would rite if he cperation were mabtittion, not adjunction. Suppose, for exemple, that we aie the bed N of NP to [Spe NJ NP nonmisial (econ, or there is no epertion to ds: ‘uss. Then we construct the sane ermal objet Wher we think of NP or Spec s projecting. ‘We might conchae tho this eae he ht result, wth such ambi ‘my inexprtad as a rased derivation. Then such operations of sel tachment (whether djuscen or substation) ate buted on grounds independent ofthe nnd dine id Let as tur now 10th case of rising af V in = complex verb sue tore, as in (115, repeated bere. (29x ve For several eeasonsslrendy dscused, the operon cannot hase 1a ‘ted VP iter as ajrction or as substitution. It most be, then, that Xe is kota proton ofthe used verb V but rather a very pase sic fom VP. as we have 5 far assumed Thus, V raises to ah ab ‘ea flied positon eruped by the light verb» that has hee ste fom the lexicon nd eds own projection. 2. V adn ov. form inal, Vo the» position i not "eeated” by Ube easing operation. For Independent eidece, ee Coline and Tidirsson 1984) The operation is permis i he tape vs a ph see raging verbal af. le eden, these contusions are required bythe properties of Gaheory iscued earher”* We haves fr sdeuepped »peablem that ais in the as of on ‘ary ead adjunction. Take, K to be X's in (10), with sing 0 ‘arpet K, which projects, forming L~ ((B(K), HK) (a K)]. Since K projets i animal Th, both inal and ine tha Imus of 1 as wel (e, he ease of elie raking), then CH satisfies the niformty condition (17). Bat supp 6 nonmexinal, a common seats Py caper inthe cae of Vrasing to orto V. Then vader sara ierpettion. (i) vated CH ot Teitinse objet at LF, and the dation ‘rather ‘That obviously the wrong result We therefore assume tat xt {EF wordie elements are “immune” to te alge that determines pie srctue aos, neat 9 (129), (025) ALLE, Xie submited to independent wont interpretation where WE ignots principe of Cy. witia X8°* WH is covert an Toe of Morphology. except that we expe it to operat composition: aly uke Merpbalogy. hs asumpacn thatthe N= LF mapping ‘form trouphout Seppe tha (12) ~~ fy. [2K} farmed by aunetion with K ecjeting, #0 that y= CH(K), H(K)). So fr. there are Iwo ways in ich L could ave been formed: by stct merger of, K (without ‘movement o 8) ring of. forming the coin CHa then messing with Kr ether case we form te structure L wth dhe thre eis 2. Ky {CBs of thee entegoy thats “visible” at the ire. whee i rus coe some interpretation. satising FI. The adjune poss n0 probley It hes CH. i eceives he interpretation asec with he trace postion i ade by ste meg. presumably pedi FR (og an adverbial adjunct tow verb). Bet there only Ont role ft LF for K and 1 (problem already mentioned), Note tht the late CHEK), H(K)) no tem ence reeves no interpretation. TEL is nonmaninsl the prob ecbiated by (125) under «vata interpretation of WL This should sfc to account for, sa), nous incor portion to verb, or ter sonporton to eawsatnes or light verbs What is interpret us covert incorporation of X™" 10 w pe he! permite sraitforwaray by the Move F thor Suppose Lit sonminimal again, we suspend the bases 10 this questionable ee of sdqncion to iavestate te problems further. We tow have sho tr, Land K, but onl) one LF eo. The sretue ‘woul! be peruse fK ache Bole. as io covert adjunction to a ‘opletive (@ ate that we now assume doesnot ex). The ny ther psbiy i ha he adel i eed at EF, king just K, When Soul ss take lace? ‘ne cat ir when the trace of sucess moverent othe ‘ype that permits iter tae decton sy, along the es skehnd Cotegrics and Treaties mm insecton 4.1—for example, hsmovement to [Spee CF with ners ate adjunction, an (120. (126) which pictures of John's brother di he expect ha fou would ey Another is is fl rouastrsction at LF eliminating he adjuneter- ‘ely, es stractre of he ype (127 inept enya the race 20 (XP ap th {An example would be “srambling” inerpected by recostrction, 2t fc to fe eiformy the case by Sato (1988, and subssyuent wor) Silay t woud Fellow thet such eonstractions as (128) mast be Con on Cvications (ander de slevantinterpretatce). at we predict & ‘iene in status beeen (129) and (12) te later escaping the vn tin beats the hed of the chain is ota ade (128) a, meet Johan Engand he dos expe tht Fill by piers of Soe, e does’ expect hat il buy (29) pictures of John's the, he never expect that you wouK! bey ‘The concsions are pluses fit apposition, howgh we enter fae ioto a roeass of dieu snd partly unsolved questions (ee Bars 19 Frain 186, Leen TAR air went a pe 3 or sense dsesson) {© via rina asin see shoul be he ony posi iis for arction (130) & word formation bs serantly vacuous target ©. delaion of aunt (trace dltion, fl reconstrction) ‘Agent fom (1300, there should te no adjunction to Oct phase G teoteasigner eran argument a pects or the XP of which ii ‘dicated Since (14) relevant to sit merger, the options for ‘he current coumerpat to “base njuration” are even narower, We wl ‘onside aljoinad adverb further nsxtion 47.52" Adivnetion therefore remains an option under natural sini = omptions, but 9 very Bnd ene with peal properties, The primary ‘nd perhaps ony cases acuneton 10 XP, fate othe oper ‘on soe) an X° Adjuncion of VP to XP dos not ely nt is ‘sal approach and fale st ll hes ey ested ene ea cape 4473 XP-Adjuction andthe Artec of Lge Theory ‘Adjuetion of YP 10 XP fas fad central place in transformation erative grammar fom is xg. Tati nent: t provides the west obvious examples of "dplaceent” of phrase from the pos ‘ios in whi they are inespreed. Bu as tbeorekal understanding has elie, to diet pats can be cerned. One pth ocsed OF the (operations foumuloed as Move NP and Mose wh ater Move oad Aflect ce pow Attract F if what precedes conrect. For thes, XP djuetion as marge, itrodueed pinay for theorynteal et sons elated tothe ECP dod so on. Anois path soupht fo undead ‘uch eperatons as extrapostion,rightnode raking. VP-adjntion, scrambling, and whatever “reariangemen ate involved in forming sich exressions as (31, {3 Took tot of pictures ou oF te ate yesterday of my iron snd thir ends it wo parts ofthe ase rok ot. epaated hy an eee ar of picnves ha should not be & phrase a al. and wo pars of the ary hea o for of [irre of DP separated by an cement of rach wider scope: Ac the i paths inezesingly diver it Domes mere and more reasonable 10 suppoe thst he proceso structures they aes Jo tot belo togebersfrterore the ter catego appear ob beter ‘pencous. AS already olnerved, the Former path doesnot ely ior Porte even lmenary sructares of many kids se notes 2, 93 “The diside has been sharpened her by ogy ino engages of the som thar Baker (195) describes in terme’ of his "poy Pvametr, with he sya in lrge part wordntereal and erpinets ‘nuached us adjuncts associated with intemal elem=nt. One might c00 Jesu that such propenies of UG appear in sor manner in nguapes for which the principe not of a fundamental ature. Consider, 52. ‘crumbling in Japanese, which sors to share sone of those properties, the scrambled element being kindof adjunct, external tothe major ‘jac truce, ascent with anita postion that deerme the semantic interpretation (kone ite oblpatory “resonmrection”) Re Ine sues ee caret being invetgated in wey diflringlangaess (Gee Barbs 198), We wil earn othe mater in ction 410, Ie carly warsformational grammar a diiotion was senses male ‘erwaen “syst” cus and eter. Inereasingy, the distinction seems Cuegore and Teno ms to be quit seat she core computationl properties we have Been consi ing dilerrarkealy in aructer fom many eter operations ofthe language faculty, and it may bea mistake to ry to integrate them thin the same anew of peter. The problems rtd to XP-adjuntion| tre perhaps «casein poi: they may tot really belo to he syste we sre discusing eve a we koxp oslo th ir of the ro cous ut ‘une, the one thats eocerad with Last Resort movement dives by feature checking within the N-=) coraputation, Hi within this core omponsnt of the language that we fd the sing proper high Fohted by minimalist guidelines. W seems ocean reasonable to die Lingus hs component ofthe language fell “Thee speciation gain mse the nartover tecnica qustion whether ‘tacking domain include YP adjired to XP, bth nonminial The ‘most dict empirical metivation fr defining checking donuts to allow ths case was aversion of Kaye's (1989) theory of partis arecment ‘Ge sections 232,32). Teka mption hain pose snd unc ‘atv, tr objet esses hough the (Sec, Ag] poston A-moveren), seckingapreenen wih the participle, and then raises to subj, diver ‘by Cise and in eperetoe movement, the objet apne tothe Ap ‘movement, agia checking apeenent inthe chucking domain of Agr thon raising timately to (Spec, CP, driven bythe operate Feature, Ta ‘articular, Kayne fond dake difeens mace wit he two kinds ‘tpi apreement. Domingue Sporiche and Philp Bronigan have observed thatthe p= ‘ator movement case ie prblemate because of Such ogenance ‘movement consiectios as French (32) 032) te tease [out + Tage Lge it fou Pete at the leer that ke hat Said that Pete tom as [envoyé sl) “the lene tht esi hat Pere set io? Raising of the operator fom # to (pertape with intermediate sep) tnd then to Spee, CP] i Tegnate sucessive-cyehe A-mevernet should yield patil ogrerent with ait in the Righer ou, incor recy. This sugasts that agreement (hence, presmbly, Case as wel) shouldbe rented to the sper postion, with “Kong mowenent™ tered by the same prin that bar objet using toa remote pos tio: the we various posible acount, depending on othr aspects of the theory of mnverent, The dae ilrences need by Kay rei ‘neeained however. there constuion ae coret we rest the checking domain of | to postions incaded in (ate than conse ie) Maa) incline the A-postions adjoin 10 a we Have assured. In bi. we ace the principe 119). which hs wel eonseguences independently a aeady Seed, 7A. Other Improprctien Let us return 10 the problem of improper moveren. We want to show thatthe wide varity of sch cass are ext on pitied ground Some fl to ple. fr exsrple, such standard eases x proper xt ing of Join fom, 101, 0 mate subject a8 (133), evens adjunction of Joho IPs petites (033) ohn islet Left ave) “Tee complement of Wega equites PRO (H & Hepa ro leave. 80 that (oull Case is esignd 1 the suet of he infin. Since er not have nul Case, the derivation eanctd by msmate e(108))” Conse ass ofthe type (134, ith fain 0 TP, again puting ‘side eervatons about such proces (034) ohn ees ha i tld hat ‘Wie do ot want to permit the iermedit (offending) tae two ee unlike what happens in (126). To diction suggests a dierent p- reach to iserediate race eto: peraps iis. reflex the pres of reorstrction,andestod in minimalst tems as in chapter 3. Toe ase ssumprin bre is tet tere so procs of eonsrecton: ether the phenomenon is coxseguerc of the formation of operator variable csetruions drsen by Fly a process that may fe sometens test) eave punt ofthe race copy of the moved cement inset at LF, dene ‘only is operon prt, The rcoratructon proses woul! then be stricted 10 the spacial eas OFA overent hails operators “That ecortucion should be barred ny A-ha ths pau om concrpualgrouods It has some empacl support as wall. Under the teleant interpretation 135} can only be understood as a Condition ‘olatoe, hough under tecostrution the ween shoud be obviate ‘with ha interpreted i the positon oft, eeomwanded by me: a8 We have see, the later coords Categorie ar Tatfamons » (135) John expected fh seers to me. 130 Benin ‘Tat the asd subj doesnot ily eeonstutishow 8s wll hy he _gussapentivestatvs commonly conferred in “surface suet” postion Gg for PRO In (138). (1360 [PRO wo appear tobe ineigen] is ard han ove might think Other reasons 1 qustion whether thee i eonsteuction in Athans arise fem consideration of “lowering eet the Kind fist dicused by Robert Ay. (037) a Gtscems that everyone en these et 1 Tenpectedfeveryone not to Be thre yt) © exeryoneseams otto be thee [Negtion cn have wide scope ove the quai in (137), an sams in (137) but cot (13% 1 0, that ideas tat thee 0 son Sraction wo the ace position in (137) ‘The quanifer interactions eoulé res fhm adjnction ofthe matin _quntier othe ler IP (ecommanding the trace of anne ad yl ing a Welormed structure ithe trace of Guat loweing is dled ong the ines of May's original prepost. But reogsrction in he ‘cain doesnot ae plas, 0 appear. Some other eats of improper movemeat are cniated within the Iramework outined Here, such as XP-rovereat psig through or ad- Joining to a pure ¥° posivon, the trace then dling. The sas of seraing might be recorsidered nd the (apport) disition in tus betwen sch stress (120) and (129) swell. The general tie eis a comprehensive review ‘So far we hve (Eta) Rept the mins assumpson tat the ‘mpottional procedure Cy i uot frm ito LF any disinction {fore and afer Spel Out ia refer of oer acts. 1 ha said Fle $0 fr about the “extension contion” of ete 3, which gunna {Rsty. The condton i empirically motioned for substinion before Spu-Out by relsiviznd minal eles and others. and it docs not old after Spel-Out ifthe Case spreemen theory of the iin p> roach score. aso cannot Hed tty fr adjunction which om only (ad inthe case of head agunction, always targets an element hina gee prjetion. 1 would be destable to sow ha thee cone ‘sequences ate deduce, rot til! nm caper Wish ear to Merge, theres nothing wo say sates the extension ‘condition Tor elementary reasonrsleady dace (se below (1) ‘Questions arise oly in connection with Altec/Move. The opertion targets K.rnsing ao avin to K or to Boome the apeifer of KK projecting i citer cae, K may be u substructure of soe structure L cay formed, Tati ceary option it the cover cement bu tot allowed Feely before SpellOut-at 3 rel of ote conditions, we hope to shove ‘With regard to over jes, there ae several cases o consider. One ‘ypeinistrate by such standard exampin a (38). (238) Sub was a pieture o taken by Bil “This i» Condon on Estrsction Domain (CED) voltion in Hute’s (1942) seas if pasive precedes wh oovement Buti derivable with no ‘olaton (incorrectly) i he operations spp incountereyec err, With Tassre following wh movement ln thi eave eye indice dey taystrngth Unies pase precedes wh overeat, he deivation ican tle by vclation of srength of he EPP). Independent, economy ‘ondtions might sake the devant distin terween the competing ‘erations. Pasive is tbe sine in both; wh movement i “anger” in the iat one iam obvious sense, objet Bang mare “remete™ from [Spex CP than subject in ters of number of XPs cried. The dine ‘ion might be capture bya proper theoy of economy of derivation — ugh the Sues nontal, e part, because we ae invoking Hee 3 “elobal” notion of cimomy of the sort we have songht to avoid. Sock prablans would be avedad in the approach proposed by Kawashina Sod Kitahara and by Groat, bil sethed in section 441 (Se alo ‘Cts 19980, Kitaara 1988, 1983), We wil aur in tion 410 10 the posi oF exhaling the counteeyelie operation under ast ie Aerpetaton ofthe pile 5) “Another las of cases re the seltvzed minimality contractions for ‘which the standard account Ruz’ (1980) ‘These fal into thre eat ove () head mower (the HMC); 2) Asmoverent (8) A-moverent ‘As dicused the ats ofthe HIMC unclear. Tn each cas We Rave 40 tuations orale oot () skipping sealed ited postion: (H) counter ‘ile operations (e, movement hat kip 2 “pores” psiton that is later filed) Station at unde he MLC, which s incorporated to the definition of Mone/ Attract. Ae for (catego (1) at peo: tend inerton i feted necesianly by pure mens, which satisfies the Cutis an Traformtons @ ‘tension condition. The problem abo cannot arse for strong fers. ust renin terete el sid eon ey be, ther that thee no net impose the enesion conten of chaper 3 on vert operations. Furthermore, athe the phone tor the covert component ean sees telecon, for reaus lead iscusd. The Morphology module indie allows variation before and afte SpelOut, as do strength of festures sed such peupetes of language asthe PF conditions on mover that inde genta i pining. All of hse conditions ref properties of he A-P intrace I seems possible to maintain the coeclusion thet the computational system qi ufo fom No LF, i hat no pre verve post Spe Ont ie lnetion stipulated. And perp we can sppreach the opine! cove son that bare ouput conditions ar if wel os pons 475 Adjunct and Sete We have so far said ie about such trucos a (19), with an ave ‘al adjoin to XP co form the wo-sement extaoey [XP, XP. ro ject rm X 3) xe, ae, x ‘We may, peas, assme that th contruction is bared if XP bas & semantic see a LF (ee (140) and note 7) sey, iF XP is predicate {AP or VP) asin the v™ structure (1408) undying (1400) sf the analysis of transite consiactionsin (U5) isonet (10) 2 som patentee Oty +, *Yoln reads fen (books, oh cildren) Such stroctues as (138) could have teen deste ether by Merge or ‘by Move. The later possibility cen peehape be ruled oot in. prio ives sce to have no morpboleieal propre that rete XP. junction (venti possible, «dubious ea, now. The expr ‘evidens alk suggests hat adverb do not form cis by XPraduction| 0 caster (seep 8). Thus am adver in pre ‘fied ie fom seme lower positon “The only option, then, is Merge. The question is whuher and how ‘hse adjunction” fn the EST sense) operates above the kvl of word formation. We bave spcuat tht i is hated sf XP is Semanal scien (Ll) Irepetive of he status ofthe word Sequences, the ‘Sseucsure assigned in (14a) sto permite ‘Ava frat apqrormation, at leant adverbial cannot be adjoin by Merge wo plrass that ae Beate rgurens or prea) 10. they ‘can te "bae-adoined” only to X' orto ples headed by v or fun ‘sonal categories, Adjnetion (0X by merger does nt confit with the ‘conclusion that X is inisible tO Cy: a he pent of adjunction, the target an XP no" ‘och constrictions a (140) have played considerable roe in bn isc theory since Emonds's (1978) sts of ferences between V: Fava noneesing lnuaes (Freeh. Ens). The bse pence slong (1400) asst by (1414-¥) both well formed in French). (44) a. So reuse hisehieen “John es fen bok A proposal someties ented is that Vass from the undeyne trite (LQ) to Form (a), Bu such easing is Barre in C11 or (Case reasons; serusatve Case i assigned to Backs by ved under a adjceney condition of the kind proposed by Stowel (1981), French dit ferrin the adncency property orin sme the Ws. “Apart fom the fac th the source coesrbton (1403) is barred if the discon cari accra the general approach i peoblematc on ‘iia grounds This framework so natural plc forthe cod tion of adnceney Furthermore, sf Case i assigned by raising (0 8 (Spec, ied poston, a we asm, adjacency should Be iret in fany event. Isls weer hy the ver sould aise at alin (181), oF Whereis ising to Teer that either the peoosed analysis is wre. ce theres problem Fr the minimal framework In fat, the empl grounds for the analy are dubious. Consider sch adverbial pases wt every da Flas nigh, which eat spa the postion af fen in (20 (042) Hot evry day muds 10 is children [Nevers we il ind the parade of (141) ition cxnnot be interpreted a Caegoet and Taras mm (13) a, Jon reads every day to his len be Moh eas everyday books “Furthermore, sna phenomene appear when ring not an option a asin aah, (048) a. John mate a decison Vos nigh, suddenly) to ase town '. John fe an bition ast igh sudden) oleae own ere the adverbial may have ani sop, tat tt within the infin case can appear betwee the N ead ad ts complement thoush the N eangot have raisin the manner ner deamon The romps incate tat lean strtures of the frm (149), we eat ‘onclde from the mati cope of Adv that has razed cu of thee ates phrase 45) [fy Ade tO VI.) In general tis dou that ring has anything todo withthe le vant pradions. ‘The plenomens sugest a Larsopign sluion, Suppose that we ex- ‘he (140) from the paradigm ently suming that ven apres a some higher poston and this doesnot exempiy (19) wth XP = VP. ‘The sreture undying (141) and (14) (140k tha iste exe i 2 is ascot under assumptions about trasiive ves sireties adoped| ‘ali (edscusson of (115), oe ‘ere VP, and Vj are proecons ofthe list verb » VPs and V5 te Proctor of read and ren es oad o Suppose tha in (14) the aerial fle. Then i= 1 he eke don hare m0 problem. Bu if f= hooks the driaion wil ees ook east rast Spe; Aas to have its Case cocked base there ar two close intervening erent: the subject fof and often The Retuivized Minsabty violations are not overeone by. Vranng 10 shen mrasing © Aes: the combined! operations ave © cose fo Age than = ok. Recall hat hooks xonot rise 10 [Spe or seasons already dicted Under this ants th Be facts olow with espn assumptions ‘There isa Case solution, butt doesnot involve jaan. The problem ‘oF opona raising eleninated along wth thove suggested by (149) and can, ‘Questions remain about eter ater, mong ths: Whit she his {or the French-Engsh distinction. and it somehow reduce to over Vrasing? Why 6o the wi-vaans ofthe veri n question tehave lie aunts, not argument" What about CED eft in the ease of such adjuncts as Adj in (147), which might he in complement postion ‘te analysis peerales? (18M they frend the book [gy ae we eave such questions withou any useful comment Another ls of questions fas vo do wth the scape of advetils in [ECM consrutions. Consider the rentene in (18). (148) «1 et ore. implore) ny stems every yea (tat hey shoud {et thir papers in ote 0 werk he) , Twould peter for my stadt every year to (gt thee papers in ‘ontime, work hard) {eve my tadents very ear to (work hard hae gotten {bei paper non tine) Unnlr the Lancnian anal just outned, eve yew shoul have ari cope in (Li), ard (14) should have the marginatus of| (140). ith embeded seope if imerpetble tal, The difrenss se to be nthe expected diction, though they sre perhaye not harp ‘one might like We wonkd incidentally expect the dition tobe ob sted ina Verasing language such as Tels, as apears to be the ca (Dianne Jona, personal communion) “The sume analysis predicts that trix wope shoud be surge st estin (149) and (150). (089) a Cheah ote ak oe ids) 1. ve prove in repented] oe Cotgo and Tearfomations mo (150) ve pred ihm repeatedly tobe it] “The cases of (149) come out about a expected, but sich examples 2s (150) hae been etd since Posts early raingtoobec wk (1974 co ‘strate matte scope. Weare left with sme wl abou Me proper Weaizatin ofthe data with extraneous factors removed ‘A plausible copcuson seems 10 me tht the scope ofthe embedded ‘semen i marrom,a y (148) (149), ad tha (150 ices the kind ‘of “rearrangement™ hat as been called “etraposion” i he pat ut ‘hat may not belong at al Within the Seaneweek of principe we are msidering: sx section 47.3. The wide scope interpretation sy shen Fal together ith such cates a (11) and (140)-(U4S), fn which overt ising 10 4 higher postion is hardy Hkly.!°° Ths 8 spout, Fever ‘Similar qualications may wel ein order with egtd vo rule adjner sete such a (1ST), partly if hey ako involve mu "ple renreagerent sn (151) 04 (131) (151) a. Jotn watched a documentary with great interest yesterday ‘ie i Boon ther doin watcha documentary yesterday that he eal enjoyed shout the French revotion bythe thor we met of the most Interesting novel that any of ur a ever ead ‘Whatever may te involved in such cases ii unlikely tha proliferation of shal is retevan. Eten if at ara sessed for maple ad nats, there ite ean to suppose that The verb raises epee fom deep in the structure rater. shell struc ig rdevan a al ‘he aditional phases might be supported by empty feds Below the nisin ver, which might Be ao more deeply embedded than in (16), aig beer understanding ofthis whole ange of onc, se Fre ‘mature to speculate. The sane kind of analyss seems appropriate for a wie ange of ery pies verbal structaes, wheter oe nt elements f the itera domain te NP arguments Consider sch cosrutions they Tote ary him, Here argr and to him ae within the internal domain of looked. thowgh thir eletve positon i uncar;sufce one maybe mile ing for reasons dicated. Seppe tat onthe araloay of sr 0 co ‘srotons (ee dcussion of (8), angry isthe complement. The st tore would then Be C2). mu taper un oe a a "oN Pee ere cena ramen eee er el ne oe ee eg (153) a this road was sce ven on Pts oud wa ven ecently on Preposion standing ofthe kind ilustatd i (15%) has varying de ers of accent, but example degrade rapidly whan an adr ‘eevenes as (15352 pecomenon that as ben wet to suppor the ‘en that V-P constructions sometimes "realy" sa new verb V. AD alysis slong the ines of (126) yield he same sus witout, Now ‘er addressing other fair qestions or example, why ate sch er {ences a which ea did John drive recently on also degrade, and why do we seein iiomatc interpretations in he “ean” css? 48 Orter [Nothing has ye een atid about ordering of ekersents Thee i no lee ‘vidence that order plye reat LF or inthe ecmputation fom N 10 EF Let ws ass that doesnt Then ordering i par ofthe phono leg eorponnt «proposal hat hasbeen put Forth eve the Yar is sarious forms 10, then it might take quite a diferent form withost Mlcting Gf Dinguage ui Invlved gresterexpresive dimensionality forgo senor manitesation at al. Te seems natu to suppose that oteringapoies 10 the eutpat of Morphology. assusag a Enea (empora, tosh) order othe ee Cees Tranfoations as mens it forms al of them XS tough not necessity exe items If ‘oret thse assumptions led further reson to seppoe th the no linear onder in the N+ LF computation, ssming tht it ht no we {othe ouput of Morphology. ‘The standard asmption has ben that order ix determin by the trad partner: languages are basicaly hentia (English) hea feat pans), with further reframens. Fuk hos peoposed that he Jad perancer provides an account of optional movement, which lr cvs is excluded under econemy condtlns apt fom he special ase ‘of equlyeconenaatertive derivations. He argues hat monet ‘tha mainsins the ordering ofthe bead parameter is "ees oer move= ent must be motivated by Greed (Last Resor. Thus, in fei Japanese, kfward movement Gerambling, passive) i eptonale wile in heal Engh, such operations must be motivated by feature sfecking and righwatd exrapestion i fc i Engh thoogh bared feJapanese Kayne (1953) has advanced a radial akeratve to thes > ‘umplion, proposing that erdr refets strc irarhy universally ty means of he Lear Corespoedenee Asem (LCA) wtih ttt ‘symietieecommand (ACC) impose near onring of terminal ee mens any category that earmot be willy oer by LCA ts tated Frees Keynes spetie formulation i follows tha thee isa univers ‘species headsorplement (SVO) oxder snd that specifiers ae in fect ‘adjuncts. A head-cemplement stctue, the, is necemarly at) XP, ich can te extended-cnely ee, on Kayes assumptions to @ twosegment XP. ‘The general ideas very mucin he spit ofthe Minimalist Program ‘nl consistent withthe speculation ta the eset eharctr of Cy is ‘ndpendent of the sensorimotor iteace Lt us eorsier ot might ‘be incorperated im the hate phrase sretire theory. That aot en sirely straightforward mutter, Because the bre theory aes mich oF he structure ofthe standard X-bar theory that plays & cro roe in Kayne’ smal Kayne offers two kinds of arguments for the LCA: conceptual a pial he ater ened in subsequent work ee panicuurty var 1993 and Kaype 1984) The coneetunt arguments show how cersin Stipulated proper of X-br theory can be derived fiom the LCA. The Empire arguments can largely be ered ener toa efron of the TLCA within abe tae theory, but the conceal ones ate problematic. a hope Firat tbe devnation ofthese properties res cuclly nt just onthe LCA, baton festurs of standard Xba theory that ae abandoned in the baz heey. Second, the conclusions re for Ue ost par immediate inthe bare theory without the LCA. ‘Let us atk Bons moi LCA eight beaded othe bare thor: ‘Tere is 00 etegry-teminal Gicion, Pee no header ds Linton and no serial conan on ceommand. Suppose we have ‘he sruture (15, which tthe hare theoy counterpart sever of he richer strcttes tht Kayne considers. ay YN. Hee Lister opi ce Jo LK ma be ete a pate ‘sear orasegnen ofboth > (par hc hacer allents) mp fag at Litt feet yaoi, ctr m orp fa anh er bal ‘aia ao minal y mite hearse Lea Soppot at Ka seat eeey and jet. sorrinan A pton AC hols (fm) and ps0) pale sid pt oul ad pon ibe peril phe Conplaeat of at ) reece a comp ng. He Sct oder spread caplet teh on fr nl Spine Spor th nod emia we ha branching with ot cote Lan Xe ner mesial or nia do 218 ‘Semi he ACC rane ae cig Sion hat Karner ab pot tt tsb vend fi wth cohen ACE ns eee Slop tk wasegen cae or Ther eno pia rob- ten br since Str rt on te pte eto a conned In athe LOA can be adopted inte ae cy with oe si tet conan Te mgneacregiy dive (rd be {Sit ero) he manned tout We dw Kaye te Comsonakot $0 or eth og ny he camp ssorecongn hana sgl Calgon and Tetrion, a» Let us etuen tothe case of L = mP wih the singe terina comple tment 7 both minimal and mastal Sine rether mor p asym tally ccommands the othr o ding is signed 01m he nen ‘rdring isnot otal and he struct nates the LCA, Tht eves to posses, Ether we wesken the LCA so that nonctleresings (ot ot “contradict” orderings) ave adobe under cetnn cones, for we coneuse that the derivation crate unis the srcture N= Temp} tas changed by the tine the LCA apples s0 that i nteral Structure relevant: perhaps N is consened by Mpg to “pho- ological wort” not subjes imerally o the LCA. suring that the LCA isan operation thar apis ser Morphology. ‘Consider the fist posit: is there a natural way te weaken the LCA? One obvious chose comes to mind there is no yarn fer the LCA to ecder an element that wil appear PF, fr expe, « mice, Suppose, then, thet we exempt aces (om the LCA, so thet (15) i leptimate fp has overtly raze, leaving & re that ex be “nore fy the LCA, The secon possi can he eld in eset the sme manner, by allowing the LCA to delete tae, Unde th ner pretation, the LCA may eliminate the elfending tac in (158 if has ‘aise. In shorn if the complement i a Snglerninal XP then it mus rie ‘very. EXD = DP, then its bead D is xc thr demonstrative oF pronominal, which attaches at higher pont (termina ther pene Aly. of by specie morphalopical prope." IF XP = NP, then N ‘ast incorporate to V (and we must sow that oh optus are Worked), itis, te, are tare Ds without complements, and noun incorporation mast be rested to “ooreteretal NPx” (as note by Hagt Bore) suming the quiseeretal ndeiat character ofa now pase 10 ‘bea property ofthe D head of DP. NP being kindof predate. Within DP. the N het of NP must rise to D (as srgued i a difrent manner by Longoted (9%4) "= We therfore expect to find two kinds of pronominal (simi, de ‘monszatne) element, spl oss that se morptilosely mathe a fies and mas elicz ad complex ones with nel rte. wich donot ce n French, fr example, the determiner D (le, lc) fd the complex eemeet hicne hens trish the simple clement 's again D. ard the complex one may even be dsconiouoes. a ae ‘each sin hat hows with determine ori (Andrew Carte, perioral ‘sommamicaton). A pheomenon tht may be elated fe noted by Esther on oper Toreego. In Spanish the Case marker decane omit in (155), but oti (1880). (055) cer dela plazanear he paz When de dleta in (1550), D = foc incorporate incre, stsying ‘he Case Fier tut that i posi in (1580) i he complex pono ‘ae not D bt a word with ihe rte, rom which ered (FD canes he cata Sine the aia propery example pronominal ize even if ‘hey are notin ial positon (ee. prvorinal objet that i speier| ‘na Larsonian shel} If foes ad more comples suture, then foes (tse) simple pronominale could chavs Hike compen pronominal If Englh-ype pronouns ae spe they foo man eikiz, thaw eal, not ring Io Tae in Romance (perhaps a2 telex of lack of overt V- fasing). The baer fo ach acres sf pcked wp i might Tow. Engl detereier sachs i ad Phat are presumably comple, ith the ial consonant representing D (asi te, tere, ete) and the residue ¢ Kind of adjective, pertans, Various comsequenes ae worth ‘ploring “Auouh appeal nt unsesonalle, the conclusions are very song his, every ehcbranching sete must end in 2 tac. on these 3 suis ‘What about erring of adjuncts and target? In Kayne’ theory. ad ets reas precede thal targets. Within the bare Ueoy. there ‘really principe contusion, 25 far ns Tea se. Ordering depend on exactly how the eore rations of phrase structure theory dinate and ‘Seommund, ae generalized o worsen apes, ‘Consider the imple case with w attached to K, which profes. (1) eo’ Suppose that K, fsx new catego he specie or complemen, S| that (186) = L (HR), K)}. Take dominate to bean ieee ce lation wth the wl interpretation. Then L dominates and K ifr sul, K, dominates an Ky ‘Seppe, however, that the operation was adjunction, forming the twosegment catepry [Kz] = [CHIK) HOR). (a, K)}. Are 2 and Ctgor aed Treforions a Ky dominated by the category (Kas Ku PP As for ecorand, et ws swine at a commands outside of thi category, thus thea ‘hain, e commands i trace, which post not bein Ky (in Head raising." But what about farther command relations inluing ‘thse within (156s? “The core intuition uadetying eeamemand is that ST) X ecommands fever Z that dinates X dominates ¥ and (h) X and ¥ are daconnetd Forcatepris, we take X and ¥ to be diconnened if K ¢ Y and nites omits the or. The wotons “Hensal” and “sconce (heee “ceommund”) coil be generale in various ways for segments “These felations ave rest 0 Jer, i the seve defined eatin the eas of (156). 100 K (= K,), and the tweserient eatery IK, Kl 1K; hat no independent state, Thee consis cotrport reasonably ‘well with the general conden tbat elements enter ino (he compat Hinal sym Cy they ae sible” a the interface. Ths, Ky say aston or eosive a semantic role, as may «(oF he chan it ends, which must meat the Chain Condition). But thee no “Our” roe et ver for Ky he women category wil be ierpreid a word by Mor ology stl WI (ce (125) K i an X°, and others lls de he arrow options Swed ee 1 that mach is comet, we concidethat i (956, TK, Kl dominates its lower sepent Ky. 56 Dat the Iter doesnot ccemmand anything ‘Gling, nt dominated by (Ka, K,} but only contained in i). “Turing ext to econsnand, how shuld We extend the notion ie onnected™ of (157) to adhe? Take adjunction to = nonmanimal ‘ra (16) in Kayne 1993, duced 0 ts bare counter) (158) ere q is ajoined to the head m to form the two-spmentextegory (mz) 8 nonmasimel X° projecting to and Reading the entegory L, ‘which ht abel oe Ri the complement of mand rts Bead and 5 (which may be compen) ie the complement of x Wit are the ‘ccomuted relation or he aur siete? SaaS toner Toman “The lowest Z tht deminates ¢ nd rm, 6 Le which also dominates loch. Therefore ¢ and mr] asymmticly command rand S. however we ere “dconncced” What are the ccommund rations vith fom 8 noed, mn, doesnot eeormand anahieg The ther Telaions depend on the iteration of "dsconnecied™ in (180) Kayne ieypets i as "X enchodes ¥." Then q (asymmetrically) © ‘commands msm kick dominates my, no hat precedes my: and in feral, an adjunct precedes the Heng ch is djing WH, Y are token to be connected” fp repent fone coma the othe. then [pesommands mt ot Pry sein pects, °° Hi Conneced™ requis sl frther dissociation of X, ¥—say hat neither isa segment of extegory that contains the etherhen no ordering | eerie for gam By the LCA T donot ey pinipid way to cheese among the varios options. Wm, isnot shea athe compen ext goe [08]. 0 that qian “xP for reasons aout vind, then g ecommards the consis of ‘my onder al inerpretations of "scone." and the aunt preades ‘he target (hehe i iirrally enrpex er 200, ‘Uc open, then, the awe of aunt ofa head 0 anther Yea, sari nding within words. Whether onder shouldbe fixed here de- ends on qustions abou inflectional morphology and weed formation {hat ae ather obscure id may have no general nse. ‘Somarizing, seems that Kaye's basi intuidon ean. be accom dated in srignfrward way i the Bre they ncuding te major “emprielconcisions, speibealy, the universal oder SVO and edjunt- target (atest for XP aunt). tthe Bae theory, the LCA pairs m0 ‘oppo fer concept arguments and therefore rests on she empircal ontequences. We fake the LCA to be a principe ofthe phonological omponent th spies to the asp of Meephelosy, oponaly ignor| ling ov deine traces. The speciferadjunet (AA) dstnetion & ri tained, slong with the posstity of matiple speiers of adjuncts though the onions for adjunction are very lini for ether reasons. "There are forher consequences with reard 1 etisztion and ther raters, whether cores o ot, donot know. 49. Espletns ond eonemy “The evidence reviewed so far ha ew o poste to fetonal t= ovis within IP for spl coestraons, Tense) an Ags. Oserense CF mata ty tte et intone Hae of is semantic bene nso rele ce ee a Wind te Chem nd prove pnn eve me tsetse ore EPMA xm! ny sca Sine ncn of tra dc end ps pun rn eg Sor ori he aio em oe apie Thee ee functor nyt on for ett pce sent Ferra ee nto co he tape ora nc nr, ony ln opps ‘Te it ep wo Nomis ft ce sree the ed Sp poo ps Mor es il ao Sti witha the of te pote Spr pone I ne frm whee No ial pss Dor 9 lage No Ato NH Tage Nom AVP fecal at haber on Are reno wih 0 ha ‘inden ie pon te ten cag “he siacure (5) thse yon oe cess (1ECs eae by ore Se Baugh 99) oes (994 ‘one ons snd Bop rent ee wh as ts fe aon pe Egat) store ans 60) Lay ee pit fy a stot yt Rowse VPI “The meaning isomething ke“ student pnt he hous’ or he neh ligiie bt unacceptable Engl counterpart (161). (ty here paint he hous stadt (who traveled al che way Fam India to dot) (10) the enliven Spee, Aw ped is he verbal ead of LP raisin stapes to T. which then rae o Ag ling the ef the subject «smd is ised Yo Spe a the objet the Bose £0 {Spec. Apt and VP conains only rac The pre-VP postion ofthe ‘objects motivated by placement of everils anc vegtion inthe over forms. The ual properties of expletive construcons hed: [Sy ‘he assacae of the expletive in (Spec. Agi. 6 nonspecific and deter ‘mines the numer of the verb inthe A psiton A Thee pected Postion are cccupied inthe overt for Cnc and agreement aye chechd fr the object onlin Sp. Ae se for the sujet er coer aning of Forma ntsc e As 30 rapier Ina TEC, Agi and T each have overt specifies, Hen, cach has a sien etere—in effect generation othe EDP 10 Aer and T inde eden. In introducing the EPP (sion 42.1 below (I). 1 noted erainambiitysbour the song featre at express iti ould be (0) Dfeatue, 2) a N-eature, oF) sera neminal feature ther or N. $6 far | have been ising the terminology (2). but neaaly among the thre choices, At th pont the choices begin to make a i Ferenc, s more cares nese. “The apcir of Ag he expletive 0 the strong Feature of Ag at least allows and may requ the etegoria ature [D). The speci of i wominl, bur there ae theory rerel reason, 10 which we wl to ‘nsecton 410, sogesting that it igh be NP rather than DP. cork (hat could tea factor in accounting forthe deitenes fl: the fae tat the associate in an expletive constuction, whee in Spee.) | lower in te class, nonmpeci, NP cathe thar DP (D assumed 10 ‘the locus of species) Farsharmore since ABfy and Ag ae the Sine cemen appearing sn two dient poston, Ay Bas 3 stone Ditcature then Ase a wel. We thus expt overt bec ang to fivor definiteapcife omits, whereas nonspecific NPS remain in ‘tu (by Procrastinte). That Seems the general tendon “The arly ofthe I stucares ithe same whether the objet is cused cor nor or even in passive an rating exper wih abject (Spee, Th {in (16(ransstion of eld examples From Jonas 199). (462) a, there ave [oy ome cake [y been tke fo the pasty] here sees ey smenne 16 Be inthe oom] Let refer 10 all of these as muiplesubjer constrains (MSCS) “wheter ranstve (TEC) oe noe DMSCé fall within the range of opsions already available, using ore uly the availa eng eats of funcional categories. These on ction slo provide ional support foc the coclsion cha she expletives in expletive smite consirctions lack Case or features. by ‘imple extension of ressceing already ese. MSCs ths fleur ev ‘ere hat he expletive spe, awe woul expect om genera rounds Tia close analy of Teed, several Faroese lets and Mila ‘Scandinavian, nas (1994) found NSCs to be contingent on overt Fasing, We know from langusges without overt Versi, Uke English that at lest one Functional category. T of Ag, can take an overt SP sifer i "sands alone” aot supperted by overt V-tasng.ASsming Cnegos sd Tonos » this cecyory to eT, Jones's generation sates hat Apr cannot have specifier unless superted by." "MSCs sucha (160) rae obese questions. (163) 8, Why do languages der with repur to MSCs, some allowing ther others no be, How are such structures permite by scosomy principe? ‘These questions presuppose some analysis of spe expletive conse tions sucha (1e4a-b). (U6!) «there artes man 1 there isa ook misting from the seit We hae found considerable evidence tht in sch constucsons the fo il features of the associate rie covery to mars checking Case and features ad factoring asf they Were in abject ether than ob Jc postion with regard to binding and control" Til comin to sume that account be cove in it ese Recurring to question (183). why do only ean langues permit MSCs" Jon's pneraizaion evant bere: overt MSC reauie overt Ving. Could tere be MSCs with only cover rating? Example (161) Suge that the posity should ot be imme discounted. AS ‘old, the senence i unacceprate (si (164, o some dese) though ‘otliile, and with other lexical choice the contraction ranges sn septal, s Kayne has esered (165) here ered the rom a man fot England there atthe stand new journal © there vised ws st night large group of people who traveled af the way from ein ‘ech contrctons have bee thought fo tesut frm an exraposton ‘perio depending in patton coneerstion of “hesvnes” bu in ou ‘esr ters there no vious source fr sich corsrctons aga fom MSCs. A possiblity that might be explored that they are in fat MSCs, wih the subject category appeasing overly atthe right Bound iy, perhaps the result ofa proces sn the phonological cerpponent that ‘ull be motivated by properiso theme the sretures, which Ip cally involve “sree” forms in sere mans. Prominence of ethene Fight require tha i be ata beusary: tothe igh sce the test tin socupit by the expletive subj. Telndie might cape this we chapter owdtion as reflex of is iteral.V-secon prope. which requis &netod for interpreting itera theses. The Fess restrictions in Engh may ret the sieaive characer ofthe expletive Mf spec Taom slo these ies prose tenbl, question (16%) may take u some tvhat tore comptes ont the MSC option may be more pera, bu ‘ith diferent manifestations depending on oer properties of the la fue ‘Question (16) lends ito ticket of complex and only pay ex plored sus. The fat thit MSCs alternate with noneagltiveconstec Sons i unproblematic dhe sieve aris from dire numeration, heyee ave not conprsble in terms of corey" But thee are ques tion abou some of the instaaion ofthe pestaared structures (166 (166) Exp Agr (Sob IF XP, Seppose we have he overt fon (167), We would expect histo Be the smarfstation of the wo dines MSCe U6) a0 (18D), (467 thre sess someone te Beis be ree (468). there ser Sereone [tbe inthe room 1, theese [yoy someone wo Be in the zoe] to (64s the mrs cane is a MSC with the sje someone oo ing matric (Spex) tn (16S) the eee clase i an MSC with oneone occupying embed (Spc, T] and the trace of there ocupying the higher Spee of the MSC. Both possibilies appear vo be litmate (oss 199 1668) (12H)? Bat we have to explain why both (68a) and (168) are eit outcomes f he sme rumeation. and svhy (163) is ered in Engi, (9) there sors 1 me, fen someone oben the room ‘We do not kave a direst conection. Principles of UG might bar (69) grey whe periting (18). Tentatively assuring that post bility to point to the reschion of te problem, we then ask why the stractrer strat in (UB) are permed while the one in (16) fred. We cannot appl tothe rmeration i hi ease, Pease 3 ‘the same inl rapes. "We also have to explain why (170) i grmited wih there rising From the positon a, where it ses the EPP ip the ebeied cause (470) ere secs [ft be emeonei the rom, Te problem bcos hare wien we a ECM consti fn hse embeded je es oy tw peson maop Fin (ib, whereas camel ein ‘sin (71) a Lexpete someone oe [in the som (0 have Been nat r) 1, Texpected [0 be [someone the rom (to have been at ebn) eit ve nnn woof pop hl eso top. Winn he aia ever weeps he ceape ose provided by invariant UG principles of economy, = * “Tre ques hve odo ath ooo hee ee weit pil shoud te Prone, whch nr cen monet te sect anteg coe detnuems sen ance erated (ad ft) io force comers Bond ta ‘vestons tear dre on base assumptions ofthe thors of eneoent ‘Leta besn by consierng the ECM cise, which contest with con tel eases sim (172, the trace of the embeded ube (072) a. expected [PRO to ave ear} expects fromeons roe eave ea te present fers, these die in properties of the Head H of the em ‘ede pia. both structures, he EP bol othe nonin fetere ‘fH sstromp Inthe contol tract (172), assis al Case to the Subject, which must therefore be PRO. Inthe ECM sevtue (1720) H ps no Case Joleen the checking dora of Ate in he ‘art clause: tore pei, ie formal fstres rie covery to this ositon, we ase. ‘There ae three hai pron about (1724 (t) Why does someane raise overt al the way to the mate postin i which i remives Ce? 2) Why someone permite to raise overtly to embedded sb po ‘on? €) Why mst someone rie overly ros the sac poi ao PRO io (tay Problem (1) evecome by Procresinite, which rues ha the se ‘ond ep in ming wo the Casrcheching potion be cover assuming that Ary & weak. Question (2) has already een unsere someon has Ms Couper scesite atures, ad one of them it extegoial feature, checks the Strong nominal eture ofthe embedded I (EPP). Problem (3) now ds ‘ras i smcone dors nt se overtly from the trace positon, the dt ‘tion hes Tet us row turn othe contri css (160) and (171. each case tne reference ot determin hy the nial mamsraton inchas a second ‘esivaton inthe ease of (168). the one that yes (170), he case of (Utley the analogous one tat yes (1718) with the tae of ras 7 ‘Our goal so show that in the ease of (16}-(17) economy conser ‘sions compel rainy of ere rom the erbedd case, while i the ‘ase of (71, the sme eonsigeratins Bleck raising of 7 fom the en tecied clause eguirng ening of somrane to ended subst pos tion to sis the EPP, The properties of the constructions suggest that the answer in Gaheor. ‘Conse fist (69) and (170, ses, the street is com ron tothe two deitatons Iv each, some singe we comtrct Y= (173, ithe sales. (079) [to Be someone in the roo) “The net step mst fil he specie pasion oft sat the EPP. Gi the nial umeration, tere are two possibiies we ex rae seme fone to (Spee ] or we cn inet there inthis poston“ The former ‘hoie violates Procestinne, the ber does not. We therefore choose ‘ie second options, fing (174) (478) [, re 10 eB ‘later stage inthe derivation we each he strate (179) (17) ose [ter tobe Convergence requires that [Spc. 8] be ie. Only one itimate option ‘ite torte re, forming (170) We therefore sect his option. “The argument based cn the assumption a a para stage = ‘deat ro numeration N, we consider the referees set RAN. fiom a highly “ocl" pint of view, eacting the bet possible (ox economia) ave nie ip R(N, 2) sage. This ere ees npronch i preferable on conceptual pounds fer the usu reasons of ‘ection of computational eomplenty ana once agin, concer tales and emia demands eine, we wou hope if We BFE On the right tack ‘Capes andTeetoratons a ‘Why the, does the sare argument nt favor (171) over ITIP The cornmon pat of te devon opin (173). We have two way ofl [Spex 1, sig of someone or nsartion ofthe ater bein prefered Suppose we insert J then rang tv fore (1718), analogous folly tothe legimate outcome (170). Bt we lady Kw tha the wepment hain U1 (1710 lacs oles ths analyse section 46) I this cases the derivation to eat, then the unwed oeome baie and ony (171) i permite If this only a ese of convergent ier isthe it bleks the dese outcome (ITI, incon. We therefore conchde that an argument wih no Golenot a leita bjt, wo lating Fla causing the derivation to rah, x eonchsion ht tra though not previoey forced. In ction 46 we reached a somechat weaker coscuson on dete rounds. There we found reason to ble that x derivation cases if fies are not propery assigned evi open the Gestion whether the problem is flue to assgn a Oro, orto rcv on, or both, We ne fave a pal answer: ae of argument to rie # ce cases he “ervatio to crash The stats of fie to nign a oe mains pen IW carer work st has bean suageded that external ole need not be aiped ip nominalzations and iin this sense open, we have vt ‘ext tha possibly, or confred i, though we hive fn reason to uppose tbat in vertsheadd eomtnictions, the question of assignment ‘of exteral Grote arses only in a somewhat diferent foro Ira com ‘euration [-VP] is frmed with Spe.) that corto jus 64 teat Ook, and the qusions, what happen if @ nondrpunent (Gn expletive) appears in ths positon, violing the &Coietion? We wil Tetum to this question eng st without dente arse. ‘Let us consider how the economy theoreti aeeoon of aking in expe- ‘ive constructions comports with ether dicussion of -npeties, 2 in (rephrased ere as (179, (196) it scm someone was ol st IIL A air stg fhe dition a (176, we ve (17, aap warm, (179 [, ws old someone that TF] Sine the numeration cade expletive, ws hve the sre two options sin th eases jst ics we can raise sme overt ite later ‘epton bang pefered, Suppee we insert hen ring io for (8), sralogourto (70 as hope (474) isa at was tld Someone that I) ut the derivation eres for reveal easons (the Casechecking fare of matin Tis no! erm, the Cave of someone not checked). Since ‘sonomy considertione act among conversnt dervtins th pre fered option of insring the expleve i (177 eaneot be enoyed and we derive (176). as requ ‘One rect of gues (163) tll resins unanswered: wy the permited MSC structure (16, repented here a (179), not blocked by ‘he sitemative (1796), nacord with he reesoing just reviewed? (OP). Bap Agr Subj TF XP b. Exp Age UCIT. Sab) oI) fo other woe, wy ce (1684) repeats ere, Bh eit? (180) ¢ there scons fy someone ft hein the eo 1. sete coms Tse to bein the rom By she retoning just outned, we would expect (80) to Hock (180) voiding the wltion of Procaniaate by over raising of soncane. Let tw delay the question until he next scion. eis worth higeltng the basi sump about references ht undies the presi dscsin: they are dtennied by te inal ‘eration, bt in ily "ocl” Fashion. Ava parte stage 2 in he “ervlion, conse oly the onions tha af permitted fem = to LE, ung what remains of the ini sumeration: the must econo fof her Hock the ethers. But weak even a mtrower question: E ‘wich operation ist yes a convergent derivation i most econaical ft his pon? Tis, we wet Merge over tract Mose if that yells 9 Conversint derivation, inespective of consequences down the Fond as Tong as th derivation converges et we slet Aurac/Move even io ling Proerastinte if tt is neesary Tor convergence. The problers of ‘computational cemplesty sre tes consderably reduced. though more romaine to he done, no dos. The sssurmpions throughout are staigh Forward, but rather delete. Wt reais to investigate further eases and ‘sonsequncs 4410 Fusctinat Categories and Formal Features ‘What precedes substntly revises the framework deetoped ia chapters 1-3. Bat we have ot yet subjected funcional categories 10 the same rials critique. tn the al ection would ike to spre tis qs tion, a course that leads wo another fui substantia modieation. Eves more than before wil speculate ether rey. The ses hat aise ae fundamen tthe nature of Cg. Havin 0 do ith the formal fears that advance the computation cman strength, vbich dives over op erations hat ae eects a he A interac) a the arin cate ovis that eons primary omtimes entry) of such eure 4101 The States of Age Functional earegones ave a cnt place the conception of tanga ‘we are imesigating. primarily Because oftheir presumed role in enue ‘hecking. wich what drives Atrac/Move. We have considered four fuetonal eategries: T. C.D, snd Agr Th int sre have Inert ale features, providing “insacions a either or both iterfce lel ‘Agr docs aot it eons of Tetrpetbte forms Fstres only. We ‘herefore ave fly diet evidence rom interac elation about TC, tnd D. but not Aer. Unlike the cer fuetcnal ator, Agr is pe Sent only for theoy-interal reasons. We should therefore look mare los at wo qusons (US1) 4. Woes does Aarappear? 1. What the estar constion oF Agr? To section 422 we tentatively assumed that Agr lacks @-eaars, ast sit fal dearly) lacks an independent Case-assgning Feature, th be ing provided bythe V ot T that adi wo i UF Agr indeed lacks Feares swell we would expect tht the @feturs af e peda Pred (Ger or adjezive ae added to Pred (opinaly asi i selected from ‘he lexicon for he numeration. We kad He war for ihe assumption ost features and 0 far i hae had Bit efecto he aly. Bu i becomes relevant as we atemp more cael anes 1 he qustoes of (G8), Tt cominue to assume that he oni! assumption Was covet, ‘etuping to the question athe ed, after having masowed Sigil the ange of eonsterations lear ‘Wie have evidence bearing ch question (181) when Agri tons 30 ‘thatthe poston és phonetic indicted by the overt categories tht raise 1 Vand T by adjunction, DP by substi in [Spec A]. The ‘hes erp isan MSC with ject rising, as inthe Telandle TEC ‘ioral asaned te lesa sams as they are drs rvs the lesen, we ‘onclde that Agr const ly ofthe tong fate ha fre eng. ‘Ceruin probe that sro i eer versions now dapeat. There 10 need to del with opsinaly song Agr, oF with the diferencsin -srength of Agr and Adxo Since Agr ston the Sst problem just 2 matter of optional lesion of an elenent (tenth of FD) fom the lexicon for the numeration. the ineducble imum: snd iferencein “Srength is inexpressbe, There sil emaies, however, a conic Between tho Etheortie principe that sare verbs ave 4 +P stature and the assumption tt ert obiet msg intra to this constrasion Gecoi Bt and p. 316, {et us um othe properties that remain Since Ar consists solely of song Feats, it exnuotatract covert ising” We tave so fr assured that Sub bi) and Oy ob) ‘ise to me checking domain of Apr, ening info checking relation ith feaes of T oF V adjcines to. Agr (etna. soined within ‘Agr te X? projection headed by Ag). Bul wih weak ABT Eoee, fret yang mus target T ane V dct ‘There now ro renon to poste Ai unless i induces overt aie ‘ng of DP to {Spec Age). What about Age It appears in MSC but seks independent metiation elsewhere 5 ates nom snd. For la aps ofthe French-English type then, Ag i tin the leon (ales (MSCé appear macpinlly, with exraposiion). Age therefore cere im phy ested ways, ‘The next question 16 10 igus into the justietion for Age with strong estore Let us fit look at Aro then nto Abt a Cones We cert attention pow 10 tensive verb consrctons which we contin 1 szue tbe ofthe form (182), poring (Spc, (the ae ff comple tera domi), aoe V ies overly tothe Light verb, Forming the comrler Vib = [Va ‘Assuming unergatives to be cone acs, the oaly other VP Censruction it that of unaccsatvs Inking the shel ot elvan, tere ‘Seppo that a derivation ha fered (182) and Agr is merged wih it ‘Agri ellcion of song fests, ster [Ds] of (VJ or bath AS rots, we need not postoate Ape ence for obtraising dos ot ont only of stone [V1. Helmbers's generalization sates, im fet, {hat cannot be jst strong [DY Let us tenaiely assure, thc, tha ‘Asti (strong [Da strong IV). The elect af ang Apr 0 comps ‘overt rang of DPC (Spe, AB] a of Vb to Ase. ‘Conse the fit property. Theres pl way 1 force overt DP- raising withosn the tional eateory Agr ately. by adding 10 vel a suong D feature (or paps he more neural strong [nominal fe Tur) tha require overt sbtitton inthe “oes Spee” of @ mulip Spex configuration. 1.08 anes 2 this peition to frm a bain (Ob itl ben the checking domain of V and theefore able to check is Case and (objet apreerent)Geatres. Recall that Subj inserted by ‘Merge in (Ses snot the euch domain of , becuse it dos sot head a nontvil ehan™ ‘Object sng, thc, takes place when he Hight ves v that ead the ‘exnsive verb construction (182) asgpned the soa feature a it raw Gon the knicon and plc! in the numeration, se section 422 “The choce is arbitrary, freed, or unavaable a5 the langunge his oP tional itor, or no overt rating, respecte. Sine Subj ot in the checking domain, jn sted st doesnot chek thi 0% Congo an Taefomatons as feature, san outer Spee must be constr fo that purpowe. One 4a) israsing of Obj hope 1o show tat al eters ar exci Suppose that en adverbial phrase Advi adjoins to and obj raising crosses, ding he Constucion Obi-Adv-¥™" That proves no reason to posite an Agr postion outide of = a strong feature ed ony be sats before a ditnrtighe ator is erated ™ (Overt objet ssing therefore sams to provide ro compelling resem {or asuming the existence of Agi. The cer property of Aesy i that st forces overt Veaisingstly to T out of ever dil vite. The motivation was theory. bu tds Fears within the moee rested framework 2¢ we wl ae, The propery x cri par of the expression of Helimbers’ssereraization iat Sect sing is cotingeat os Vrsisng, but to ntsoduce that consider ton justi postulaton of Agr tele. For VE. st les, seems that we shoul dispense with Ae Comsider adjectival eonstucions mich (55), Hrs, U8) John is Lage ty Abra niet) We assumed that Jl merged in the poston of in AP as [Se Adi | (ubjct of ingen a this ease). tng to (Spec. Age] for DP tine agreement then on to mie [Spe] for DP agrement > Do we need a song funcional extpory (Age hae to esd the sal tus complement of the copula? Assuring that [Si Ad] is anata. ‘us to speier or complement of Vand more geal, thet the ‘complementarity of ®heery an checking theory ho inthis ese a8 ell as others then agremen wll no te checkod in this position of meer. We assumed thatthe (—Iuerretabe)detues ofthe ade five Adj are checked by overt raising of Hs suet Sehj to (Spe. Aer ‘dof Adj to Apr—the lanes problematic. x mentioned, bitte Ag i cal in Engst (ee ete 3). We cam now avoid tat problem by stn ating Agr and edoping he analy jst proposed for over object | ing: Adj is signed the featre srg [oman ast is raw fom he leven, and [Subj Ad iss to the cater Spee regi hy the sone Fes, ening the checking domain of Ad, To thi cathe derivation ‘wil converge only if the sions feature i mkt, 50 the choice 6 fn effec oblignory. Note that features of Subj cannot adjoin covery 1 the Adj as a review of the possible eases shows (on plausiNe sumptions. a capers ‘We therfore eininate Agr this ae ton, sing Simple mechaisns nd ovetcoing a care probe abou uaexpeed eet raising. The ‘sructure of predcte adil eestuctions is ox (183) bot rather ash, 184 Sohn ag a inteligen} For sal causes. we bate sorting ike the ovina assumptions of ‘Stowell (1978) on which mach ofthe rock onthe topic as been based. ‘au consent with otber asupion that we ae now adopting Teall of this turns out to te cotet, we can eliminate Apr Fem the lea inventory entry for any guage Tuning 1 Api We need to consider only MSCS, which Bae the surtice order [Exp-V Subj. Our “sumption sofas that the subject Sub in [Spex] and the expletive in (Spec Awl. aed that V has red to 4p, Suppee instead. we fo low the in of reasoning aged fr Apo, limiting Apr nod adding an optional strong festre that asin an cuter Spec to T. The situation tiers rom the ease of Ao. [Spe] in (12) is equ for indepen ‘ent Greate resins 0 oly one new Spec is required for objet Ing. Te conta, T requires no Sec, soe ave 10 aeormede ‘Spec thn are induced nly by Tete strength. Independent, we ave To eowount forthe fat that he cdr not the expected (185) but at (LESS, slong with eter abseried properties. (185) a. Esp [Sub (TPT Exp To Subj XP [MSCs appear only when the EFP olds. The question of thir ature ies, hen only when T lend bas strong [renna festre, which {dake when chesked hy DP or NP in [Spex] Suppose that che eivation fs eicod the stage TP with strong, and the mation ‘cena an unused expletive Exp Then Exp ean be inser by Merg 10 Sly the EPP, and we have a ordinary expete- esos conse: ton. The rng feature of T deletes ad furthermore eres ince the ‘env converges. Hence, overt MSCs exist only if T has x parame ‘ead property of the hind iced carer (e below (38) which silows 8 Inerpreale feature Gin this ease, the strong [eominat] feature to escape erasure when checked. Ube option ssl, the ‘here must be kip Spee construction, with n+ ¥spcier if he ‘pton is exered tine In 3 angunge wih the EPP but no MSCS the stong nomial Fate of Ti introduced into the derivation with Cte ed Trees ss ‘n=O, tence erated when checked In tetandic he decitve fers {odie that "= Gor n= I: in the later cae. T hos two Spec, ‘Lec sce where his course lead. lniating Agr from UGG ely and. a lest for our parposs bere, keepin 0 fonction esepories vt ntiasiepropertisthat are raft atthe intertace level The questions thax sre ae atin ater dette, Let ws ely a ec iven Lpnton of them unt see urbe roundvork seid 4102 Core Concepts Reconsider “To accommodate te change Irom an Aur-tesed toa sine Spec the ‘We have to simplify tbe notions of equidance to closeness that ‘entered into the deiiton of Atract Move, Thst wef expresed inthe pipe (7), epee hee. (186) Bis loser 10K tan if fesommands ais notin the rininal domain of CH, where CH ithe chin ade by 7-7 Meine within the z0seve projetion HK) ‘Bu this no longer works: withthe eiinaton of intervening beds in: nal domain cotape. We thefore have to eacide noni chains | fiom the account of equiistance, relying instead on the mich tore Aiflerenited anahsis of features tow avaiable and the immobility of facestha onthe Fac that only the bef» shan canbe “Se” Dy Ksceking the closat to aac Tn the eat formulation, the tsi eas (5), repeated ere in de ‘more geal form (187) oacsommedat adjunction of 10 X as wo 8 ‘sbsinson of ain (Spe. X] (where X may already be the Head of ‘simplex zeo-leve proeton. ay wh Ye Wien sss, trgeting sit erentes anew pon x), which my ‘lier be (Spec, X] or uid wo (Y-XI (=X?) ell 1 the ger in citer case. "The minimal dorsi ofthe chain CH = (¥.0 ints 36 ars Spec, and ZP slong with eX) formed by raising of, which i within 22? or ZP. Cras, Spc, i wth the “meghbortood of X" that is ‘ignored in determining whether lose enough to be attracted by X (echrcaly, by is projection). That sasumption was necessary inorder toallow xt ros Sper wo reach eOS). Ina uanstve verb contraction, for example was asumed hat X= Ag, Spe = Subj Yi the verbal ‘ement tat ais to Apr and Obj s within is ZP complement. Obj as to rise to the checking domain of Agr Tor Festre checking either ‘overly or covertly ining that it "es clone” to the tng Spey Most of his now bade the point. We ave eiinated Agr and its proeton from the inventory of elements. For the care of overt objet Tris he struct formed sno tng (157) ith X = Age ad (8) = [Spr Azr but (188. witha extra Spc in VP (sy YP. se we wo bis he vert clement or isis, if tbe cmples has aed ure 0 ain to "TY" and YP are projections of the light verb io whe V tad to fren Vb ZP [Obi y the eae of Vand Specs tpt) era hy the ring operation. Spe; is Subj and is only acest that be no clore to Ue arget Spe an at ZP. For {his purone, lcs to simply (16) keeping just othe il ebin CH HOR) (ie ead of K) ned is misiral domain. We therfore estat (18) 2 (189), (1) and are equidistant from Fy an arin these inal oman Hence, y= Specs an f= Spee, ate iia fom ‘lusratie example just ies ‘We now define “los” for AtracMove in the obvious way: 3f 6 commands sand i heart of sing then (490) fis loser eK than uns isin the sume minal dnsain os Gar ori Obj in he Cacao ant Transformations oa ‘Wie thus ave two eases to consider. We ask (ease (15a) wheter and ‘ar euiistant fom 2. and (ase (1900) whether Band ae uid fant fom Wither is tr, then doesnot br raising oft T case (190i), and «are inthe misimal darn of HOR); and in case (1908). ‘anda are m the minimal domain ff for some heed In ease (19) Isin the “neighborhood” of H(K) that i enone inthe sense of eae, ‘expostion By case (190s), Obj within ZP in (18) lose enue to be attracted ‘by Y (YP, a0 this poinn, since Spe; tin the minimal domaia of H(Y) and is therefore not closer to ¥ than Obi: Spey sod Specs =) ate ouistent fom Ob. Therefore, ether Se in Spex, oF Ob Gn ZP) ‘an se to the new outer Spe. Spe, requ by the srong feature of Both Obj and Subj mst aise For Cae chackings and something st Ini (o ches the Case feature of T (ero some higher entegory Hf Tis 2 ibnginfintval as already discwsed). By case (190), overt object, faitng to Spe does not prevent sbiet ring from Spec. Because Specy and Spec, ave equidistant From any higher target: both ore inthe sinimel domain oft. How about ist rang of Ob fom within ZP. targting T. crossing Subj and Spec? That Based by the MLC, since Subj and Obj are not equcstaot from T, given the ¥-VP analysis of ftanstives: they are in eiferet minimal domains. We wil turn a ‘loser analysis reviewing other option kine hee Consider the following eousertpument.Soppose the langue has {he EPP ard optional objet suing Trees (Spee, T] and’ permits fn omer Spec, Spe, Beyond Sub ia Sper, (bath over. Suppene tht bj ass to Specs hen eases dete to [Spe]. sting the EPP. “That much x permitted, Subj ana T have not had Cas etre checked ‘et tat ean be overcome by cover rag af Subj targeting T, which ako permite. So the derivation converges incorrect. But this devia ‘ion i blocked by economy’ conditions. involves thts raking opera ‘ons and two would sufice for convergence abject raking followed by sue rasing to [Spc.T] (in bo nse. wth wo vations of Proc. ‘inte. the minimal number with to strong fetes) So the unwanted fers of steps, though permted is bare’ by economy considerations: shorter derivations block longer ones “The computation is Local: after raving the abet, we choose the oper- ‘tos that wl lead te the short cosergent deren: ring of Subj to Spec. T} We alo have empire support forthe tentative asumption ‘made cuir thot shorter dervations, lolly determined inthis ses, Hook longer ones (se dsctsaon of (14), [Note that we have lst Holbergs generation and ober fs ‘of Wralsing cn exersion of chins: tal fa consguence of exons ‘hans Go the deiniton of "owen." Such generalizations, valid, trou ow have to besten tems of property of Vib i (IEE ican ve scond ter Spe only if eee. These no obvious reason ‘why this should be 0 any event the err, more complex defo of equitisnce and _ozenes tno nesematy and act no psi. The notion of equ ‘itance ay sl be med, a cases jus evewed ad others, But tas rarower Sone “The conclson tht suidistance is stil vended ree on a fc as sumption tht could he calengds thatthe tong fea of st be {ats by the outer Spec, Spee, of (18), not the iner Spe. Spe, A ‘we Brom, however, tha soe Se of ris tivated y cosiertions| ‘of thoy (host the exereal arpunent) andi therefore independent ‘ofthe stank of the ater Sp x present only to check the scat featre But oth Spce are within the minimal domi of so ether svlable for Oasking of the extemal apument of «tase ver, Suppose we allow this possi, ro thatthe outer Spee can bos the enema angument In that cine we ein drop the noson of equiistance nity, sping (190) tthe statement that is ltr to the tart K than ecommnnde follows he, that Obj can only tae to the Ine Spc, Spee, of (18), to check he strength faye and undergo ‘overt Ca marking. W oer bjt sing takes pace, tea Subj wl be Inergd in the outer Spc to recone the exert ele prides by the ‘coufiguation, With "loser ar” este to e-corme. only Sub | the oer Spc cam be attracted by tot hat Subj aay bas featur that wil check sublabls of T). Therefore, Obj i Gozen in place afer ‘vet eject rang, andthe concision eal above flow diet. ‘On these ssunpthng, fellows that Subj sbways c-cmmands Obi itn Pi particule, ti re i expletive constrtons, wheter ot ot oc sing has akon place tht peat tobe general the east wth seme unexplained exceptions ace Jonas nd Babak 1983). We Ssohave a yornewhat more tal accunt of green, with the ine Spec uniformly ering into the ration (the Spee @-posion is ot fat to i for remonsaleedy disased). I also should te the c= that enly the nner Spe in lie Spee constriction cam be Bibb fassming that lay ener erusally into nding in one of eer pewile ways) thous contol mayb more fe, as tan is (= Categories ad Teanctoreatons 3 in 14.2 ht apes ote theese Hopi Us, pol om "eaton, Father queso eben we tun f tn Wi op inemal aren ttre whch Teno fe Leta trp sale the wo ons forthe nton “cows than” in hee ht he Oe a eed pr, ec fo pe id ate (ana ean erp ppc, wl ree the examples cow othe ton hat aes ving io teeter Spe to ensure hat the equ commu follow ten unde it tore compe erative: yt cee at te gum Akouch or sin sme clones eso econ _ tha bet ang othe tone Sper td ony te eae Spe | attracted by T. m= * a We alo hive to sete some quetons abou! aunts tt have bee Jef open but become more prominent Ho this mich more rericive Framework, cover adjunction beng the mos interesting eae. Emprcl fidence for covert operations andthe stuctre they ere bien than for their over counterparts, bit ent, and cooeptua tguments ao cary us sere itn et et ‘One reesonable guiding dea tat interputve operations at the ne lerface should be as simples posible. Baring empl evidence to | the conrary, we assume that th external ystems are inperished ural extension of minima nttions fo the npg Tcly Ove ‘roo, incading the systems (possibly dedi to fnggs) atthe “other side” ofthe inerfae. That mears thatthe forms that teh the | LF tee! muat teas soar as typogial variation perits--uniee, ‘hats posible These assumptions bout the ieee pose fai Te Sticte conditions on application ard ordering of operations, ctng | down the variety of cerputation, slay a wokome raul for reasons ‘Seay discs, At the A-P terface, overt haiettion providen ad ona evidence. Such evidence slp unavaiinble the Cine ace, bt the eeperalceaceptal consideration jist reviewed cat sme | 89h We have been pity eying on then throughout for ex He. to conclude that covert fetes adjoin to the heed of = chain (epee, x raised ver, not co the trace or optional to ser see (4)aod discon, The central problem abou covert adjunction concerns he strtire of ‘Te at LF. Consider st the sche! ene a TEC with ject raising \etanic, Pung aside the obatrd postion of T°", we asume e Form at L¥ woe (ES), epee hee, with YP an instanceof (18). xe Chane 9b pts cTe= Ye fp and Sub ae sci of he T head of T°, wih ery fonction Te'Vb-“[ Ve) Taisen T™" &(192) and he cm fewen v8 (9 am oN “ ame &éy zy ro <™. Here 08) ad fy ae species ofthe when of VP ‘Suppose V rates overt ad Obj does uot (French, or optionally Be- lund) The complemen of Firs rm (195) ha sacks he over Spee oocupi by Obj which remain in the position oy, The Kral feat FF) into TP fr ere ee, Rfors hs ove cperation takes place T°™ i again (192), To maxi sian to he UF ovipt (182), FF(OB) must ain to the complex form (192) se. Forming (194), not to the deeply embeded ¥ within Vb, which actly contin the leant features fr ebecking oT rey oN wy “The operation is permite, since she features of V ate subabels ofthe target of edjunson: uni stds the conditions on “closest tet” (Goa being “nv,” iseused) ae frration of complen X's. As Song thi to he the general ue. we conch shat adjunction Categorie and Tranoomans vat ‘alway tothe maximal axre-eel projection X=. never “ntrally” to ‘one ofits constituents the simples assumption nay eve, ‘Consider un Enpstvpe language with over eosing of Subj bat not Vor Obj. To achieve maximum smporershment atthe interface, we ‘want T° at LF to he a sir as posable (18) fat, Heil ‘cept that in pace of Vb it has FE(VE), since Verising i cov, [FF(Ob)) therefore cannot rae 10 Vor tothe verbal complex Vb before ‘Vb adjoins to Tf te, the stryctre formed would be que eifeent from (19) Aer covert Veasng, FF(Ob) adi ro T°". again foxring (198) at LF (with FF(VD) in place of Vb). The ondeing i force by bare output corditons, if the conjecture about poverty of Interac inerpecation score. Suppost the largsge lacks overt raising of ether Subj or Obi, 30 that both FF(Sub) and FF(Ob) mise covertly to TP. The pover-ot- inurpetaton conjecture ees tat Vb ase 0 T tor either set ‘or chject ssing; we thus have (192) once npn desired, whether V- fasing i cvert 38 na VSO langue) oot in which case Vb in (192) {is replaced by EF(Vb)). TP now attrac Sub, whch i cows than Ob orming strate dentin co (198), except thas the ears FF of (488) bappen 0 be FF(Sub) rather than FF(OR). Bot i also ie sary for FF(Obi ora. That i now posable, sine the tice of Sob (one Sab sel) is not cle to TP, bang inazesable to Arta ‘Move Therefore eds up a (95), oy oT Frio PRS) wy Rca that ina neal expeiveconsrction, the strong D-eaere of “Tis satisfied by an eaptoe rather than the rasoé Subj the cave of Merge that was asad to Atract/Move (oe end of seton 4.) tn ths case T°" i again (19), and [Spe. Ts ocupied by the expletive ‘As captive construction thus has # certain struct sesennes tow "¥SO configuration 3s has been pt trophont sa hope We conctude that aunction i tothe maximal X® projection XP and that eds ese before cements of thir domains, conditions remi- rine of eycy, These are decrpive pnealizations that we deve, tot principles that we mule they fallow fem the miimalis pine [He GF poverty of interpretation st the iterfoce. Simple and plausible “suinptons ufc to puarante vitally the same LF form for T™. ‘ver the (olan! range we are considering. So far thes concu- tins are mevivated only bythe conceptual requirement of maxiniing Uniformity of UF eutps. They supplement te eter consi that traces of A-moverent neve eter inl Anat Move, whether overt oF ‘Suppose tha output contons at the LF interface rule out (198) under he rather natural requirement that FF(Ssb) must e der consideration: (1) “lose thas” i deine interns of cera sone an overt cect ring ea only be othe iter Spex (2) “lose thane dfned n tems of command and equkstanee, nt the Objet nay (Gethps ent) rs tothe our Spe. Again et ws west aten- (Cueprc and Transfers 1 tot to the more complex variant (2; under (1, no problem ats, 5 ‘can readily be coche. Sopose then, that overt sustittion i inthe ter Sp of: Sp of (IE), this case» as a stone fominal] feature H cannot te checked by merged Subj a5 we have seen so tere mast Be a eta Spee sin (IB. ore exp (206 om sre, OY smh \ ww ca) [Apain, Vo isthe compen form [,V Subj & in Spry for @heoree fessons unrcand to tenth, Spec, must bei overly to remove the srg feature of before 3 tint larger category is Terma i this ease, Before merger with T. “This can be done by Mere or AtryMove. We hve already excteded Meigen argument inserted this position does not eta check: ing relation (0 that she strong fetwe i nt checked) and ako Wcks 8 ‘oe olatng FI; he dernation crashes. Expeives cant be merged fn this postion, a8 we have sen. The enly option, then, i ating of ther Obj or Say. We want to low only rising oF Oi wich 3 me fave sen, then permis the dentin fo converge by easing of Sub 0 the checking domain of Tandon hat way ‘We have brie (ad incompletely) considered why easing of Subj vo ‘Spe: of 6) is bata. Let ook at the ossitis more closely 10 iy he ises Suppese tat Subj is raed 10 Specs in (200, 1 in the checking domain of V, and checking rations are esas for Case and ‘sures. I features mismatch, then the derivation is canceled. I hey ‘ah then Subj receives aociatve Cie an object ugreemet, nd the (Case and 6 feaies of Venn The Cae of Oj il hs to be check, So that Wil ave to take pice in the checking domain of T. But wh raed Obj cannot rach tht poston as we have ee, Because Subj o costes loser to T. The eof Suh Sry i inl o AtratMove and therefore dacs not prevent sng of Obj, but Subj ise dome “The only posit the, i tht Sub ene Fart, to the checking onain of T. Now is sce in Spe is evil, and FF(OR) ean wie to T The tac let in Spec, deletes, and at LF the res dent to the result of the derivation in whic Sp was never cossrictd The ‘rong fare of has no fect onthe PF or LF ouput in th deriva tion and therefore cannot have bezn selected Tor the numeration, bythe ‘economy principle (76). This option is therefere excel. Wf Specs exis tall test be formed by over objet using ‘As the derivation proce, the next ehekingrlton that can Be taishe substi in the fs (Spee. Tats Frmed! (EPP), tr by Mere or Antal Move Por Mere, the only option i an exiletive, icf pe exabshes checking relation only with the strong [or ina eau 7, requring overt smote ring sing of a ee tive to thi poison werks the same way. The only remaining car raising of an angunent to (Spee. necessarily Sut, as we have sec ‘Then snares ener ito checking reins wih he “moe rotate” ‘bates ofthe target inthe ebious Ses: the features of V (ich tare the only ons). and the Case feature of! Vb cance ive 3 ‘tong feanse at ths stage of We deiation, bu» checking relation i ‘sablshed with he sang forint] feature of T tat fred the over Substitution, The subiats in checking tlaons ease if matched) those of Subj thee i a mismatch, the derivation neared Suppose ether T lacks e song cial fete otha fete bas ready ben erased ty subtution of Exp in Spc. Then FF (Sa) as toadjin to T'™ forming (198, modi igh hee amt mrsap w oY Checking proceeds exactly 2 in he [Spee Tease, eating the dsve- tion if there is @ mismatch, erasing —fecerpetable fetus if there is» rah. IF Ob; has raised overly t [Spee VB ste Features ae checked there and undergo vo cover asing. H Obj has ot mised very, hen FF(O}) esses tT fring (199), pete here. Cnzoesané Tanonations m om or wri res w oY ‘The Coe feature of T has already been erased by Subj, 30 PFO) checks the Case feta of V eacling the derivation uae town (Case feature i accusative. Nominative Case and subject agreement na early coinage. ‘We rey cricilly here on several earlier asumptions, among. he thar AucaciMove “sss” ery the bead of om Actin ei that mit ‘match of flexure canes the derivation, I dissing te air topic, ‘we considered the weaker principe (107) tka allows the devation| to proced after mismatch, posxbly converging in some ferent wa, Largely on concep grounds, we rected that opin in favor of (108), mich cancel the derivation under rama (1) fs referee beens ‘tetuees computational eoplety. We now se thal the deion »a5 necessary on empiri routs as wel if we were tallow the Broader ‘as of demvations, certain cies forthe (options) Cae es of DP weuld have allowed the derivation 1 ecmmerge peopel, with fccative subject and nominative object. Onc asin, the picks ‘hesen on conceptual grounds are enpiially confirmed, a8 we Would ‘expect if the Mintmatst Progam is capurng Somehing re nd poe tant about human langage ‘Thirst am crhawsine review of posites, but itinchads range ofan tht sem tobe cena pi ‘The probens that eran hav to Go prima with MSCs. We ave tone a considerable distance toward the coacusion thatthe srucuse (ODD repented Heres the only posite frm for MSCS (with without ‘overt objet asin). (205) Exp [Subj(T?™ Xe “The structure pert, nd we ave seen that a numberof unwed pasties are eacladed. Some romain, however, and thre it lap ‘ubsanial conceptual problem: how do we acount for the range af ‘ptiens? m apr We are concer wih the “double EPP” structure (210, 210) fp Pe YETI) We saat 1 reste the option 10 (208). Let us break the pron dows Ingo evra pats, (211) 9, XP an YP cannot both be Exp. 1. YP cant be Exp. (© XPoe YP must be Esp. so, we ave (200). “Te eonepual question is 212 (212) What optional sices ca allow tre options (3) 20 Spee of T (¥503. (6) one Spec OFT (EPP). c} two Spes OFT (MCT? {Case (212 is 210, wich we hope to rest to (2). We have to ask ‘hw he tre options of (212) are speed along wth combinations et ‘hon for example, lesan, with pions 2128 and (2120, Let us begin with 2a) Thefts are steigiforvar: there are no steoctues of the frm Exp Exp. and furthermore. the cna obs ‘ation generalizes to @ broader eas of cess, intung the perennial ‘roublaker (213. (213) thee is believed thereto bea man in he room) “Thee sructies are ached i we take the tong [oom fxture of [D] to be (84 not [nominal] neal inclined (D). That essonae move: ais the pst eases t the ones We find im DPS fenelly, namely, N+ D raising and it has the conssuence (last ‘Feo unproblematic hat the ssoae ofan apie aust Be nonspe ‘ie, wheter is wn ordinary expletive constrvction or an MSC wih he soca im (Spee, T] (se pp. 342,350, Let us adopt this assumption. “Then Exp. Exp consis ae sraighiforwadly hed. Exp isa DP. so the trong festure of the outer Sp wll ot be checked by msi of the inner one: only DP-NP uicurtres are possible, analogeus to ox tary eapletiveasociate pire. As for (213), the features of the asc ‘2 man ean ie to embod Tin he normal wa, oth [N] then asin to tere, to check te tong fete; ba thre sures itt to LE. the derivation eeashes (ising of here wil Hp, ec itis & DP) ‘Since Exp-Exp i bared, to sabish(211b) ve have to show only that Argunen- Exp is impoase in socessve Specs. We know tha ‘vey point ofthe denstion Exp nthe poston (Spec, T} an thal Cetin and Trcfomatons » ‘ust be in the Exp-Subj order of (209) at some stage ofthe derbation for is D-cature 10 be deleted by Nerang fer Subj. What fas to Be shown, the, that once 2a) i formed. xp cannot ate to Becoe [Spec. HI. with a subsequent operation forming (244) (XP the outer Spec Hh 21) ExpSubj XP lon ap AT) “The problematic dvivation is haved stesighaforwarly (21a) the [Neuro Sub adja 1o Exp to dete [D, But ths operation eo- ‘et featere rising sed eannot be flowed by overt ralsing 1 Spee to fora 218). ‘remains to eb (211) ad to answer the question of (212). Let 1s bepn by asuming (2c) to te true (we wil ern to tht esr tion). We now have to adres (21) special we tave to deermise ‘hat the chice of opens mst be. I xy Yo ake the dsinetion Detween ne Spee ad some Spc: that a mater of svaiabity ofthe strong [nomial] eae for T- The bard protien i 16 distin eases (0120) and (2122, each assuming tha T a the steoe froma) fe ture How can that distinton be expressed within the int esoures| sealable? ‘Suppose here 8 pormetr tat sows two chs: one Sp oF wo ‘pect That propota fail, because the two Spee eption permits Subj and Obj both to raise to (Spe. 7. ilating 2 i he curently ‘sve framework, nothing prevents ring of both Sub and Obj to the Sper of (210). Suppose we modily ths spproach, appealing Yo an ‘conomy prinsple that allows the twe-Spee option only the deneation Seoul pot converge without ia wrant of 76). That ile to, teense ‘thar even the desired ExpSubjstrcture (29) as we ave sen, the 0), aad e+ Limes if 90 Exp sill fo ner "ion, Each such application violates Precsstinate, which suggests that the parameters valved in 212) shoud be fase in tems of such sioutions, 1A soltion of Procrastinate that is equted for convergence isnot 28 ‘conom voltion: one that not requized fr convergence i an econ: ‘omy vielaon, To fata expesion et us sash veminologially betneen the wo (adallydiferend kines of violation of Procrastnat, speaking of fred vars (or convergence) and unforced voatons (ere econony violations. Ina VSO strstr (pe 212) with weak hee are no vilaions of Proerasinate. fn an $-VP siracare (ype 2120), there sa mest «| Sine forced votion that ine Exp i seabe for meron. In 2 ‘MSC an unforced violation is tolrate: that 2 wilaton of Prorat ate ot eoquied for convergence. ‘We therefore have to options 215) a, Tay be strong ort, b. Timay ormay not okrate an unfrced viation of Procrasinate Option (21S) i ken ener fom the ABHT sytem: T may have the strong feature fromina (EPP, or it may nt (VSO). option 2158) Select bat not (2150), then thee sone Spe bu hee can be ao forced violation of Procrainat: Spe, T] Exp its availabe dw sais argument iTLis now. IF (218) is selected as wel, there are ab trey many Specs tha il too weak conclsion, 0 (2136) mus be sturpened "To clarify the iss, let us compare the Age-T approach to MSCS with the one we ae now exploring (2150) isthe eounerpart to the p> rameter (216) of he analysis reviewed in section 49. (216) T may or may no ave Spee In the system of sections 41-43, (Spc, Aptis sinaysavalae and (2124) ant (224 (ingle aod deubie EP) die in the valve of the Parameter (216). The right outomes are guaranteed by the prince ain, (217) Onty one argument (nel, Su) ea ase from VP ro (Spec Caters nd Teaformaons ms But as we have seen onthe basis of 3 mor careful analysis f properties ‘of features and te empire phenomena that wet o be expsne. rin- ‘ple (217) i untenable: Subj, Obj (and eter argument) have tn relable formal fstrcs and ts can be ene even alter feature check Ing the MLC iste, ae tin the reevat eases once we testi ‘trstes omnis pritcipes, dispensing with apparats hai on- ‘epualyonneesary and empinially defecive. The apparent generat ‘ofthe carer anaysisis therefore pons, ‘Suppose angtge sels (2158) a a eption an thus allows & ve- Intion of Prorasnae. This ils exvety the ish dbl om ‘Snctions (land), bu it alko permis MSCs with more than two ‘Specs. We therefore hve 10 resi (2150) to the minal mer of ‘Soltions of Procrsinte-exactyone—revising (215) (218). Qt) 2. Tisstone 1s Teolertera singe unfored vilntion of Prosasine ASO lenguage asa negative sting for (21) (negative 218: T Js weak An EPP langue lacking WSCs bas poste 218] and reet- tive CIR Tie sro bat dees no double-Spx. An Hester Ianpiage with optional deublesubjec MSCs bas [positive (218) and {postive @1sb)p. The famework dus covers the typologies! range we ave fa considera 'A pretier picture would ine another option, the positty of ‘mote than a snse unforet violation of Proctainte. Suppor, the ‘we have ancter parameter that disngishe two yes of langue. (219) 7 tolerates etitanly many vforedvoltions of Proeastinae ‘The langungos we have w far considered ve the neptive vale for (210 no more than one unforced violation i okra and mo Fred ‘ioltons, i he seting of 18s) diallows song 1). Could there be langage wt [posites (219) lowing arbirary many unforced il ‘ions of 210, whieh we may interpreta meaning maa unfored ‘olations? to suc languages at arguments are exacted fea postion ‘Suis of I. The endate that comes to mind is langage in whi Al arguments do appear ese OP, tee sya ok bing ed ‘xed only by some ration to an clewext within the renaining complex ord: pro with matching fetus or perhaps trace with apne > eating in MSCs. so, then (219) his the favor of Bakers (1995 pol Synthesis arate. we chaps le scons poaity worth exploring Wit makes sere, then we woul Interpret te sytem of parameters to imply that wnfered lations of Procastinat can be esllowed, eal (ne), cena (a3) Newlss to say. we ce nam pel ar eu on en ‘Sill untied case (2110 the impesbiity of double EPP co. striction with ao Exp. the two ocatences of (Spc, T] being occupied by Suty and Obj (in ther crder) (an Kedande-ype language that allows MSCs wath Subj-Okj, for expe). The proposed answer 10 (212) bar thi posnitilty. The Inguage would have 1 allow strong T snd a unforen lation of Procrasinate positive (2184) Bok ony ‘isimally(oegatve (219). Thar sto choirs leone) allows the Soublownfrced vition eaguved 1 extrac eaely to argument S| (Spex. T)peatons. "The options (218) and (29) ca be formate ey in ter ofthe mechanisms for multiple Spec mentioned eater (below (58). nel. by allowing # feature to escape erasure aller eheching. The option in (Gib) is ejected i that posta is dso for the song emia] fear of T, chosen if the posit i llowed. The option ia (215) 6 rete if the song [oomina| feature can espe ease ones adopted i iemastesape erase a often as posible ‘Consider the ansarrativeconstnition 220), where Nam i the sle em or ‘Seppo tha Tin strong. Then iter Nom rss overtly o (Spe, T Exp sista in that postion. Exp cannot raise frm 2 lower aus: ‘hat woul Be teed by the closer Nom. But suppose Exp i mere [Spee VP] non pomton) hen red. The oseme emt, Bo the derivation sagan berrd: bythe condition tht share derivations ‘ur longer ene from the sme numeration. The colin adds ‘ther weight to the assumption that hiss re economy condition. Fu thermore, nel the let gap in extaishng (16), the tae deseiptne ‘oberation that recs apie: to [Spec, TL! The observation {otlows (nontedondant) fom independently motivated eecnomy cond ons 3 we would hope Nothing hasbeen ad ofr about quantifier rising (QR). The stats ‘ofthis phevormenon has been the apc of rch comrovery, which Ti ‘ot atempe to review. Soppse the eperation ents. Since 8 covert, it ‘ust be ete rising: quantifcatons festure quan ass to ain to some X°%* dha sa potential host (presumably T or wbich, we ‘ight ssn, have osional fx fextres lowing the to hes [a sppene 20 fqn nterpreable and therefore esd sot te checked ‘Thea eae af he foreign etegores is options, therfore cho sen if makes dference”ananng that the eeonom condition (76) ols fr this case, which eller rom ener we avez ar eosin ‘Sine ther sno PF ef, theres would be olen QR when i eas to sinc imerprctation, in some sce tat tus be made prc, an ea proposed by Reina (1998) in broader amework and dev! ‘ped with considerble empirical support by Fox (1994)." I falls at rly ino the framework already ouine. ne discsson fon te right track, then a vanety of apperent reasons for inclusion of Agr im the Iexial inventory have Bee eli nated. The question ofits existence therefore narrone tough not ‘inated Not alt rgumets for Agr have been conser!" The ‘usin Hat hae based onthe assurpion tht Agr has no features that thse Features se assigned to subeative Kei ems a8 they ae ‘raw om the econ IC Agr eta the cas of ears, a a een more etre role ad unig tat han belo, with no appar fet impet for the core compotationt proceses; that seems dubious, 21 feast Forte ease f sujet apreerent, these appatentfuncons of At ‘could perhaps be acornedated with the system just caine by as Simla it with T: by ssening, that i, thal as Tis drs for the levee forthe numeration t to i optionally ssgnad Gfeatures (as ate nouns and, Ihave so far ssumed, verb ad adjectives). Note hat {his caves bck 0 semaine the analysis that was convention ‘fore Poles (158) highly productive spi theo) hough now cor- siderably revised sud in que 4 diferent sting. Toe agreenen- base jusileation or Agr woul thercore reduce to adsesive and eect agree sent To carry the rater Forder ite necesay to lok ite a range of ‘empirical questions ha go well yond the compass of his ingir: As mates sind hr. sens reasonable to conjecture that Ay does tot ent and that fates ofa predicate P, though —Inerpreabe, ar ike he Hoterpretble fates of nouns hat they are pat of Pt the numeration aed opera as Ps suet om the lexan. m tapers AML Summary Reviewing vie ie scems that we may be ble to linn the theory of phase statue enti, derving its properties on principled grounds “Many consequenss fallow for te theonte of rivement und eon shen these concsons ae combined with eter mininalit esuiryins “Taking the ater eriousy, wea do aftr formation of Ue theory of the computational system tat rates form and meaning, and 10 a sharpening and improveren of economy nd ther cena We have been concerned with the computation Cy that maps 2 mu rmeraton N selec from the lescon to pi of imeroce represen tio (= 2}, at PF and LF, reepetvay. Ata airay pot in the [N-+2 computation, the epertion SpellOnt slips say phonological Feats, ich ener the morpholoniat component sre linesry ordre fn then ae Nanded ovr to the operations tat mep ther to = The N=, deviation procols ia large meat independent ofthe exis eco” rae thatlnguage i marifsted i senornctr eyes Ihe sole fect tay be the neal to construct derivative caine involve Categories to satis PF convergence (and more bron the existence of macy Mave i the Fest place). Apart fem the mapping to PE. Co ‘css te conions of uifority upd (useless. Kom Sits of Wo operations, Merge and AuractMove. which incorporates [Norge Ezonomy conditions tht ae in Inge part teal commputabi select among convergent derivations. Something lke Merge i inescapable i 20y langage seem, but the operation Atracy Move reflects peur of humia tngtage mong them the merphalog)-diven “lst reser” proprtis to which We Juve dived speci atention Fugsionalcatepores ad th forma fetus cee 2 cea posi ‘ion in the working of Cy. IF he general approach of section 410 can ‘be suttred, then the only functional categories are those with Fexares hat survive through the derivation and appeat atthe inet, where ‘hey are werpetel, OF the funciona categories we have considered, ‘only T, Cand D remain, Sons fates. hich lay a considerable oe jn vet manifestation ard anguage vation, ate maroy Tit in bution, We have See ne veason to suppose that N ot V, the basi ubstacive categories, have song features. The sength property ean terested, perhaps. tothe nonsbstamtve elements Tad that head Cage snd Tresfortons a the mskr projections wihin the lus, and to complements that serv as mood fre indicators bu ac te 139), The various refinements and simplicaions along the way sharply rove the problem of exponenial Bowup of economy’ cleuations nd overeome a variety of conceptual protien concerning Last Resort ‘movement and “shortest move" (MLC). They ako appear to be com fred empl in am atretng range of enue. Most importa. they fa out natural frm minimalist assumptions, which requir tht we ep sry to operations om fetes and cl lations emor them {capores ony deatvly taking ito acsount the ere Gxnctin ewe -nterpretbl etures that edstermind ty bate cot con tons. Henc. we have cnfemation tha the inescapable enditions ch language may bests in something ike 3 “ben possible” ws) ‘On isa assunpsions, hese ee 1 be he it kinds ove, thovg doubles net ye the ight oes Like ear fase bythe samme gous they ase many question ad i pause al for Boed (el of tahinking of what ha been esd. ‘More neal it seems nt unresionable o cone ha langage ‘may appoxinate x “perfec” eye nthe sue deserted inthe nt ction Iihis inci scerate, it would make pod vers to pres it {0 the its 1 se what cane discovered abou this erous and io ‘resingly mysterious cermponemt ofthe human mind. To progress for ther along ths ond we wl veo find ot theater fa wide range of empiial questions hat have come to view and that ae sometimes Formalble in furty clea ways. We ae let with hid and challenging problems of a new one of dept, and prospects fora theory of laine _uape wit propre hat are ite supe, Notes CO te backround of theca, se toto. Thanks parents the MIT karen offi 93a etc ary ae ices twit ms wa 3 soe er, dnt ceo ‘ce. Tanks parc we Cee Cali, Som Foss, eb Frampton Sa ‘Stir ot Howrah for cess on eat Sens a ete ‘efor Choma 19a oun Ure cee er gre ‘eons de vero. [The P bol ia xo ini ond unrated oar th prose, at ones fron the eu of te apn of ye es 0 BE EN sss p ay. 0 chased 2 or my on views on we 6 he ts. ee Corky 1936. 1, tone 2. Res ht he orig of patos astra, arene paste pep reso te engage acy othe bi, na tesper eee ed te ths royce he tenner at pr hve» eters er {fran ty ay tefl scree propeiet He Stina pers role Pet and Set oe msm er we Re col entra escape yt + tg phone ou amo. whatever Sa mht be rage) Pt wel er ot y rand ‘Seca’ 983) -eorhernt coon” Gee pase by Pre nd Sooke (993 ns ¢“somobvots humgon” hey hae “Rando ‘Svar. bt tic hrd oc ow ha cob oie septa only (Gini cues neninad) vara mpd tunptors reat arty, ‘Spine wit aneng uneven speci wal {conte ern ithe cpaaion to Pt rsig ie qesion [Fh hy er tan spesch Thy oy oe tr at theta ot poste posses (a ae hao ert en sepes ‘ah At ps ens el at Grete nd ete (9) are ‘Src eng tht pons tee fC ee ba, paps Spun tem some ee seri. SIs sms pop fei to the eat oe McCarty and Mice {50 apart sth concen ha hen) devcop Price nd Solera. thee neo fra pat er ah we). ee pon te cnn hy ah ing) of prtsion "Tey gs at ‘Sere poker arcs eel inte ory of neg Trach oes ‘ime dave ey sy do hota ea or ec tiry mee Bt fou fea sens mater nd and ecg fv acy These ‘och ned to add p-out renons sme ind (tues) Tad ‘eval sorting tak to Pal an eed n rete hoy tee ar Tod ph et “Tahini toms he moten "se ‘honda embodies abt ma recs and ‘eons an he cof es tnon mc, mas Seta ax). Meany ad Pi G93 prone that “tune he Tarced te inpuroupa codon, ba what iy get ses to hate Me ‘seme oe santa! pote ps -Memly ewes ipa md ep i te aly neve sd) To ake rt efi. tie om meri arr cn) rope spent hn of part ef ond ge. ns pcp en rae ‘ening ture ceo we cnt ok tow nape nes ef Ot ‘oy They cir eee a wt gers er rows wade (Tt ae ue ation ig ition a on al LF cargos a Tretorn, mI sea into nates So he AP sytem th obit hi ti. Buch gees wel! cat to id nto bs of nna os cig the acta pecs of reer and Bk) tare ale an poy uae ‘tes. Sx Chel) 1738 1B, adhe 998 1. Note tt cnserons of his ature eon be veh ony within «iy acpi mma oppreath Th wh erty eh ol Soke ‘tthe eortepar any cet (nde a eds} cn ey Be eos oe etre Thare bn ena tey bewse aa eg new Und fees und towing cer te Tonal dev we se Thi the (bly equines ae pred ey wb Teel ty ‘ome proper of neue 4 Lah ey caer wef tae in wih aes wa eu 9 For nef oda ith sme stk oesioes. te Et thing 10. tndewions of ei an trator are ot ontsne in eal ito Nova yod prt ofthe oun pone ats pel anu re Common ute eer. Under the Ds Morphol tee) ot ae ‘Sid Marne (193) and Nero), pe ars dot pena ‘lin Ge N= computa, wore ing Sngustd (yy dg ve ou ‘iia he phonsog ern, 1. One sch example, ake due dela of ire te, ‘stich muy meets eee iu Bing eons, sore Se seen 1. {gy Note that hi sein with cont th bide a LF asprettion i anaphor ring 2 Syecbe operon se scing $3 Iiplenenation of sah ideas fay sgn tw sip chal ‘ti mrs hy do mit efor at feos, a whe nde. 12 Teal guetion arte su they fe acer) aie 2 ure ofa ts bn hh. Te sips supine ers ‘ham are comida uni ne aed by an operation, Tht at tae toa uso ht ca bees en Nat ay ito ‘ows corse, 1 Thos pert ef the meta fr Dead Molo: ee mae 10 Pina ice Gc to the onary amir: for exe, Comme rer Romance perc of Enh wont the ato ae sae fre ‘coparatve conrad Gale shi Son pest tat see robes of nein elation the ese had at nthe roche ete Ta ot oy athe ts ot ress bot Fas ie roncnbre do 18. Tha, een formal feats 1pcly have seman ors at sani ope ace Cae ad aay. orem 15. Uns, of cours, be sem stows te Cie to na eo ‘meted aviv inthe phoenlogy a spice apa at wel hae w ae cgi arongy most be asi sesh, ting wl eer the pesbity ‘Ssed are hat “ure fe" might be aecesed tl terse. Te unearth AP nef igh cs an ented covey 16 "se ance naan wt be fed No. The onion wi be siy edi ston 410 tenet conto wil bel et Prat ight te tat for reases spr re are engi, tat | ope 1 Ei ‘Reh pon We wll tre othe mer 17 Foran exo ef tity tems of com ta wi ica i a spe eee Kaan 994, 1S Howes ean aed er ‘sep ha how ese cman eNotes seg sl be "ost pnt ton "ere by the operon. at langues sck conn afar eke, na a hyo eal nie ‘ttn fers them Ale react re the Kins of extra sinc on 32, income with Hae ad Keser (19%) Oey. 20, See Lak 95, whee es soa sat Ee ad Engi i, ad ‘ha Engi uals ir tom in ein ‘erat Tar te liner on temo the sy of ara gue Sona Se ote. 2. For example we sil hae mo goed phase suc er or uc sie ‘ane eae ast ene lund aan een 23. The nc fore ws sumption dpe om popes of pass at are ‘asic We ests tose han er 24 See Fu 186, Spe 196, Ont 150, al rein 192, mone ene From epemial pst of othe seating ed abo 2 ets {hat spe tote, ul om a deri ponpete 2 St poh ‘here rel que rats, ee eb eng "Yo hat wee as ‘ea ere ihe at sr sge derailed y the operon ‘hc cr hee ie. 25 In present ems elation of crane CM ty end i tet ‘eine teose Tats bead Hog of EM ote ex of ahrton ot a tot ant cm behest refoc moworet, Furr quetion ae thet Tein neo td eer nse Optinaly hw oo shoul ede we ‘srs ne Chon, SP), pe a mach prin i se 26. Noting event changes if eet ia toe comple consrsicn| item femurs bu me te Categories a Toesforatons 8 21 As points ou by Cre Colin cal qucon ai i head ov ‘deta es wre sak con pet oe he ae ‘abe ters th nen ir efacd-an savant ought onc ‘ha states Weean le he tne wnt, edn ne sg, tons et ope tou theatre oe 2 Auriga be HNC is val pin, wich i wo eu. We wl ‘murs gute 28 pee pry hc a fer or Chin ‘peat ta cn be “ens ea seamae fo ater} hep to ‘hep ese of wommenberel chars: We i ea oth ae. 20 necting he ner cation mote ely we mis ht the ee sith ror Fie (F(T K)} (2a) overcame the tempat to tisk fo Iva sre dso sate. at dom te pate ce pat (i ure vara). wi oom sch en Bi, Nave hah 2 voli ofthe nae condo, The serane is that th epefeton aes pce wit he phovlgse compete sing ck betwee sme evel whith component and LF (eae 13) Fre ‘ners copy inom ee Tce 192 22. Na tobe confused with he worm conto on dshtons dius ‘air, or eer ther fete of "uaori” cae caer I} oh tenn though na row chin fee whee heed owed cence Se isimorecoumds the potion fan oer ac 233 Sex Cline 199 for ns pyar oot cy. 24 Sex capers 1-3 ad Cheoaky 19865, 14, Cah 10 ape, Un 954 235 sl ler serio wold allow he operon tie ofthe site ‘art incerkea Yeo hs pity, mer dagen i 20) ey. 24. In carer work (secon 35) aed sch cee Jain ane ¢ ‘hater er le of perl ta To tts Ba tt Tne for sonst wh ve wl xin, they ry een We nen wesc of pce popries feat Case 37. See Chay 196 nd cet work bd on mich ber pcg and "opblepel ede parti nd wort rare) 3. Noe that Os modifi conform ete hry developed io {onto spree 29. Ths is margial phenometon a hes sie Merge does ot ee ‘ase ceckig doin We ter oi ay evant peepee ‘sued by se mere portant ef Move {The tac mumpbon here in tha on Monee nays as pil fing fo ener. As need eae, tha each too mone Ie eh ‘vo shou be roped tar seers th ot am ‘chapter 41, An ub ih ai if he fs Fat ite ene i testa se le et hae io econ od o saan in Spel sume ita is oxime a us Prt Go msc SOc Setter tec sone appt tment {2 that edna ef WP cor wih ogre ti gus he pts son soa nek rh Shay aie ns se gn or pis Fer ‘Soci lay fees new ne sk resem ete Seat a sol coe rete mc fo Teck roy he ak Sth ah may SSuctimat neces sna ow ore A ar amare te {$5 ed ons ry eran oe Suu Ringe ech. ty be [5 Avs at nen, gesnen wth at nonce ren tec fou eh er yan oe (Sieunirs aise Sheraton heer cm ort Mov conte eke eo et seers ay ms Bae os ea 1S ome nig tovos- lee ipo ei Seb ate ST ee rsa Bp a ‘owe neue) cdo separ o Oe BN ee TES Saehns e's prop nr ce ropa en N= {sh than hh Egor an ae Seago rn Se Nor ot rete oe [Socom 2 efor po iu Nee shee nur of he deters ls Tatu fry te tng et Sng mee oe cca ait ine mec oo ges ‘Sinton Some t? {St wana Crd and MiSs fie at Ty bane ‘av eo ve mcune n Ta esos SarvmmteCaFese Oey ton meget ep the rd an tah ay ga ‘nthe pena pimonevn, ee Feonater , Alsin (3 {hem Cn, golem. The ne arte ‘rite (Dizon Hons pero corm) and Maina Seadineioe {Geom seg Fr Ta) fr Teamsome wn une ouabain La (Cage sed Trasfomations ws i eee hy corte tuto itd in) Dom het echingdonain "rie minds so eva. supped telow. ts ne [ator sop. bt Twi lave he mate fee, Real th the peer ‘ijt dstton docs not ge erie wah he AA dsion ese at ‘ove motion eae fr the nan of enue UF sec eet upset whan eto robles fe ECP ed hd Matis ha do i (eas nh Tora hn mos ernce e. 4% Cine the ream 0 sare come. we rah ext tat cto ‘aur ashi weet ee apes muy der se ecding ot. The aa comes dee te 2 ening i oe at of Wa GN wh rok ‘Sho f Case aspen yeti offerte cree y the opto, The "heey ca te seped hy acing foun snow toe ea {ta dena Case ane of Van Fi ot ele y ose sagen at ‘nt be deed uci fer somerzence. wich rogue tang he ‘hcg damsin feo caer Sere dan Skene ged ‘o entre thatthe ne Case of V aed Tse” ly ae be ‘Sess Cis nonce ese Fenre cathe deat sth 50, We are eacing anion hae tee N= caption, psig we heft tat ~ tepals iy have» PE ree That tae ‘ee ceckingreton cnr wich ae ree ese ae towph thar phonetic rex sexsi Tee msc af mays 10 poe, ‘depending parton the rower 0 aisons open show @ fees ef § ites ene would boone vec of een coms ot ropinsctpa proper Race sp! tt Set On 31. Awami shut o paths pu. The pf esing i posse fly the agcine raves on 0 Ar a cnnpse eure of Ego, wh ‘Ea Apr heghout Me nl ren toe room neon 410. 53. See Cards and Guat 191 for er eee fr a stn She (55) On the analy nar the rer tory had on Mone a Crd ‘Choma 18ts. Not tha Iam evoking hr etry sip ile een cen sure ad aes seers 39) fy te frm ‘tug pon fret) The str of Fl seven LF bu et ‘hoe neti fe ch anes fer ching 53 The whoa const en xg (ene, propre inal seven ate, Pps hy nlf a sucess el ‘be mtg, Jet oder He ta melee day ees hat appear peo 4. Thetis em oot ey to determine, eae of may inefring ton- Se Gru ane orth 18 fr dnc of ting fom sje in Seip i Romanian: agement east nthe embed ln ‘Sf oe can shure hat somantve Cue sted wih pesos oft ‘hrc tect wl fit the sae pater The ater Ot oa casper sigh a oner Went’ (1953) thor of Cae vation a a 400 ‘hese ti snene on able ease ems ‘ceo usta Oe o crpenerne, sn ener Seton ‘teones Thank tal Gis (or cop tea 48 docsion Foe Soyee ny sr Une no pina ers that Aone ‘ca cnted aoe we cry of ro Urn 994) 5. On the lane se Lasik 9568. Bring csi teil nd ets ‘taht alow Asmovenan esa ern Case posion They. homes. tee aso wy tt bared ep oe verson of the foray ‘oie nese i Chom 18H 193-194 sition te moun pets Sith lnguesspeie dealt ase sh prune nod not be chase esl hat Cae hry ce er eco state Cae Into Cone. wich ssp oly 4 Gren tothe Case ais | ‘Saco phowinenon I bag! bd ao reid. aly an te ‘ino ther (he “uoriyconion” af Choraky 1885 1831), "ough somes io rtrd Cae pions, various nays. The ici ‘noting ae ut ts ot rn shown ou ary Benign tetra Ce ‘heey te pperety ge a popes wert esas 56, Se Reine 181- Ore pec ep, ee Ure 1 Ao see Miyepae 19 eeu 198 59. Tose mull te checing i erat to el verb constve ‘ety Cote 89. SO be inf annoge re se ft of ple Be. ce son 49, mana hed ery vet oem lt Ja! held so ‘che the counterpart ven tgsges et pm 3 59 Ashe snp ae cried, we ld Bea o dss he posse ‘il come uch loot epee hve an fam tee sermon See Cestata 199 fr ose Ent confemation nt ferns te etre coe! of alee. {61 Pipe an cap ott serio” ae see pope ty Hale rant, Maree(198). (2 Onthismmter, 2 pactly Cho 19. (6. Witham excepion to which we wil em sy ep. 22 4 For os. the 1-0 eptn is reste in agit oot caso ‘Shatner we wl etre eu ugon tot hy ths mig be tea Stoner te aon othe ey th suse the tore ee tv reter bt hw vane ier cee yay win he eeepc Eenipenent (Both Rehr a Tchr tat te gs spent ane iis ‘Sacral sn arpa dino tie they propos, on tx ie a Svat ka pio fos vile tat be hat tespcsaton A ‘Renton sata ede Spc nai ae Ta coven nga Cotgoes and Tamfomatons » 6 Oe nigh expec that some pscopen cordon. paps 1m Seahyos to tbe Waco popes ans othe eno fs teak ‘higtor proce tient vray e188 Ere epee ‘etcas of fonranen (hater he posi arnt tala ere {Tha th comparsns ih et to oy cotati i gpd ‘Gotan 923, Rechar (19), eo (05) eo conection ofa of te serrate ip Rls wes nd CPi ee eC 159 = “ Soe Chomsky 934, ste 59, Un 185, eee el. For met Sscrion ef suey etal opr Tema, Ejeet Kone 19ah The wtf te stecrenon hot te Cer Snr agen, sochastheconrast Been a Ot ed owe ae Oe whose ata Bho id eu pesado tat © suture ose {4 St yu perme soo to the acaba the wba a a ca send cole ing ita fr hs reson lee he ea ces ede wie at pre Sys enn, Oe opens ha vee ed a ee eo conan hes terms 1 pan yo teal wh ke ethan edd ope ae (© who wonders at ho og “iat old be oo ach, Ronee, releing + pee for wciton of ‘ho Thin we opremtesmersion n (Gi shat censor wo peak ‘here areany te roe. 18, Se Choos 8 osc oe th his rob, nw ues. 21, The davai f 0) sis fc wo reso Ja ack aaa ‘Tsoi Ce. The ier pope ero se en). 1 te bral nese of eee abot, he gpench ieee mis be laste onthe snp tha ¢Sareton tf sehen re ih av a (0 ec ctn of apron in 0). Ba at Tee pcs ae he {26 otic do na sor 2 For sme fens and ein te ico 14 posite hat he yet ie fr argues hve co whe opie oF Se spine es of Resa (51) wor etal long ner ey ne 11. olios tse sgzesion oJ ramon’ Ser al po. SK 25. epeding on actly bow tr eprint ened the mane fer of mrt spe pan ‘nut fn abo hun ke rene ere ee we capi gt ton oh prach we a tine i he ‘pat rk nytt ed a LF hat se em he ss ‘Fup (ana? We hos crv eopres ef he ECP ae teeta ‘Ecotu publors of saber fx eb 1.1 (6) ad tet) a acon cS) Noe ue sa th prar coef cn ustorey Sac re ‘ihn oly os evr formty these ds as “oni {iy ih rapt PF se property. The oper deal tee wast Fetes An coupe props ene tan te pepe {tengctiers Sager ra potion Exons of he eas tht cae thon he neporaton a nym ces song a.” 176. The orclaion gus Beyond ha as Len shown, wich ely th {ee ofa srg att Be ose 71 The face oie ot a dr rion ton of oni fs icing “tes ant eter poet sertond ype. The ion of gee testy sper ceantenent ches Mes to nly ened nO ut thocondusone re asomuble sd resupply by heft ‘corm Ch sing the Pic Rncon 4 DP with Ped. SEIN U?eine manor oth does net fet exons Even th ih eb» o which nm basa eid very en {0 Lavin Prceyge langues tha woe tal the ane. The rae Ital in ofthe con nih terns he "perboed o F ‘ie foe dtcring eens dest cont 49, See preling te, Thanks 9 Visa De or expan forces Treva prcad ux wh she eer eomdeaty ore coe ta edema beret dd er nd many eer factor od. rng thom che ech: ers Gen ordering of PP nd las ‘Sent, nd om chess (wich ge shrger tne) ve tN: fms Baier Teg tbecrang. se psa! cman pei {hatwide 0 Spe. the alone of 96) mar sexe a sb stm the dbl cet ese a 3 Thue ech br mies race ago toe) 50. One omy ltt uo, dace in Lasik o apes a to do with the eposoty of auch exper re yy econ gh ith he Tune set lug oc fame wo eam er sh epee tthe ftloog Hla Key (19980), Socres of up Ih for do XtL icy ie ane eae contacto of ee 2d or [tne to wich wow turn th cane 5 we wilt sujet eee ‘eu [Spe ich stone Te guste sl ea or Enc. ‘Tore sso ete hehe ht ever pope roe some 1 ‘rns ety en VP babe ston ais 11 Ur Atg bso Dts tha cet jc ning ete fr engl sec Keun 183 190 for teary of the gr sont Va a ln the ora ht ie suc ote. Sate eons | Cage an Teetematone a {© usization i ete foe muti Spee consrtion, eitig t- 1 We ase tat i in the cating demain of HOR fe te. Tht feos fm he eft of Atacov houph rtons emai she the 1. Can Babes to ura cenit to Ole spat ie he oan adel by 2 I a the ston would be arog to anes cece ‘sd fB— Tor Vinnie, Ba Te propel mote spt of te es tht Oc are ob ewer congas dw es “Gril robo at we ly |S Inthe cre they flo Mone ed Gia te operation nul fe ernie isvegten ft hectng rey facia ee ers ha he compris edit 11) 6a nx quay eration ‘Choy 99. Cen (1938) rons ta tah ie mph ste und a sia of tts awh we es Farle pec oe es nse Hedy 19% Boe 19, 46 Thecieusson of meomtrcog ux oem vent Fincher gt ‘ie of arunent espn the concen ht dation Gash ste 157 An angen oh omar in chp hick tate Form Chin, 90 {eng btu eran of a Reson gen ae 8% Chachng cry aples ntha aw at shee Ft se op ‘ins desea the end of acon 45 Ie Ge te fe opto the ee ‘Srimert apis. though to. move righ sss th rng for arp th ‘enaon D tha ys Jot ome ie Bl wih the te onion ‘in, Wockog ihe ore vation Wel aie Joke 4 eae wt 19. They tetra arpumortinave me. howe Se, One cs of Baris praizntion tere ined ecuned for (tao ar iach eoxrto thes meta Th cc ese fetraieritan rca Cow) Te te etm ane wre ee oe ‘huh gtr eer rg tht dom net spn seat Cos ‘he hypothe xan, The jt ho a oe stg pose to rase Case, surg the Soe emenement ht as «Dw Greg ‘eons (oe Us 180, 5. Tec is et ite co ee 52. Se Mier nd Choy 196 rd Choy 195 for ee 9. Note. houeve Aloka teat wk he 950 sows "hs ey pnamens tn fel wag he sch ere spain hee peste ‘tate with sscunts of be rt a ky ss ‘al win te mach mee reat tna bth ar ‘liner aun) and rein enon the bee an of contro Re ‘eto pric ennai nat Sree 22

You might also like