05 08 23 29AUB Foreign Languages and Literatures 2014 No. 2 PDF

ANALELE
UNIVERSITII
BUCURETI
LIMBI I LITERATURI STRINE
2014 Nr. 2
SUMAR SOMMAIRE CONTENTS
LINGVISTIC / LINGUISTIQUE / LINGUISTICS
GABRIELA ANIDORA BROZB, A Few Remarks on the Vowels of

Kenyan English ..........................................................................................
ANDREI AVRAM, Syllable Restructuring in Early Solomon Islands Pidgin
English: An Optimality-Theoretic Analysis .............................................
MARIA AURELIA COTFAS, An Objective Look at Object Control Instances in
Romanian Subjunctive Complements .........................................................
SABINA POPRLAN, Le verbe faire pluri-fonctionnel et ses quivalents en
hindi et en roumain ......................................................................................
DANIELA BORDEA, Dynamique du figement : ladjectif en franais .................
FRANCESCO VITUCCI, Gairaigo within Japanese Language: Language
Suicide or Casual Culture? ..........................................................................
JING DENG, On the Diachronic Development and Pragmatic Features of
Chinese Public Signs ...................................................................................
ALINA BUGHEIU, Linguistic Landscape, Microtoponymy and Unconventional
Use of Anthroponymy on the Border: Vama Veche, Romania ...................
119
*
Recenzii Comptes rendus Reviews .....................................................................
Contributors ...........................................................................................................
135
153
3
15
37
63
79
97
109
A FEW REMARKS ON THE VOWELS OF KENYAN ENGLISH

GABRIELA ANIDORA BROZB*
Abstract
Kenyan English resembles other African non-native varieties of English in that most of
the vocalic phonemes have undergone restructuring, resulting in a system which has far fewer
vocalic elements. Phonetic evidence is provided in support of the phonological assumptions
made. The acoustic analysis of speech samples obtained from 10 subjects highlights a number of
characteristics of the vowels of Kenyan English speakers, although the amount of data is
admittedly far too limited for generalizations.
Keywords: monophthongs, ATR, variation, vocalic processes, acoustic measurements.
1. Introduction
Research so far on English varieties in East Africa has taken the form of
studies on English in the area as a whole or as a cluster of two or three varieties
(Abdulaziz 1991, Kanyoro 1991, Simo Bobda 2000 and 2001, Trudgill and
Hannah 2002, Mutonya 2008, Schmied 2008).
Kenya features a multilingual ecology made up of 69 indigenous
languages, according to some sources (Ethnologue 2009), or 40 languages,
according to some others (Abdulaziz 1991: 391). Crystal (2003: 107) writes that
Kenya has about 2.7 million speakers who use English as a second language,
but, according to the Ethnologue (Lewis 2009), more than 65% of the total
population of Kenya use English. Of course, individual variation in the use of
English will depend, to a large extent, upon the quality and quantity of exposure
to the language combined with the attitude of the speaker towards English.
Schneider (2007: 196) notes that Kenyan English (henceforth KenE) is during its
third phase, i.e. nativization is still going on and the language is spreading gradually.
In the present study I analyze speech samples obtained from 10 subjects,
from the Speech Accent Archive (SAA), in order to verify some of the features
of KenE. In processing the sound files I have used the Praat software (Boersma
and Weenink 2010). The phonological standard used for reference and ease of
University of Bucharest, Department of English, brozba.anydora@gmail.com.
comparison is Received Pronunciation (henceforth RP). The standard of

comparison used for vowels consists in the lexical sets in Wells (1982).
2. Previous studies
The research carried out by Schmied (1991a) has served as a yardstick for
most studies on African English so far, especially for those focusing on East
African English. His comparison of Standard British English (SBE), in Figure
1, and two regional varieties of African English, namely West African English
(WAfrE), in Figure 1, and East African English (EAfrE), in Figure 1, proves to
be particularly relevant as a point of reference:
Figure 1: The vowel systems of SBE, WAfrE and EAfrE
(adapted from Schmied 1991a: 61 and Mutonya 2008: 437)

Vowels
systems
African
systems
tend to merge in East African varieties of English towards vocalic

made up of five vowels, compared to the seven vowel system in West
Englishes. Also, even though most of the vowels in these simplified
coincide, the mergers are different in one variety or another. Finally,
GABRIELA ANIDORA BROZB
the mergers may also differ within varieties of a larger group, as will be shown
in the case of KenE among East African Englishes.
Mutonyas (2008) study is particularly relevant as his findings disconfirm
some of the data in Schmied (1991a) regarding the vocalic system of KenE as a
variety of East African English. As the Figure 2 below shows, the differences
lie in the STRUT and BIRD vowels which merge with the TRAP and START vowels
to [a], rather than towards [o] for STRUT vowel and [e] for the BIRD vowel, as
predicted by Schmied (1991a).
Figure 2: The vowel systems of SBE, EAfrE and KenE

(from Mutonya 2008: 442)
3. Present study
In this study, I aim to look at the mergers in Mutonya (2008), to discuss a
number of implications that acoustic measurements may have at the
phonological level, and to show whether there is inter-speaker variation (and

which are some of the factors generating it). Finally, I touch upon some vocalic
processes which are specific to East African varieties of English.
The analysis is based on 10 KenE speech samples1 from the SAA corpus
(Weinberger 2010), all of which are in terms of style the reading of the
following passage: Please call Stella. Ask her to bring these things with her
from the store: Six spoons of fresh snow peas, five thick slabs of blue cheese,
and maybe a snack for her brother Bob. We also need a small plastic snake and
a big toy frog for the kids. She can scoop these things into three red bags, and
we will go meet her Wednesday at the train station.
3.1. Results and discussion
The KIT and FLEECE vowels are almost invariably rendered as [i]2 in
KenE. Consider next the acoustical analysis of the KIT and FLEECE vowels in
KenE. I have cut and pasted the words thick and meet, belonging to the KIT and
FLEECE lexical set respectively,from the corresponding sound files of the male
subject Gusii1:
Figure 3:
KIT and FLEECE
in KenE (Gusii1, SAA)
The two vowels are similar in terms of duration: the [i] sound has 159 ms in
thick and 131 in meet. This difference is not statistically significant and it can
1
The names of the samples indicate the mother tongues of the recorded speakers.
This /i/ is short, but also [+tense], a feature which characterizes long vowels in RP
(Brozb 2011 and 2012).
2
be explained either as a result of phonetic environment or as an instance of

intra-speaker variation. These explanations appear to hold as I have also looked
at two examples of the GOOSE vowel3 produced by the same speaker of KenE in
the words scoop and spoons: the sound [u] has 106 ms in the former and about
150 ms in the latter. The two words have been cut and pasted onto the
spectrogram below:
Figure 4: GOOSE vowel in KenE (Gusii1, SAA)
Given the expected length leveling observed, the first step taken was to
measure all tokens for the KIT and FLEECE vowels so as to ascertain whether the
length difference is indeed unimportant.
The measurements of the current data show that things are not quite so
clear-cut. There is a lot of variation whose sources may be multifold.
The table below displays the realizations for the potential targets of the
KIT and FLEECE tokens in the text for all the 10 speakers whose samples were
considered for this analysis. The length measurements show that the short
monophthongs vary between 40-60 milliseconds, while the long ones exceed at
times the double of those values, varying between 110 and 160 milliseconds. I
have delimited some of the cells with borders in boldface in order to underline
breath or sense pauses, as they should be marked during the reading by the
subjects recorded here4.
The speech samples at my disposal do not comprise any words which could allow me to
make any clear statements from a phonetic point of view on the status of the FOOT vowel.
However, it is my intuition that the FOOT/GOOSE pair behaves similarly to the KIT/FLEECE one.
4
This will prove important, as they do not always overlap with the ones made in RP.
WORD
please
bring
these
things
with
six
peas
thick
cheese
we
need
plastic
big
kids
she
these
things
into
three
we
will
meet
RP
target
vowel
F
K
F
K
K
K
F
K
F
F
F
K
K
K
F
F
K
K
F
F
K
F
Gusii1
F
F
F
F
K
F
F
F
F
F
F
F
F
F
F
F
F
K
F
F
K
F
Gusii2 Kikuyu1 Kikuyu2 Kikuyu3 Kiswahili3 Kiswahili8 Luo1 Luo2

F
F
F
F
F
F
F
F
K
K
K
K
K
K
F
K
F
F
F
F
F
K
F
K
K
K
K
K
K
K
F
K
K
K
K
K
K
K
K
K
K
K
F
K
K
K
F
K
F
F
F
F
F
F
F
F
K
K
K
K
F
F
F
K
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
K
K
K
K
K
K
F
K
F
F
K
K
K
K
K
F
K
F
F
K
K
F
F
F
F
F
F
F
F
F
F
F
F
F
F
K
F
F
K
F
K
K
K
K
K
K
F
K
K
K
K
K
K
K
K
F
F
F
F
F
F
F
F
F
K
F
F
K*
K*
F
K*
K
K
K
K*
K*
K
K*
F
F
F
F
F
F
F
F
Nandi1
F
K
K
K
K
K
F
K
F
F
F
K
F
K
F
F
K
K
F
K
K
F
Table 1: The KIT and FLEECE vowels in SAA samples of KenE

Legend: K = KIT; F = FLEECE; F = FLEECE, K* = KIT*
Note:
KIT* = we + will > well
KIT: 40 - 60 ms
FLEECE: 110 - 160 ms
FLEECE: > 200 ms
Table 1 reveals that the difference is still maintained between the KIT and
FLEECE vowels in most of the cases. Therefore, can one really speak of a
KIT/FLEECE merger? Before moving on to clarifying this issue, or at least
attempting to do so, a few more remarks are in order. First and foremost, the
pronunciation for these is sometimes (in the cases of Kiswahili8, Luo2 and
Nandi1) [ds] rather than [iz]. A twofold explanation can account for this
unexpected pattern: one can expect an invariant use of this (which is neither
unexpected nor uncommon for a variety of New English) or we could be
dealing with a case of allophonic vowel length which might function as a cue
for final consonant voicing, in the sense that the voicing of the word-final
consonant could be predicted on the basis of a quantity-sensitive distinction in
the preceding vowel (see Roach 2000: 50). Secondly, we encounter a lot of prepause lengthening5 which is visible in the extra long FLEECE tokens (marked in
boldface in the table). This being the case, one could claim that they could serve
5
Attested in Singapore English (see Deterding 2007: 38).
as prosodic cues. However, this is not what happens as their placement does not
serve prosodic functions6.
Two speakers (Luo1 and Gusii1) display a similar behavior. What they
have in common is a late English onset age, namely that they started learning
English when they were 12. In their case, KIT tokens are used only in functional
words. So do the two vowels merge to a single vocalic position in their speech?
Let us consider the sample of Luo1 first.
Lexical set
FACE
DRESS
NORTH
THOUGHT
GOAT
KIT
FLEECE
GOOSE
TRAP
BATH
BIRD
STRUT
Word
snake
Stella
store
call
go
thick
meet
spoons
plastic
ask
her
brother
F1 (Hz)
502
577
557
591
448
435
316
409
745
766
758
740
F2 (Hz)
1949
1752
1150
1135
1072
2131
2099
987
1463
1437
1397
1433
Duration
(ms)
153
150
270
151
252
150
115
204
107
128
190
106
Table 2: Mean F1/F2 values of Luo1
The table above shows that there is no length distinction between the vowels
that make up the KIT and FLEECE pair. However, the vowel plot below will
uncover some interesting facts.
FACE
F1 - Backness (Hz)
22
00
21
00
20
00
19
00
18
00
17
00
16
00
15
00
14
00
13
00
12
00
11
00
10
00
90
0
DRESS
GOAT
600
700
800
They occur rather randomly than at sense phrases.
F2 - Height (Hz)
THOUGHT
400
500
Figure 5: Monophthongs of Luo1
NORTH
300
KIT
FLEECE
GOOSE
TRAP
BATH
900
BIRD
1000
STRUT
Luo (Dholuo), which belongs to the Nilotic branch of the Nilo-Saharan

languages, has two sets of vowels, distinguished by one main contrastive
feature, i.e. [+/ATR] (Owino 2003: 45). Therefore, the expectation, under the
influence of the mother tongue, would be for the KenE spoken by Luo1 L1
users to have ten vowels.
The results of the measurements show that there are nine vowels (eight
plus one coming from FOOT7, for which there are no tokens in the text). Two of
the monophthongs have emerged by the monophthongization of the diphthongs
in FACE and GOAT. As mentioned above, there is no difference in terms of
quantity. The [+/ATR] distinction is preserved in the high area, it is restored in
the mid area by the monophthongization of the two diphthongs, and it is lost in
the low area (based on the data available). It is worth mentioning that the
association between vowel length and the [ATR] feature does not apply here.
The following table displays the results for the Gusii1 speaker:
Lexical set
FACE
DRESS
NORTH
THOUGHT
GOAT
KIT
FLEECE
GOOSE
TRAP
BATH
BIRD
STRUT
Word
snake
Stella
store
call
go
thick
meet
spoons
plastic
ask
her
brother
and
F1 (Hz)
429
484
543
532
474
318
311
327
709
657
676
654
535
F2 (Hz)
1939
1699
1224
1207
1011
2014
2046
905
1403
1386
1408
1429
1465
Duration
(ms)
194
114
138
130
238
159
131
150
128
118
170
111
113
Table 3: Mean F1/F2 values of Gusii1
Gusii (Ekegusii), which belongs to the Bantu family of languages, has

nine vowels, and the main contrastive feature is length for high vowels, while
mid vowels are distinguished by means of the [ATR] feature (Cammenga 2002:
37). This would allow one to hypothesize that the KenE variety spoken by Gusii
L1 users will have a vocalic system made up of nine vowels.
I operate on the assumption that the GOOSE/FOOT pair behaves as the KIT/FLEECE pair,
which is the case in many non-native varieties of English. For an in-depth description of the cases
of Cameroon English, Black South African English and Singapore English, see Brozb (2012,
chapters 5, 6 and 9).
10
FACE
F2 - Backness (Hz)
2000
1800
1600
1400
DRESS
1200
1000
800
NORTH
300
THOUGHT
GOAT
500
KIT
600
700
800
900
F1 - Height (Hz)
400
GOOSE
FLEECE
TRAP
BATH
BIRD
SCHWA
1000
STRUT
Figure 6: Monophthongs of Gusii1
The measurements show, however, that there are only eight vowels
(Figure 6): two originating in the same diphthongs (i.e. FACE and GOAT) which
have been monophthongized, one comes from the paragogic schwa and the
remainder are the result of various vocalic mergers. The length distinction in the
high area has been lost. In terms of quality, this time the KIT/FLEECE and
8
FOOT/GOOSE pairs have merged to one vocalic position. The [ATR] distinction
is restored once again in the mid area with the help of FACE and GOAT. Just like
in the case of the Luo speaker, the BATH/TRAP/BIRD merger and
THOUGHT/NORTH merger are to be found again. As mentioned before, schwa
occurs sporadically as a paragogic vowel.
By looking at the vowel plot in Figure 5, one can see that BATH, TRAP,
and BIRD vowels have merged towards one single vocalic position, as is the case
of THOUGHT and NORTH.
3.2 Some vocalic processes in KenE
In spite of the low number of speakers and the shortness of the text, the
samples analyzed in the preceding section are illustrative of some recurrent
vocalic processes, which are characteristic of new varieties of English in
8
I suppose this is the case for the GOOSE/FOOT pair as well. From the data in Schmied
(1991b: 424) on the Africanisation of RP phonemes in KenE, it appears that the RP GOOSE
vowel is more or less approximated in KenE, whereas the RP FOOT vowel is lengthened in this
variety.
11
general, as well as of others, which seem to point to the uniqueness of KenE

among new varieties of English as a whole, and African Englishes in particular.
Consider the examples below:
(1)
a.
b.
c.
d.
quickly
Wednesday
and
toy
[kwikili]
[wnzdei]
[and]
[tju]
(Simo Bobda 2000: 263)

Luo1
Gusii1
Gusii1
In addition to being cases of vowel epenthesis, examples (1a-b) also illustrate

vowel harmony in terms of the [+/ATR] feature. This may be accounted for as
an influence of Luo phonology:
(2)
a.
b.
c.
[ATR]
[k] to prepare
[kl] to scatter
[kn] to refuse
[+ATR]:
[iko] to bury
[kelo] to bring
[kuno] to preserve
(Owino 2003: 57)
Further evidence of such harmony effects comes from the adaptation of English
loanwords in Luo:
(3)
a.
b.
c.
d.
conductor
cinema
station
television
[kndakta]
[snma]
[sitesen]
[telefison]
(Owino 2003: 89-90)
Note also the occurrence of paragoge in (1b, c) as well as in the following

set of examples:
(4)
a.
b.
c.
book
hospital
spring
[bk]
[hosptal]
[sprn]
(Schmied 2008: 162)
As is well known, the occurrence of paragoge has been pinpointed as a hallmark

of East African Englishes. As far as the nature of the paragogic vowel is
concerned, there are three main options: the default vowel schwa, as in (1c) and
(5a-d); a vowel harmonizing in terms of the [PLACE] feature, as in (1d); a vowel
copy, as in the examples under (4) and in (5e).
(5)
a.
b.
c.
d.
e.
bring
end
raised
told
went
[brigi]
[nd]
[rezd]
[told]
[wnt]

12
3. Conclusions
The current sociolinguistic picture shows that KenE is still a non-unitary mix
of elements from the whole pool of features (in the sense of Mufwene 2001).
The analysis above has highlighted the existence of considerable
variation with respect to the vowels of KenE at this stage in its evolution. There
is both intra-speaker and, more importantly, inter-speaker variation, which
undoubtedly reflects the influence of the first language of the users of KenE.
This is not surprising given the diverse and complex linguistic ecology of
Kenya. The predominance of Bantu languages in the ecology as well as the use
of Swahili as a lingua franca in the area will determine the selection of features
based on their frequency in the future development and stabilization process of
KenE (cf. also Lim and Gisbone 2011).
It remains to be investigated whether factors such as less careful speech
styles, different levels of education and non-interference with native varieties
will lead to different results and will have different implications.
REFERENCES
Abdulaziz, Mohamed H. (1991), East Africa (Tanzania and Kenya), in J. Cheshire (ed.), English
around the World: Sociolinguistic Perspectives, Cambridge University Press, Cambridge,
391-401.
Boersma, Paul and David Weenink (2010), Praat - Doing phonetics by computer (version 5.2.03),
http://www.praat.org/, retrieved on November 29th, 2010.
Brozb, Gabriela (2011), The treatment of monophthongs in the New Englishes, in A. Cuni,
F. Florea and M.-O. Punescu (coords.), Regards croiss sur le TEMPS, Editura Paralela
45, Piteti, 153-160.
Brozb, Gabriela (2012), The Phonology of New Englishes, Editura Universitii Bucureti,
Bucharest.
Cammenga, Jelle (2002), Phonology and Morphology of Ekegusii: A Bantu Language of Kenya,
Rdiger Kppe Verlag, Kln.
Crystal, David (2003), English as a Global Language, 2nd edition, Cambridge University Press,
Cambridge.
Deterding, David (2007), Singapore English, Edinburgh University Press, Edinburgh.
Kanyoro, Musimbi R. (1991), The politics of the English language in Kenya and Tanzania, in J.
Cheshire (ed.), English around the World: Sociolinguistic Perspectives, Cambridge
University Press, Cambridge, 402-419.
Lewis, Paul M. (2009), Ethnologue: Languages of the World, 16th edition, SIL International,
Dallas, http://www.ethnologue.com, retrieved on December 22nd, 2009.
Lim Lisa and Nikolas Gisbone (eds.) (2011), The Typology of Asian Englishes, John Benjamins,
Amsterdam/Philadelphia.
Mufwene, Salikoko S. (2001), The Ecology of Language Evolution, Cambridge University Press,
Cambridge.
13
Mutonya, Mungai (2008), African Englishes: Acoustic analysis of vowels, in World Englishes,
27, 3-4, pp. 434-49.
Owino, Daniel (2003), Phonological nativization of Dholuo loanwords, PhD dissertation,
University of Pretoria, http://upetd.up.ac.za/thesis/available/etd-02092004-112729/ un
restricted/00thesis.pdf, retrieved on March 1st, 2012.
Roach Peter (2000), English Phonetics and Phonology: A Practical Course, 3rd edition,
Cambridge University Press, Cambridge.
Schmied, Joseph J. (1991a), English in Africa: An Introduction, Longman, London.
Schmied, Joseph J. (1991b), National and subnational features in Kenyan English, in J.
Cheshire (ed.), English around the World: Sociolinguistic Perspectives, Cambridge
University Press, Cambridge, 420-434.
Schmied, Joseph J. (2008), East African English (Kenya, Uganda, Tanzania): Phonology, in R.
Mesthrie (ed.), Varieties of English, vol. 4, Africa, South and Southeast Asia, 150-63.
Mouton de Gruyter, Berlin New York, 150-163..
Schneider, Edgar W. (2007), Postcolonial English: Varieties around the World, Cambridge
University Press, Cambridge.
Simo Bobda, Augustin (2000), Comparing some phonological features across African accents of
English, in English Studies, 83, 3, pp. 53-70.
Simo Bobda, Augustin (2001), East and Southern African English accents, in World Englishes
20, 3, pp. 269-284.
Trudgill, Peter and Jean Hannah (2002), International English. A Guide to the Varieties of
Standard English, 4th edition, Edward Arnold, London.
Weinberger, Steven (2010), Speech Accent Archive, George Mason University. http://accent.
gmu.edu/, retrieved on October 2nd, 2010.
Wells, John C. (1982), Accents of English, vol. I, An Introduction, Cambridge University Press,
Cambridge.
SYLLABLE RESTRUCTURING IN EARLY SOLOMON ISLANDS

PIDGIN ENGLISH: AN OPTIMALITY-THEORETIC ANALYSIS
ANDREI A. AVRAM*
Abstract
The present paper proposes an optimality-theoretic account of the various syllable
restructuring strategies used in early Solomon Islands Pidgin English. These consist of prothesis,
epenthesis, paragoge and consonant deletion. It is shown that vowel harmony plays no part in
determining the quality of the epenthetic and paragogic vowels, contrary to claims put forth in the
literature. Also, the repair strategies conducive to syllable restructuring in early Solomon Islands Pidgin
English are strikingly similar to those attested in other early varieties of Melanesian Pidgin English.
Keywords: Solomon Islands Pidgin English, Optimality Theory, prothesis, epenthesis,
paragoge.
1. Introduction
The present paper is an analysis of the strategies employed by Solomon
Islands Pidgin English to repair illegitimate syllables1. The theoretical framework of
the analysis below is that of Optimality Theory (see e.g. Kager 1999).
Solomon Islands Pidgin English2 is one of the varieties of Melanesian
Pidgin English whose earlier stages are hardly documented. Samples of early
Solomon Islands Pidgin English are presented and analyzed by Mhlhusler
(1987), Keesing (1991a and 1991b) and by Tryon and Charpentier (2004). They
provide valuable information about e.g. the syntax and the lexicon of the
language at the end of the 19th century and the beginning of the 20th century, but
provide little insight into the phonology, given the faulty transcriptions or the
normalizing, i.e. anglicizing orthography frequently restoring English-like
forms3, as in e.g. the samples in Keesing (1991a) and Tryon and Charpentier
(2004: 236240). Consequently, except for a small number of early attestations,
*
University of Bucharest, Department of English, andrei2.avram@gmail.com.

See also Avram (2007).
2
For a general presentation of the language see Jourdan (1988 and 2007). Some
characteristics of the Anglicized variety of Solomon Islands Pidgin English are discussed in
Jourdan (1989).
3
See Hancock (1977) and Avram (2000) for a discussion of this problem.
1
SYLLABLE RESTRUCTURING IN EARLY SOLOMON ISLANDS PIDGIN ENGLISH: AN

OPTIMALITY-THEORETIC ANALYSIS
15
the data examined in this article consist of forms used by older speakers or of
fossilized forms, taken from Keesing (1988 and 1991b), Jourdan and Keesing
(1997), Jourdan (2002), Lee (2008). It is assumed that such forms reflect,
faithfully enough, the phonology of the so-called bush Pidgin, which is close
to that of earlier Solomon Islands Pidgin English. Theoretically and
methodologically, this is in accordance with what Rickford (1986: 162) calls
feed-back from current usage4. Empirically, as noted by Goulden (1990: 54),
bush varieties of [Solomon Islands] Pidgin English conserve archaic
material and thus provide insights into the history and development of MPE [=
Melanesian Pidgin English].
All examples are rendered in the orthography or the system of
transcription used in the sources mentioned. Intrusive vowels appear in bold
characters. If known, the date of the early attestations is mentioned.
The article is structured as follows. In section 2 I analyze the adjustment
of illicit onset clusters in the etyma. I treat /s/-initial onset clusters separately
given the frequent exceptional behaviour of /s/ in English. Section 3 focusses on
the treatment of complex codas. The reflexes of etyma with simple codas are
discussed in section 4. Finally, section 5 summarizes and discusses the findings.
2. The treatment of onset clusters
2.1. Reflexes of /s/+ oral stop clusters
Jourdan and Keesing (1997: 409) note the use of interconsonantal []
vowels to break up illicit three- or two-consonant onset clusters with /s/
followed by an oral stop. Consider the following examples:
(1)
a.
b.
c.
/skr/
sikarapu (< E scrub) bush
/st/
sitoa (< E store) store
/sk/
1916 sekool (< E school) school
The prohibition against onset clusters made up of /s/ and a voiceless stop can be
captured by means of the high ranked constraint *ONS/sO:
(2)
*ONS/sO: onsets made up of /s/ and an obstruent are disallowed.
See also Avram (2000).
ANDREI A. AVRAM
16
*ONS/sO dominates DEP-IO since these illicit clusters are broken up via
epenthesis. Given that deletion is not an option, MAX-IO also outranks DEP-IO.
Both *ONS/sO and MAX-IO are ranked higher than CONTIG, the constraint
militating against medial epenthesis or medial deletion5. Finally, the constraint
L-ANCHOR, which bans epenthesis (i.e. vowel prothesis) or deletion at the left
edge, also dominates DEP-IO, and CONTIG is prohibited. The ranking of DEP-IO
and CONTIG is irrelevant to the outcome. It follows that for such forms the
constraint hierarchy is:
(3)
*ONS/sO, MAX-IO, L-ANCHOR >> DEP-IO, CONTIG
The evaluation of a relevant form, e.g. sitoa, is given in the tableau below6:
(4)
/st :/
sto.a
to.a
so.a
si.to.a
is.to.a
*ONS/sO
*!
MAX-IO
L-ANCHOR
*!
*!
*!
DEP-IO
*
*
CONTIG
*
*
Two other strategies appear to have been only sporadically used. Thus,
deletion of /s/ is attested only in one form in the corpus (Tryon and Charpentier
2004: 368):
(5)
1920s tarch (< E starch) starch
Similarly, vowel prothesis occurs once in a sample of bush pidgin (Jourdan

2007: 80), which is presumably illustrative of earlier stages of the language:
isteret (< E straight) good
(6)
Interestingly, a fact gone unnoticed by Jourdan and Keesing (1997: 409), one
form in the corpus illustrates both vowel prothesis and vowel epenthesis:
(7)
/st/
isitapu (< E stop) to stay
This constitutes a violation of what Kager (1999: 105) calls minimal

epenthesis7. Not surprisingly, such forms are rare since, as shown by Kager
5
In this paper I do not decompose this constraint as e.g. in Alber and Plag (1999).
In all tableaux dots indicate syllabic boundaries.
7
Kager (1999) uses epenthesis as a cover term for both prothesis and epenthesis.
6

17
(1999: 105), epenthesis always applies minimally, that is, precisely to the
extent that is necessary to improve structural markedness.
The examples under (1), (6) and (7) suggest that [i] or [e] are the default
epenthetic vowels8. However, /s/ + voiceless stop onset clusters are also broken
up by an epenthetic [u], if the vowel after the cluster in the etymon is /u/:
(8)
a.
b.
/sp/
supun (< E spoon) spoon
/sk/
sukulu (< E school) school
Jourdan (2003) writes that speakers will tend to insert epenthetic vowels in
Pijin words in order to avoid these clusters and that the choice of the vowel is
directed by a rule of vowel harmony9. In fact, the phenomenon illustrated by
such forms is that of vowel copying. In optimality-theoretic terms, this can be
straightforwardly accounted for by assuming a constraint COPY, defined as follows:
(9)
COPY: add a copy of the etymological vowel
The ranking of the constraints is:

(10)
*ONS/sO, MAX-IO, L-ANCHOR, COPY >> DEP-IO, CONTIG
The high ranked constraint COPY ensures the elimination of the competing
candidate [sikulu], with the default epenthetic vowel [i], while securing the
selection of [sukulu]:
(11)
/sku:l/
sku.lu
ku.lu
su.lu
su.ku.lu
si.ku.lu
is.ku.lu
*ONS/sO
*!
MAX-IO
L-ANCHOR
COPY
DEP-IO
CONTIG
*!
*!
*
*
*
*
*
*
*
*!
*!
2.2. Reflexes of the /sl/ cluster

Although no relevant items occur in my corpus of early Solomon Islands
Pidgin English, modern forms give us a clue as to what must have been the fate
of the /sl/ cluster in onset position:
8
9
For other relevant evidence see section 2.5.

See also Jourdan (2007: 110).
ANDREI A. AVRAM
18
(12)
/sl/
silip (< E sleep) to sleep
The onset cluster /sl/ does not surface, even though it would not violate the
sonority sequencing generalization. This can be captured by positing the
following constraint:
(13)
*ONS/sL: onsets made up of /s/ and a liquid are disallowed.
The constraint is ranked high and dominates both DEP-IO and CONTIG since the
onset cluster at issue is resolved by epenthesis. The hierarchy of constraints is:
(14)
*ONS/sL, MAX-IO, L-ANCHOR >> DEP-IO, CONTIG
This is confirmed by the evaluation in the following tableau:

(15)
/sli:p/
slip
sip
lip
si.lip
is.lip
*ONS/sL
*!
MAX-IO
L-ANCHOR
*!
*!
DEP-IO
CONTIG
*
*!
*
*
Finally, note that [i] is again the default epenthetic vowel.

Additional evidence can be adduced from reduplication in the verbal
morphology. The basic pattern, consisting in the reduplication of the first mora
of the verbal root10, is illustrated below:
(16)
/silip/ (< E sleep) to sleep [sisilip] to be sleeping
Reduplication is quite productive in the substrate languages of Solomon Islands

Pidgin English (Lynch 1998, Lynch et al. 2002, Jourdan 2008). Given the
lexical meaning of silip to sleep, the reduplicated form must have existed for
quite a long time in the language. It therefore provides evidence that /sl/ onset
clusters are broken up by epenthesis in early Solomon Islands Pidgin English.
2.3. Reflexes of /s/ + glide clusters
Records of early Solomon Islands Pidgin English do not include any
reflexes of /s/ + glide onset clusters. However, the treatment of these onset
10
Contra Jourdan and Selbach (2008: 183), Jourdan (2008: 477).

19
clusters, which do not violate the sonority sequencing generalization, can be

safely inferred from contemporary forms (Miller 1989: 58, Beimers 1995: 50,
Jourdan 2002):
(17)
/sw/
a.
b.
c.
suea (< E swear) to swear

suim (< E swim) to swim
suit11 (< E sweet) delicious
The glide /w/ in the etymon does not surface in the Solomon Islands Pidgin
English reflex, but appears to have undergone vocalization. This can be
accounted for in terms of the constraints *ONS/CG and *GLIDEVOCALIZATION:
(18)
(19)
*ONS/CG: onset clusters made up of consonants and a glide are disallowed

*GLIDE-VOCALIZATION: vocalization of glides is prohibited
Clearly, if *ONS/CG outranks *GLIDE-VOCALIZATION the glide /w/ surfaces as

the vowel [u]. The constraint is also dominated by MAX-IO, L-ANCHOR,
DEP-IO and CONTIG. The constraint hierarchy is therefore:
(20)
*ONS/CG, MAX-IO, L-ANCHOR, DEP-IO, CONTIG >> *GVOCAL
This ranking is supported by the evaluation for suit:

(21)
/swi:t/
swit
sit
wit
su.wit
su.it
*ONS/CG
*!
MAX-IO
L-ANCHOR
*!
*!
DEP-IO
CONTIG
*GVOCAL
*
*!
*
*
Consider again evidence from reduplication in the verbal morphology:

(22)
/suim/ (< E swim) to swim [susuim] to be swimming
Since reduplication is quite productive in the substrate languages of Solomon

Islands Pidgin English (Lynch 1998, Jourdan 2008) and given the lexical
meaning of suim to swim, it can be assumed that the reduplicated form has
existed for quite a long time. This provides additional evidence that /sw/ onset
clusters undergo glide vocalization in early Solomon Islands Pidgin English.
11
But swit (Link Komik n.d.), in presumably Anglicized Solomon Islands Pidgin English.
ANDREI A. AVRAM
20
2.4. Reflexes of stop + glide clusters

Stop + glide onset clusters do not violate the sonority sequencing
generalization and are, in principle, better candidates for retention, even in
earlier stages of English-lexifier pidgins and creoles12. Unfortunately, the
corpus of records of early Solomon Islands Pidgin English at my disposal
contains no forms illustrative of the fate of such onset clusters. However, an
examination of data from the modern variety suggests that the glide /w/ must
have undergone vocalization. Consider the following examples (Miller 1989: 30
and 62, Beimers 1995: 31 and 2006, Jourdan 2007: 151, Link Komik n.d.):
(23)
a.
b.
c.
d.
kuaet (< E quite) quite

kuik taem (< E quick, time) quickly
kwiktaem (< E quick, time) hurry
tuentifala (< E twenty, fellow) twenty
The occasionally different spelling reflects variation between [w] and [u] in the
modern variety. This would suggest that the retention of /w/ characterizes the
more recent, possibly anglicized pronunciation whereas the vocalization of /w/
in stop + glide onset clusters represents a prior stage. If so, this confirms the
generalization expressed in (18) and the correctness of the ranking in (20). The
ranking of constraints is therefore:
(24)
*ONS/CG, MAX-IO, L-ANCHOR, DEP-IO, CONTIG >> *GVOCAL
The following tableau shows the interaction of these constraints:

(25)
/kwk/
kwik
kik
wik
ku.wik
ku.ik
*ONS/CG
*!
MAX-IO
L-ANCHOR
*!
*!
DEP-IO
CONTIG
*GVOCAL
*
*!
*
*
2.5. Reflexes of stop + liquid clusters

The treatment of stop + liquid onset clusters in early Solomon Islands
Pidgin English is better documented. Although this type of onset does not
12
See Avram (2005, chapter 4).

21
violate sonority requirements, it is disallowed, as shown by the examples

below:
(26)
(27)
(28)
(29)
(30)
(31)
/pl/
a.
bilai (< E play) play
b.
piles (< E place) place
c.
pulande (E plenty) lots
/br/
a.
barata (< E brother) brother
b.
birek (< E break) to break
/tr/
a.
tarae (< E try) to try
b.
tarake (< E truck) truck
c.
turu (< E true) true
/dr/
dorop-em (< E drop, him) to drop
kalaem (< E climb) to climb
/kr/
a.
karae (< E cry) to cry
b.
koros-im (< E cross) to cross
c.
The relevant constraint is *ONS/OL:

(32)
*ONS/OL: onset clusters made up of an obstruent and a liquid are disallowed.
As can be seen, onset clusters consisting of a stop and a liquid are

invariably broken up by an epenthetic vowel. First, in 9 of the 13 forms in (26)
through (31) epenthesis involves vowel copying, and not vowel harmony, as
analyzed by Jourdan (2003). Epenthesis with vowel copying obtains from the
following hierarchy of constraints:
(33)
*ONS/OL, MAX-IO, L-ANCHOR, COPY >> DEP-IO, CONTIG
The correctness of this ranking is demonstrated by the following evaluation:

(34)
/tru:/
tru
tu
ru
tu.ru
ti.ru
ut.ru
*ONS/OL
*!
MAX-IO
L-ANCHOR
*!
*!
COPY
DEP-IO
CONTIG
*
*!
*!
*
*
*
*
*
ANDREI A. AVRAM
22
Second, epenthesis with vowel copying is the preferred solution, for the
resolution of the illicit stop + liquid onset clusters, but not the only one, contra
Jourdan (2003), who claims that the vowel is exclusively dictated by vowel
harmony. In three of the forms in (26) and (27) [i] is the epenthetic vowel. In
sections 2.1 and 2.2 I have claimed that [i] is one of the default epenthetic
vowels. This is supported by the existence of the three forms with epenthetic [i].
Compare barata brother with a copy of the vowel between /b/ and a liquid
to bilai play and birek to break [i] in the same phonological environment.
The constraint hierarchy is:
(35)
*ONS/OL, MAX-IO, L-ANCHOR >> DEP-IO, CONTIG
The evaluation in the tableau below confirms this ranking:

(36)
/brek/
brek
bek
rek
bi.rek
ib.rek
*ONS/OL
*!
MAX-IO
L-ANCHOR
DEP-IO
*!
CONTIG
*
*!
*
*
*!
Finally, pulande lots is an instance of trans-categorial assimilation13,

in which a vowel takes on the place of articulation of a neighbouring consonant.
The assimilation at issue is labial attraction: the [LABIAL]14 consonant /p/
enforces the occurrence of [u] as the epenthetic vowel. The relevant constraint
is CLAB-VLAB, defined as follows:
(37)
CLAB-VLAB: insert a [LABIAL] vowel after a [LABIAL] consonant.
The constraint CLAB-VLAB outranks both DEP-IO and CONTIG. The constraint
hierarchy is:
(38)
*ONS/OL, MAX-IO, L-ANCHOR, CLAB-VLAB >> DEP-IO, CONTIG
Here is the evaluation of the candidates for the input /plenti/:

(39)
/plenti/
plan.de
pan.de
lan.de
13
14
*ONS/OL
*!
MAX-IO
L-ANCHOR
CLAB-VLAB
*!
DEP-IO
CONTIG
*
*!
A term proposed by Clements (1993: 109).

Place of articulation features are considered to be unary.

pu.lan.de
pi.lan.de
up.lan.de
23
*
*
*
*!
*!
*
*
2.6. Reflexes of fricative + liquid clusters

My corpus of early Solomon Islands Pidgin English unfortunately includes
only one example illustrating the treatment of fricative + liquid onset clusters:
(40)
/r/
torou-em (E throw, him) to throw
The illicit cluster in the etymon is resolved by epenthesis of a copy of the vowel
after the cluster. Therefore, the constraint hierarchy securing the emergence of
this candidate is the same as in (33):
(41)
*ONS/OL, MAX-IO, L-ANCHOR, COPY >> DEP-IO, CONTIG
The ranking is demonstrated in the evaluation below:

(42)
/r-/
tro.wem
to.wem
ro.wem
to.ro.wem
ti.ro.wem
ot.ro.wem
*ONS/OL
*!
MAX-IO
L-ANCHOR
*!
*!
COPY
DEP-IO
CONTIG
*
*!
*!
*
*
*
*
*
3. The treatment of coda clusters

3.1. Reflexes of nasal + oral stop clusters
The corpus contains only an extremely small number of forms illustrating
the treatment of complex codas. Consider first reflexes of coda clusters
consisting of a nasal and an oral stop:
(43)
a.
b.
c.
/mp/
siambu (< E jump) to jump
/nt/
wande (< E want) to want
/nd/
endi (< E and) and
ANDREI A. AVRAM
24
On the basis of the description of the phonetics and phonology of modern

Solomon Islands Pidgin English (Jourdan 2003, Jourdan 2007: 110, Jourdan and
Selbach 2008: 174, Lee 2008: 57) it can be assumed that sequences made up of
a nasal and an oral stop must have been realized phonetically either as such or
as homorganic prenasalized stops, [mb], [nt], [nd]15. As will be shown below, the
exact phonetic realization of these clusters in the etyma is, however, irrelevant.
If the reflexes of the etyma in the examples under (43) are [mb], [nt] and
[nd] respectively, they cannot occur in coda position since they violate the
constraint *CODA/NS:
(44)
*CODA/NS: coda clusters made up of a nasal and an oral stop are disallowed
All these clusters are reduced by paragoge. This clearly violates both DEP-IO
and R-ANCHOR, which prohibits epenthesis (i.e. paragoge) or deletion at the
right edge. Since neither deletion nor epenthesis is an option it follows that
MAX-IO and CONTIG outrank DEP-IO and R-ANCHOR. For the forms in (43b, c)
displaying the default paragogic vowel [i] or [e] the constraint hierarchy is:
(45)
*CODA/NS, MAX-IO, CONTIG >> DEP-IO, R-ANCHOR
This ranking ensures the emergence of e.g. [endi] as the optimal candidate:
(46)
/nd/
end
en
ed
en.di
e.nid
*CODA/NS
*!
MAX-IO
CONTIG
*!
*!
DEP-IO
R-ANCHOR
*
*!
*
*
In the form siambu the paragogic vowel is [u]. The selection of [u] as the paragogic
vowel is triggered by the /p/ in the etymon, since both are [LABIAL]. This can be
handled in terms of the constraint CLAB-VLAB, formulated in (37). In the constraint
hierarchy CLAB-VLAB necessarily outranks DEP-IO and
R-ANCHOR:
(47)
*CODA/NS, MAX-IO, CONTIG, CLAB-VLAB >> DEP-IO, R-ANCHOR
This tableau below demonstrates the correctness of this ranking:

(48)
/mp/
si.amp
si.am
15
*CODA/NS
*!
MAX-IO
*!
CONTIG
CLAB-VLAB
DEP-IO
R-ANCHOR
*
Prenasalized stops occur in the substrate languages (see e.g. Lynch 1998, Lynch et al.
2002, and Jourdan 2003).

si.ap
si.am.pu
si.am.pi
si.a.mip
*!
25
*!
*
*
*
*!
*
*
*
Consider next how the same etymological coda clusters are adjusted by
speakers who realize phonetically /mp/, /nt/ and /nd/ as prenasalized stops [mb],
[nt], [nd] respectively. These are prohibited from occurring in coda position by
the following constraint:
(49)
*CODA/SPRENASAL: prenasalized stops are disallowed in the coda.
The relevant constraint hierarchies are given in (44) and (45):

(50)
(51)
*CODA/SPRENASAL, MAX-IO, CONTIG >> DEP-IO, R-ANCHOR

*CODA/SPRENASAL, MAX-IO, CONTIG, CLAB-VLAB >> DEP-IO, R-ANCHOR
These rankings also enforce paragoge. This is demonstrated by the evaluation of

relevant candidates in the following two tableaux:
(52)
/nd/
end
e.ndi
(53)
/mp/
si.a.mp
si.a. mpu
si.a. mpi
*CODA/SPRENASAL
*!
*CODA/SPRENASAL
*!
MAX-IO
MAX-IO
CONTIG
CONTIG
DEP-IO
R-ANCHOR
CLAB-VLAB
DEP-IO
R-ANCHOR
*!
*
*
*
*
Note that a consequence of the phonetic realization of /b/, /t/ and /d/ as
prenasalized stops is the avoidance of the closed syllables in the etyma.
Consider next reflexes of the same coda clusters in the so-called bush
varieties. Typically, if the oral stop in the etymon is a coronal one it is deleted,
and the etymological cluster is thereby reduced:
(54)
(55)
/nt/
difiren (< E different) different
/nd/
a.
giraun (< E ground) ground
b.
han (< E hand) hand
Deletion of the coronal stop is enforced by a high ranked constraint:
ANDREI A. AVRAM
26
(56)
*CODA/C+t/d/: clusters made up of a consonant and a [nasal,

permitted in coda position.
CORONAL]
stop are not
Since epenthesis is disallowed DEP-IO and CONTIG are also ranked high.
Further, since paragoge would incur their violation, max-io and r-anchor are
both outranked by *CODA/C+t/d and CONTIG. This yields the following
constraint hierarchy:
(57)
*Coda/C+t/d, Dep-IO, Contig >> Max-IO, R-Anchor
The hierarchy of constraints if verified in the tableau below, for the input /hnd/:
(58)
/hnd/
hand
han
had
han.di
ha.nid
*CODA/C+t/d
*!
DEP-IO
*!
*!
CONTIG
MAX-IO
R-ANCHOR
*
*
*!
*
*
3.2. Reflexes of stop + fricative/affricates clusters

Coda clusters consisting of a nasal stop and a fricative or an affricate are
represented by just one example in my corpus:
(59)
/ns/
Pranis (< E France) French
However, their treatment in early Solomon Islands Pidgin English may be

inferred from descriptions of the modern variety (Jourdan 2003, Jourdan and
Selbach 2008) as well as from forms attested in Miller (1989), Beimers (1995),
Jourdan (2002), Beimers (2006), Jourdan (2007), and in Link Komik (n.d.).
Relevant examples include:
(60)
a.
b.
c.
d.
/ns/
fenis (< E fence) fence
/n/
manis (< E month) month
/n/
branis (< E branch) branch
/n/
oranis (< E orange) orange

27
First, note that in coda position the reflex of the affricates // and // is the
fricative /s/. Second, as can be seen, the reflexes of nasal stop + fricative/
affricate contain the default epenthetic vowel [i]. The relevant high ranked
constraint is:
(61)
*CODA/SF: coda clusters made up of a stop and a fricative or affricate are disallowed.
The constraint *CODA/SF dominates both DEP-IO and CONTIG. Since neither
deletion nor paragoge is resorted to, it follows that MAX-IO and MAX-IO are
also ranked high and that they also outrank DEP-IO and CONTIG. The hierarchy
of constraints for the reflexes of the clusters at issue is:
(62)
*CODA/SF, MAX-IO, R-ANCHOR >> DEP-IO, CONTIG:
Consider e.g. the evaluation for the input /fns/:

(63)
/fns/
fens
fen
fes
fe.nis
fen.si
*CODA/SF
*!
MAX-IO
R-ANCHOR
*!
*!
*!
DEP-IO
*
*
CONTIG
*
*
One example in the corpus illustrates the treatment of oral stop + fricative coda clusters:
(64)
/ks/
1909 bokkis (< E box) box
The same form and other relevant ones are attested in the modern variety
(Miller 1989, Beimers 1995, Jourdan 2002, Beimers 2006, Link Komik n.d.):
(65)
/ks/
a.
b.
c.
d.
e.
akis (< E axe) axe

bokis (< E box) box
fokis (< E fox) fox
nekes (< E next) next
sikis (< E six) six
First, such forms confirm the generalization expressed by the constraint

formulated in (61). Second, such illegitimate coda clusters are resolved through
insertion of the default epenthetic vowels [i] or [e]. Third, the ranking proposed
in (62) accounts for the reflexes of oral stop + fricative coda clusters as well:
(66)
ANDREI A. AVRAM
28
/bks/
boks
bok
bos
bo.kis
bok.si
*CODA/SF
*!
MAX-IO
R-ANCHOR
*!
*!
DEP-IO
CONTIG
*
*
*
*
*
*!
3.3. Reflexes of fricative + coronal stop clusters

This is another type of cluster whose reflexes in early Solomon Islands
Pidgin English are extremely poorly attested in my corpus:
(67)
/st/
a.
b.
c.
d.
tas (< E just) just

tes (< E just) just
das (< E just) just
des (< E just) just
Similar forms occur in the modern variety. Consider e.g. the following
examples (Miller 1989, Beimers 1995, Jourdan 2002, Beimers 2006, Link
Komik n.d.):
(68)
/st/
a.
b.
c.
d.
fas (< E fast) stuck

jes (< E just) just
mas (< E must) must
pos (< E post) post
In the reflexes of etyma with /st/ in the coda the illicit coda cluster is normally16
reduced via deletion of /t/. In optimality-theoretical terms this can be handled
by the constraint *CODA/C+t/d, formulated in (56). The coda cluster /st/ in the
etymon is reduced by deleting /t/, in violation of the constraints MAX-IO and RANCHOR. Therefore, this shows that the constraint *CODA/C+t/d is ranked
higher than MAX-IO and R-ANCHOR. Since the illicit coda cluster at issue is not
reduced by epenthesizing a vowel, it follows that DEP-IO and CONTIG also
outrank MAX-IO and R-ANCHOR. Here is the ranking of constraints:
(69)
*CODA/C+t/d, DEP-IO, CONTIG >> MAX-IO, RIGHT-ANCHOR
The interaction of these constraints is shown in the tableau below:
16
There are just three exceptions in my corpus: pristi (< E priest) priest in bush
Pidgin; isti East and westi West in a song, possibly composed during World War II
(reproduced in Jourdan 2007: 8485).

29
(70)
/st/
dast
das
dat
das.ti
da.sit
*CODA/C+t/d
*!
DEP-IO
*!
*!
CONTIG
MAX-IO
R-ANCHOR
*
*
*!
*
*
4. The treatment of simple codas

The way in which etyma with simple codas are adjusted in early Solomon
Islands Pidgin English is relatively amply documented in my corpus:
(71)
(72)
(73)
(74)
(75)
(76)
(77)
(78)
(79)
(80)
(81)
(82)
(83)
/p/
a.
antafu (< E on top) up there
b.
isitapu (< E stop) to stay
c.
koafu (< E go up) to climb
d.
robu (< E rope) rope
/b/
/m/
a.
finisitaemu (< E finish, time) worker whose indenture has elapsed
b.
kamu (< E come) to come
c.
sut-imu (< E shoot, him) to shoot
d.
talemu (< E tell) to tell
/v/
muvu (< E move) to move
/t/
a.
bata (< E but) but
b.
oraete (< E alright) alright
c.
waetemane- (< E white man) white
d.
wokaboti (< E walk, about) to walk
/d/
loti (< E road) road
/s/
bisinisi (< E business) business
/z/
bisinisi (< E business) business
//
finisitaemu (< E finish, time) worker whose indenture has elapsed
//
siosi (< E church) church
//
kabisi (< E cabbage) leafy greens
/l/
a.
olo (< E all) all
b.
raeholo (E < rifle) rifle
c.
solodia (< E soldier) policeman
d.
sukulu (< E school) school
e.
tale (< E tell) to say, to tell
/r/
ANDREI A. AVRAM
30
got bagere (< E got, bugger) forget about it
/n/
a.
ana (< E and) and
b.
Diapane (< E Japan) Japanese
c.
taone (< E town) town
d.
wanekaeni (< E one kind) indefinite article
e.
waetemane (< E white man) white
/k/
a.
gobeke (< E go back) to return
b.
laeka (< E like) like
c.
seke (< E check) to check
d.
tarake (< E truck) truck
e.
wawaka17 (< E work) to work
/g/
bikibiki (< E pig) pig
//
a.
banga (< E bang) to bang
b.
bilongo (< E belong) of
c.
longo (< E along) in
(84)
(85)
(86)
(87)
As can be seen, the simple codas in the etyma include oral stops /p, b, t, d, k, g/,
fricatives /v, s, z, /, affricates /, /, nasal stops /m, n, /18 and liquids /l, r/.
These simple codas do not surface in the early Solomon Islands Pidgin English
reflexes. This suggests that in reflexes of simple codas the constraint NOCODA
ranks high. A vowel is added to permit syllabification. Note that this appears to
be true including of word-internal codas in the etyma, as illustrated by the forms
waetemane, finisitaemu, bisinisi, solodia, wanekaeni.
Since reflexes of word-final codas are much better represented, in what
follows I will focus on paragoge. In all forms with a paragogic vowel, NOCODA
and MAX-IO dominate DEP-IO and R-ANCHOR. In addition, the examples listed
in (71)(87) show that there is considerable variation in the quality of the
paragogic vowel. A frequently occurring case is that of vowel copying, with the
constraint COPY also outranking DEP-IO and R-ANCHOR:
(88)
NOCODA, MAX-IO, COPY >> DEP-IO, R-ANCHOR
Consider the evaluation for e.g. the input but:

(89)
/bt/
bat
ba
17
NOCODA
*!
MAX-IO
*!
COPY
DEP-IO
R-ANCHOR
*
With partial reduplication.

Paragoge in the forms ana (< E and) and wanekaeni (< E one kind) suggest the absence
of /d/ in the input. Otherwise, as shown in 3.1, the expected reflexes would be either [anda] and
[wanekaendi] or [anda] and [wanekaendi] respectively.
18

ba.ta
ba.ti
*
*
*!
31
*
*
Another attested case is that of labial attraction in which a [LABIAL]

consonant in the etymon triggers the selection of [u] as the paragogic vowel. In
all such forms, the relevant constraint CLAB-VLAB is ranked higher than DEP-IO
and RIGHT-ANCHOR:
(90)
NOCODA, MAX-IO, CLAB-VLAB >> DEP-IO, R-ANCHOR
The following tableau illustrates the interplay of these constraints:

(91)
/rp/
rob
ro
ro.bu
ro.bi
NOCODA
*!
MAX-IO
CLAB-VLAB
DEP-IO
R-ANCHOR
*!
*
*
*
*
*
*!
Note that muvu may reflect either the effect of COPY or of CLAB-VLAB since both
constraints would ensure the selection of [u] as the paragogic vowel:
Finally, and not surprisingly, [i] and [e] are the default paragogic vowels
as well. The relevant hierarchy of constraints is:
(92)
NOCODA, MAX-IO >> DEP-IO, R-ANCHOR
This ranking is illustrated in the evaluation below:

(93)
/rd/
lot
lo
lo.ti
NOCODA
*!
MAX-IO
DEP-IO
R-ANCHOR
*
*
5. Conclusions
The evidence presented in sections 2 through 4 points to a strong
tendency towards a CV syllable structure in early Solomon Islands Pidgin
English. This is not surprising given the influence of the substrate languages19
19
For an outline of the phonology of the substrate languages see Lynch (1998), Lynch et
al. (2002).
ANDREI A. AVRAM
32
(Jourdan and Keesing 1997, Jourdan 2003, Jourdan 2008, Jourdan and Selbach
2008, Lee 2008, Avram 2011).
This article has also analyzed the various strategies used by early
Solomon Islands Pidgin English for the restructuring of illicit syllables in the
etyma. The strategies at issue are mostly epenthesis or paragoge employing the
default vowels [i] or [e], epenthesis or paragoge with vowel copying, epenthesis
or paragoge with labial attraction. Consonant deletion appears to be more
restricted. It has also been shown that vowel harmony20 plays no part in
determining the quality of the epenthetic or of the paragogic vowel.
Identical illicit onsets or codas are occasionally subject to different
adjustment strategies in early Solomon Islands Pidgin English. For instance, I
have shown in section 2.1 that /s/ + oral stop clusters are normally simplified by
means of epenthesis, but occasionally via deletion of /s/ or via both vowel
prothesis and epenthesis. As seen in 3.1, the clusters /nt/ and /nd/ are reduced
either by means of paragoge or through deletion of the oral stop. Similarly, the
/st/ cluster in coda position in the etyma is simplified by deleting the /t/.
Simple codas are not excluded altogether. First, as seen in 3.1, [m] and
[n] may occur in word-internal codas, with speakers who have /b/, /t/ and /d/,
whereas with those speakers who realize them as prenasalized stops, i.e. as [mb],
[nt] and [nd] respectively, nasal codas are excluded even word-internally.
Second, [s] is attested in the reflexes of stop + fricative/affricate coda clusters,
as discussed in 3.2. Third, simple codas occasionally appear in reflexes of
simple codas. As shown in 4, reflexes of simple codas normally display a
paragogic vowel. However, the simple coda may surface as such. Actually,
etyma ending in an identical consonant may have reflexes with or without a
paragogic vowel: compare e.g. birek (< E break) to break, where [k] appears
in coda position, with seke (< E check) to check, with a paragogic vowel after
[k]. This variability has been reported for the earlier stages of other Englishlexifier pidgins and creoles as well (Avram 2005).
Clearly, as shown in sections 2 through 4, identical illicit onsets or codas
are occasionally subject to different adjustment strategies in early Solomon
Islands Pidgin English. This also accords rather well with the variability typical
of early pidgins, abundantly documented in the literature.
Further evidence in support of the analysis of syllable restructuring in
early Solomon Islands Pidgin English comes from early records of Tok Pisin
analyzed by Hall (1943), Murphy (1966)21, Mhlhusler et al. (2003), Tryon
and Charpentier (2004), and of Bislama analyzed by Crowley (1990, 1993 and
1998) and by Tryon and Charpentier (2004). The strategies employed by early
20
See also Avram (2007b, 2008 and 2009).

This is the second edition of the book, reproducing the first one, published in 1943. It
therefore presents the Tok Pisin of the 1940s.
21

33
Solomon Islands Pidgin English for the resolution of illegitimate onsets or

codas present many similarities with those attested in these closely related
varieties of Melanesian Pidgin English. This is not surprising given the
historical-linguistic circumstances in which the three varieties emerged22. The
similarities extend to the different vowels selected for epenthesis or paragoge.
Thus, early Tok Pisin and early Bislama also have two default intrusive vowels.
In the former [i] and [] function as default intrusive vowels (Hall 1943: 16,
Avram 2005: 151), while early Bislama has [i] and [e] (Avram 2005: 90, 153
and 198-199). Both varieties also resort to vowel copying (Avram 2005: 90, 153
and 198-200). Also, in both early Tok Pisin (Avram 2005: 150) and early
Bislama (Avram 2005: 153) the labial intrusive vowel [o] is selected due to
labial attraction.
Similar strategies are also attested in early records of other (unspecified)
varieties of Melanesian Pidgin English. For instance, epenthetic vowels
breaking up onset clusters are found in forms recorded by Churchill (1911).
Even the rare case of violation of the so-called minimal epenthesis via a
combination of both vowel prothesis and vowel epenthesis in reflexes of
/s/-initial onset clusters is attested in early records of other (unspecified)
varieties of Melanesian English: Schuchardt (1883/1980: 22) lists forms in
which both a prothetic vowel and an epenthetic one break up /s/ + oral stop
onset clusters. Note that in these cases the intrusive vowels are either [i] or [e],
as in the early Solomon Islands Pidgin English forms discussed in 2.1 and 2.2.
Finally, /s/ is occasionally deleted in these clusters, as shown by Mhlhusler
(1997: 134).
To conclude, this analysis of syllable restructuring in early Solomon
Islands Pidgin English will hopefully contribute to a more comprehensive
picture of the syllable structure typical of earlier stages of Pacific Englishlexifier pidgins and creoles as well as of the repair strategies used by these
varieties for creating well-formed syllables.
REFERENCES
Alber, Birgit and Ingo Plag (1999), Epenthesis, deletion and the emergence of the optimal
syllable structure in creole, Rutgers Optimality Archive 335. http://ruccs.rutgers.
edu/roa.html.
Avram, Andrei A. (2000), On the phonological interpretation of early written records of English
pidgins and creoles, in Lancaster University Working Papers of the Centre for Language
in Social Life, 117, pp. 1-15.
Avram, Andrei A. (2005), On the Syllable Structure of English Pidgins and Creoles, Editura
Universitii din Bucureti, Bucharest.
22
For a summary see Holm (1989: 526-529). For a detailed account see Tryon and
Charpentier (2004).
34
ANDREI A. AVRAM
Avram, Andrei A. (2007a), Syllable restructuring in early Solomon Islands Pidgin English, in
Bucharest Working Papers in Linguistics, IX, 1, pp. 219-230.
Avram, Andrei A. (2007b), On the alleged occurrence of vowel harmony in Solomon Islands
Pidgin English, in Analele Universitii din Bucureti. Limbi i literaturi strine, LVI,
pp. 87-98.
Avram, Andrei A. (2008), Vowel harmony in Pijin?, talk given at the Research Group in
Phonology, 14 February 2008, Lancaster University, Lancaster.
Avram, Andrei A. (2009), The epenthetic and paragogic vowels of Pijin, in Revue roumaine de
linguistique, LIV, 3-4, pp. 365-382.
Avram, Andrei A. (2011), The epenthetic and paragogic vowels of Pijin: Internal development
or substrate influence?, in P. P. Chruszczewski, Z. Wsik (editori), Languages in
Contact 2010, Philologica Wratislaviensia. Acta et Studia 4, Wydawnictwo Wyszej
Szkoy we Wrocawiu, Wrocaw, pp. 7-23.
Balzer, Trevor, Ernie Lee, Peter Mhlhusler, Paul Monaghan, Denise Angelo, Dana Ober
(2008), Pidgin Phrasebook, 3rd edition, Lonely Planet Publications, Hawthorn.
Beimers, Gerry (1995), Wei fo Raetem Olketa Wod long Pijin, Solomon Islands Christian
Association, Honiara.
Beimers, Gerry (2006), Pijin dictionary. www.pijinplus.net/lexicon/lexindex.htm.
Churchill, William (1911), Beach-la-mar. The Jargon or Trade Speech of the Western Pacific,
The Carnegie Institution, Washington.
Clements, George N. (1993), Lieu darticulation des consonnes et des voyelles: une thorie
unifie, in B. Laks and A. Rialland (eds.) Architecture des reprsentations
phonologiques, CNRS ditions, Paris, pp. 101-145.
Crowley, Terry (1990), Beach-la-Mar to Bislama. The Emergence of a National Language in
Vanuatu, Clarendon Press, Oxford.
Crowley, Terry (1993), Pre Pionnier and late nineteenth-century Bislama, in Journal of Pidgin
and Creole Languages, 8, 2, pp. 207-226.
Crowley, Terry (1998), The Bislama lexicon before the First World War: Written attestations,
in Papers in Pidgin and Creole Linguistics, 5, pp. 61-106.
Goulden, Rick (1990), The Melanesian Content in Tok Pisin, The Australian National University,
Canberra.
Hall, Robert A. (1943), Melanesian Pidgin English. Grammar, Texts, Vocabulary, Ams Press,
New York.
Hancok, Ian F. (1977), Recovering pidgin genesis: Approaches and problems, in A. Valdman
(ed.) Pidgin and Creole Linguistics, Indiana University Press, Bloomington & London,
pp. 277-294.
Holm, John (1989), Pidgins and Ceoles, vol. II, Reference Survey, Cambridge University Press,
Cambridge.
Jourdan, Christine (1988), Langue de personne, langue de tout le monde: le Pijin Honiara, in
tudes Croles, XI, 1, pp. 128-147.
Jourdan, Christine (1989), Nativization and Anglicization of Solomon Islands Pijin, in World
Englishes, 7, 3, pp. 25-37.
Jourdan, Christine (2002), Pijin. A Trilingual Cultural Dictionary [PijinInglisFranis] [Pijin
EnglishFrench] [PijinAnglaisFranais], The Australian National University, Canberra.
Jourdan, Christine (2003), Pijin phonology. http://www.pidgin.ca.
Jourdan, Christine (2007), Parlons Pijin, Histoire sociale et description du pidgin des les
Salomon. LHarmattan, Paris.
Jourdan, Christine (2008), Solomon Islands Pijin: Morphology and syntax, in K. Burridge, B.
Kortmann (eds.), Varieties of English, vol. 3, The Pacific and Australasia, Mouton de
Gruyter, Berlin New York, pp. 467-487.

35
Jourdan, Christine and Roger Keesing (1997) From Fisin to Pijin: Creolization in process in the
Solomon Islands, in Language in Society, 26, 3, pp. 401-420.
Jourdan, Christine and Rachel Selbach (2008), Solomon Islands Pijin: Phonetics and
phonology, in K. Burridge, B. Kortmann (eds.) Varieties of English, vol. 3, The Pacific
and Australasia, Mouton de Gruyter, Berlin New York, pp. 164-187.
Kager, Ren (1999), Optimality Theory, Cambridge University Press, Cambridge.
Keesing, Roger M. (1988), Solomons Pidgin pronouns: A further look, in English World-Wide,
9, 2, pp. 271-292.
Keesing, Roger M. (1991a), The expansion of Melanesian Pidgin: Further early evidence from
the Solomons, in Journal of Pidgin and Creole Languages, 6, 2, pp. 215-229.
Keesing, Roger M. (1991b), Substrates, calquing and grammaticalization in Melanesian Pidgin,
in E. C. Traugott and B. Heine (eds.) Approaches to Grammaticalization, vol. I, Focus on
Theoretical and Methodological Issues, John Benjamins, Amsterdam/ Philadelphia,
315-42.
Lee, Ernie (2008), Solomon Islands Pijin, in Balzer et al., 53-96.
Link Komik (n.d.) Honiara.
Lynch, John (1998), Pacific Languages. An Introduction, University of Hawaii Press, Honolulu.
Lynch, John, Malcolm Ross and Terry Crowley (2002), The Oceanic Languages, Curzon,
Richmond.
Miller, Laura (1989), Peace Corps Solomon Islands Pijin English Dictionary. Peace Corps,
Honiara:
Murphy, John J. (1966), The Book of Pidgin English, 2nd edition, W. R. Smith & Paterson,
Brisbane.
Mhlhusler, Peter (1987), Tracing predicate markers in Pacific Pidgin English, in English
World-Wide, 8, 1, pp. 97-121.
Mhlhusler, Peter (1997), Pidgin and Creole Linguistics, expanded and revised edition,
University of Westminster Press, London.
Mhlhusler, Peter, Thomas E. Dutton, Suzanne Romaine (2003), Tok Pisin Texts. From the
Beginning to the Present, John Benjamins, Amsterdam/Philadelphia.
Rickford, John R. (1986), Short note, in Journal of Pidgin and Creole Languages, 1, 1, pp.
159-163.
Schuchardt, Hugo (1883/1980), Melanesian English, in G. G. Gilbert (ed.) Pidgin and Creole
Languages. Selected Essays by Hugo Schuchardt, Cambridge University Press,
Cambridge, pp. 14-23.
Tryon, Darrell T. and Jean-Michel Charpentier (2004), Pacific Pidgins and Creoles. Origins,
Growth and Development, Mouton de Gruyter, Berlin New York.
36
ANDREI A. AVRAM
AN OBJECTIVE LOOK AT OBJECT CONTROL INSTANCES IN

ROMANIAN SUBJUNCTIVE COMPLEMENTS
MARIA AURELIA COTFAS*
Abstract
The paper sets out to discuss instances of object control in Romanian subjunctive
complements (both s- and ca-subjunctives), starting from previous analyses which argue that
verbs of (object) control (ask, convince, encourage, order, urge, oblige) impose
obligatory control readings on the (necessarily) null embedded subject. We aim to show, starting
from clear empirical evidence as well as the results of a questionnaire, that neither the controlled
pro analysis, nor the one(s) arguing for PRO as the empty embedded subject are on the right
track. The complements of the object control verbs above will be shown to freely display NonObligatory Control behavior: they allow disjoint embedded subjects (overt or null and thus
retrieved by the distinct phi-specification on embedded inflection), control shift and (very
frequently) split control. Such evidence goes hand in hand with similar findings regarding subject
implicative verbs (the try and manage class) and corroborates the more general claim that
Romanian is a subjunctive language and as such does not exhibit syntactic control.
Keywords: obligatory control, optional control, object control verb, infinitive
complement, subjunctive complement, null subject, overt disjoint embedded subject.
1. Introduction
The paper is divided into three main sections, as follows. Section 2
introduces Control Theory and its main tenets as discussed initially for
English infinitival complements: the Obligatory Control (OC) vs. NonObligatory Control (NOC) dichotomy, the characteristics of (OC) PRO and
previous analyses of object control in Romanian (a much less discussed topic as
compared to subject control). Section 3 elaborates on the idea that what have
been dubbed obligatory object control verbs do not impose OC in Romanian
and hence we cannot speak of syntactic control in the subjunctive complements
selected by these verbs. Supporting evidence for this is provided not only from
naturally-occurring examples, but also the results of a questionnaire. A brief
section 4 summarizes the main findings and draws the conclusions.
University of Bucharest, Department of English, maura_cotfas@yahoo.com.
MARIA AURELIA COTFAS
38
2. Briefly on Control Theory

2.1. OC and the properties of (OC) PRO in English infinitival
complements
As an empty category (the subject of infinitive complements in English),
PRO must be licensed, i.e. its postulation must be derived from some general
principle of grammar: The Theta Criterion (which states that the subject role in
any verbs theta grid should always be discharged to some syntactic position)
and the Extended Projection Principle (EPP), requiring that sentences (in
English) have (overt) subjects. Thus, PRO discharges the theta role of the
embedded predicate (i.e. the Agent role of leave, in (1) below). Importantly,
however, it is not case-assigned in virtue of the very fact that it is null (it has no
phonetic content). Thus, if it moves from Spec,VP to the specifier position of
the clause it does so not for case reasons, but merely to satisfy the EPP.
(1)
I forced themi [PROi to leave].
PRO therefore differs from DPs in point of phonological content and case,
which are actually connected since case cannot be assigned to null categories
(but see refinements in more recent literature, according to which PRO can also
bear case). However, the internal properties of an empty category (EC) and
thus of PRO as well represent a subset of the set of properties that
characterize lexical DPs. Consequently, an R-index should be among the
obligatory properties of PRO for reasons of semantic coherence and wellformedness. Besides the R-index, phi-features must also be specified, for cases
when the EC itself functions as an antecedent of anaphors, as in (2):
(2)
a.
b.
John was asked [how PROi to behave oneselfi in public]

Theyi didnt know [how PROi to prepare themselvesi for the event].
Unlike other empty categories (NP-trace, pro, wh-trace), the case of PRO
is somewhat more complicated, given its specification as both [+anaphor] and
[+pronominal] on account of the fact that sometimes it behaves like an
anaphor (3a) and other times like a pronoun (when it receives arbitrary
interpretation) (3b). This points out it should obey both Principle A and B.
(3)
a.
b.
Johni wants [PROi to learn from his mistakes].

PROarb To err is human, PROarb to forgive divine.
AN OBJECTIVE LOOK AT OBJECT CONTROL INSTANCES

IN ROMANIAN SUBJUNCTIVE COMPLEMENTS
39
The fact that GB PRO is both bound and free in its governing category
leads to a contradiction, solved by the claim that PRO is ungoverned, i.e. it does
not have a(ny) governing category (the PRO Theorem).
The ungoverned position of PRO, its specification as both [+anaphor] and
[+pronominal] and the fact that besides its lack of case it has to bear a distinct
theta role from that of its antecedent, hence be visible for theta-marking at LF,
represent complications of the theory of PRO in GB, which were later tackled in
various ways.
In GB Control Theory and the interpretation of PRO were approached
using the basic claims and concepts of Binding Theory (BT). In other words,
control was seen as an extension of BT, such that obligatory control was treated
on a par with anaphoric binding and optional control instances were likened to
pronominal reference relations. More plainly, PRO in OC instances was
analyzed as an anaphor which needed to be bound in its domain governing
category (the main clause), whereas NOC PRO was analyzed as a pronoun free
to pick up a (non-)local antecedent or otherwise bear arbitrary reference.
The GB problems of ungoverned and caseless, but theta-marked PRO
found a solution in Chomsky and Lasniks (1993) proposal that PRO should
bear a special type of case, null case, assigned by non-finite inflection. The
authors propose that PRO, too needs case so as to be visible and interpretable at
LF, on a par with lexical DPs. However, unlike the latter, which bear structural
nominative assigned by finite inflection, PRO will bear null case, a special type
of case that only non-finite inflection can assign and which only PRO is
compatible with. The fact that PRO is incompatible with any other type of case
explains why it is disallowed in regular case-marked position as well as its
complementary distribution with lexical DPs:
(4)
a.
b.
Romarioi tried [PROi to score the winning goal]

*Romario tried [Bebeto to score the winning goal]
(Cornilescu 2003: 227)
Hornstein (1999, 2001) takes issue with some of the earlier accounts on
the interpretation and distribution of PRO. Thus, among others, he maintains
that the distribution of PRO can be explained without recourse to the notion of
government, tackles the problematic different interpretation of PRO in OC and
NOC and draws attention to the theory-internal flavour of null case. His answer
to the above problems is to provide an alternative to the classical theory of
control, proposing to analyse it as a case of DP movement, advocating a
modification in the GB chain theory. In his view, the controller which at PF
surfaces in the matrix clause has moved from the lower infinitival clause where
it was originally projected, stopping to the next theta-position of the main
clause and finally to the case position of the matrix. According to this
40
assumption, the same DP actually bears two theta-roles (the one of the
complement verb and the one of the verb in the matrix). Thus, the chain in (5a)
contains two theta roles of the type (DP Agent, t Agent):
(5)
a.
b.
He tried [ t to come].
He seemed [ t to be crying].
Assuming a movement theory of control, the only difference between

control and raising that still holds is the number of theta roles each construction
has: two in control (5a), but only one in raising (5b). For the rest, they are the
same since both are the result of a moved constituent.
By this token, the null hypothesis that underlies his study is to assume
that OC PRO is identical to NP-trace, i.e. it is simply the residue of
movement (Hornstein 2001: 37). The basic assumptions of his theoretical
framework are the following: (i) theta roles are features on verbs and can trigger
movement; (ii) Greed is enlightened self-interest; (iii) a DP gets theta role by
checking a theta feature of a predicate phrase it merges with; (iv) a chain can
have more than one theta role. Theta roles are therefore treated as
morphological features, a pre-requisite if we are to assume that OC is the result
of movement. When a DP receives a theta role, it means it has checked the
relevant features of the predicate. The radical innovation is the one in (d),
assuming as it does that control forms one chain (similar to raising
constructions), bearing (at least) two theta positions.This, Hornstein (2001: 38)
purports, brings about a radical simplification of the grammar of Control and a
derivation of the basic properties of OC structures.
Finally, Landau (1999 and subseq.) presents a different analysis of
control in English (and crosslinguistically), one that departs from the view that
control is an extension of binding theory. At the same time, Landau (1999)
points out the shortcomings of the movement analysis of control (see above)
and is the first account to finds another more reliable factor able to tease
apart OC from NOC, as well as to operate a distinction within the OC class
itself into Exhaustive and Partial Control. The key ingredient is the tense
specification of the complement, encoded in its C head. As such, Landau (1999)
is the first account which analyzes control as mediated by the (temporal)
specification of the C head of control complements a view that we have also
adopted in our account of the Romanian data, showing that OC readings are
directly linked to the lack of tense specification of the complement. (see Cotfas
(2012) on subject control in Romanian)
Starting from the common observation that not all infinitival
complements are alike as far as control is concerned (Landau 1999: 13),
Landau departs from the classical view that interprets (6a) below (Landaus (5))
as OC, versus (6b, c) as instances of NOC on the basis of the (im)possibility of
a for-complement:

(6)
a.
b.
c.
41
John tried (*for Mary) to win the game.

John wanted (for Mary) to win the game.
John wondered how to win the game.
Contrary to the classical view, (6b, c) display OC properties, such as

impossibility of Long Distance control or strict reading under ellipsis (see
below). The conclusion Landau draws is that the examples in (6 b, c) belong to
the OC category as well, being however different from the OC construction in
(6a). The intuition is that the former are more flexible than the latter while
sharing nonetheless most of the characteristics of obligatory control.
To account for this irregularity within the OC class, he proposes the
distinction between Exhaustive Control (EC) and Partial Control (PC),
differentiated in terms of both the type of identity relation obtaining between
PRO and its controller as well as the temporal properties of the infinitival
complement. Of the two, only EC still abides by the classical view that the
relation between PRO and its controller must be one of strict identity. PC
introduces a new (empirical) problem in the discussion of Control Theory,
pointing out that there are many instances (particularly in English, the language
Landau (1999) focuses on) where the controller is merely (referentially)
included in the set that PRO denotes, in a subset-superset relation.
The author breaks the domain of infinitival complements into seven
classes according to the semantic characteristics of the matrix verb. Although
languages differ as to what sort of verbs are included in each class and even
though some of the verbs may also take finite clauses (or gerundial non-finites,
in English), the classification is (claimed to be) universal. These predicates are:
(i) aspectual (begin, continue, start); (ii) modal (need, have to, be able, etc.);
(iii) implicative (manage, dare, etc.); (iv) desiderative (want, prefer); (v) factive
(regret, hate); (vi) prepositional (claim, maintain, believe); (vii) interrogative
(ask, wonder).
The question is which of these verbs count as EC verbs and which fall
under the PC class. If as stated above EC requires strict identity between
PRO and the controller and PC requires that PRO merely contain the controller,
there must be some test to verify this assumption. This is the (im)possibility to
control the PRO subject of a collective predicate. In other words, whenever a
collective predicate is allowed in the infinitive complement, PC obtains (since
PRO need merely contain the antecedent, not be identical with it); on the other
hand, when such a predicate is disallowed in the complement clause, the matrix
verb that bans it must belong to the EC category.
(7)
(8)
a.
b.
a.
b.
*The chair managed [PROi+ to gather at 6].

* Mary knew that Johni began [PROi to work together on the project].
The chairi preferred [PROi+ to gather at 6].
Maryi thought that Johnj didnt know [where PROi+j to go together].
42
Generalizing to a higher level, Landau observes that besides the

distinction between EC and PC in terms of the semantic type of the matrix
predicate, the two are also syntactically distinct function of the presence vs.
absence of tense in the complement clause. This is actually the crucial factor
distinguishing the two classes (1999: 70). The author acknowledges the
existence of conflicting temporal modifiers between the matrix and the
infinitive clause, suggesting the presence of distinct tense operators. He then
tests all the seven classes of verbs that take infinitival complements, concluding
that modal and aspectual verbs take untensed subordinates, cf. (9):
(9)
a.
b.
*Yesterday, John began to solve the problem tomorrow.

*Yesterday, John had to solve the problem tomorrow.
Thus, EC predicates are untensed, as opposed to PC predicates, and this

distinction correlates with the properties of PRO in the two types of (infinitival)
complements: while in PC configurations PRO is a group name (that is, it is
semantically plural) and it need merely include the reference of the (singular)
antecedent, EC constructions impose a relation of strict identity between PRO
and its antecedent (i.e. strict agreement in phi-features)
2.2. A word on OC vs. NOC
In GB, control configurations were divided in two types: obligatory and
non-obligatory (i.e. optional) control. The former designates constructions that
become illicit in the absence of an overt controller, which alongside the
infinitive complement must be a co-argument of the matrix predicate. Which
argument of the main verb is designated as the controller is partly a lexical
property of the former. Consider (10) below:
(10)
a.
b.
c.
I forced them [PRO to leave] /*I forced [PRO to leave]

I promised him [PRO not to perjure myself]
I tried [PRO to give up smoking]
In (10a) above, force is a verb of obligatory direct object control in that it is its
direct object that controls the empty PRO subject of its infinitival complement.
Promise and try in (10b, c) are verbs of obligatory subject control because they
force their subjects to control the empty subject of the infinitival complements
that they select. Verbs of obligatory control always select PRO-TO
complements and disallow FOR-TO infinitive clauses as complements

43
In OC configurations the controller DP and the infinitive clause which

contains PRO are always co-arguments of the same matrix verb. Conversely, in
NOC the infinitive need not be controlled by a clause-mate DP. This can either
be absent altogether (cases of arbitrary control (11)) or may be in a clause
higher than the one containing the infinitive (long-distance control (12)):
(11)
(12)
a.
[PROarb to err] is human, [PROarb to forgive] divine.
b.
[PROarb to vote for Vadim] would be a tremendous mistake.
Maryi knew that it would damage Johnj [PROi/j to perjure himselfj/herselfi]
There is significant empirical evidence distinguishing the two

constructions:
(i) Long distance control is only possible with NOC, not with OC.
As already stated, in LD cases the controller of PRO and the infinitival
dependent need not be clause-mates. It usually describes the situation when a
local antecedent is available in the structure, but it is not the controller of PRO,
which is instead co-indexed and controlled by a more remote DP available in
the structure. Actually, neither is there the need for c-command, i.e. the
controller can be hierarchically lower than PRO (cf. (17b)):
(13)
a.
b.
Johni said that Maryj thought that [PROi to shave himselfi] would bother Suek
[PROi storming out of the room like that] convinced everyone that Johni is
immature.
(ii)
Arbitrary Control is only possible with NOC, not with OC.

Since OC requires a local overt controller, PRO cannot be interpreted as
having arbitrary generic reference in such syntactic contexts (i.e. be interpreted
as one). In NOC, arbitrary readings of PRO are perfectly acceptable, as we
have seen above.
(14)
a.
b.
*John tried [PROarb to be quiet]

*Mark remembered [PROarb not to smoke in the classroom]
(ii)
Strict Reading of PRO is only possible with NOC, not with OC.
Strict vs. sloppy readings are visible in contexts with gapped material,
which are ambiguous as to the interpretation of the gap. Thus, the reconstructed
constituents tend to be paired with overt NPs in different ways. Consider (15) below:
(15)
Bob persuaded me to pick up the sandwiches and Tom the liquor.

a. .and Bob persuaded Tomi [PROi to pick up.]
b. .and Tom persuaded mej [PROj to pick up.]
44
The interpretation in (15a), in which Tom is the one who picks up the
liquor, illustrates the sloppy identity (identity established with the closest
possible antecedent), while the interpretation in b), according to which I/me is
the one who picks up the liquor illustrates the strict reading (identity
established with the remote antecedent).
As far as the OC/NOC opposition is concerned, only the latter allows for
strict readings (i.e. identity of PRO with a remote antecedent), whereas the former
requires strict identity between the controller and PRO, that is sloppy reading.
(16)
a.
b.
Only Billi expects that it will make a strong impression on Maryj [PROi to
read the play].
Johni tried to leave early, and Billj did too (OC, only sloppy identity possible,
i.e. Billj tried [PROj/*i to leave early]).
Therefore, there is clear empirical evidence so as to support the classical

control dichotomy into obligatory and non-obligatory control structures, the key
differentiating element being the necessity or otherwise lack thereof of the controller
and the infinitive complement to be arguments of the same matrix predicate.
As for more recent developments, Landau (1999) shows that PRO in
NOC is a logophor, i.e. a semantic anaphor (whereas PRO in OC is a syntactic
anaphor) and that the distribution of NOC PRO is different from that of regular
pronouns depending on the semantic prominence of the antecedent DP1. The
author maintains a unitary nature of PRO, in the sense that PRO is always an
anaphor, syntactically identified in OC situations, semantically in NOC. Also, a
more comprehensive picture of NOC contexts is offered: besides the known
cases of Long-Distance Control (controller and infinitive are not clause-mates)
and Arbitrary Control (there is no argumental controller for PROarb), other
instances are discussed: Split Control (two matrix arguments jointly control a
(syntactically) plural PRO); Implicit Control (controller is not syntactically
expressed); Control Shift (semantic effects, either the Agent or the Goal can be
the controller of PRO)
Split Control is an interesting case, because it closely resembles PC.
However, the two should be teased apart: whereas PC describes a situation
where only one controller is overt, which, together with the infinitive
complement are co-arguments of the matrix PC predicate and which, when
singular, can license a (semantically) plural (i.e. collective PRO) (17a), Split
Control is an instance of NOC whereby two overt controllers which occupy
distinct syntactic positions in the matrix jointly control into the infinitival
dependent. (17b). Needless to say that these DPs need not be clause-mates with
the infinitive (17c) and that in such cases a syntactically plural PRO is entailed,
1
We do not insist on such matters here, since they are not relevant to the present
discussion. For details, see Landau (1999).

45
such that syntactically plural anaphors are licensed in these contexts (see the
italicised anaphor in (17c)):
(17)
a.
b.
c.
Johni told Maryj [hei didnt know [which club PROi+j to join together](PC)
Johni agreed with Billj [PROi+j to kiss Mary].
Maryi thought [that Johnj said [that [PROi+j helping each other is vital]]].
2.3. OC in Romanian (and Balkan) subjunctives selected by object

control verbs
Let it be noted that most of the studies dealing with control data in the
Balkan languages (Romanian included) have mostly focused on subject control
i.e. on verbs whose semantics imposes the subject argument as the controller
of the empty embedded subject. This is also the case of English: a look at the
seven classes of verbs taken by Landau (1999) as the domain of control are
all subject control verbs.
This may be due to the fact that at least for Romanian the subjunctive
complements of object control verbs like a ruga ask, a ordona order, a
convinge convince, etc. can easily select ca-subjunctives (i.e, complements
introduced by the specific subjunctive complementizer whenever embedded
material is dislocated to the left periphery of the complement clause), unlike
(some) subject control verbs. Actually, the (im)possibility of a subject control
verb to take a ca-subjunctive has been taken as evidence for the presence or
absence of OC, such that Controlled-subjunctives (in Landaus terminology)
disallow the complementizer whereas Free-subjunctives freely accept it2.
Since object control verbs of the type mentioned above freely select casubjunctives, their status as CP projections is straightforward, so that they
should be analyzed as phasal domains (with saturated T and available
nominative position). By the same token, their being more than TPs clearly
discredits the availability of a raising account of such constructions such as
has been proposed, for example, for subject control verbs by Alboiu (2007).
As far as the control abilities of such verbs are concerned, there are two
routes that authors take with respect to the nature of the (null) embedded subject
of their subjunctive complements. This is either taken to be a (controlled) pro or
a PRO (a case of finite control). Alboiu (2006) takes the first option, in
claiming that object control verbs in Romanian select CP subjunctive
2
For a detailed description of the typology of finite control as discussed by Landau
(with a focus on subject control), we refer the reader to Landau (2004) and Landau (2013). For a
detailed discussion of the advantages and disadvantages of adopting the bi-partite classification of
controlled subjunctives in Romanian (again with a focus on subject control), we refer the reader
to Cotfas (2012).
46
complements with an available Nominative
controlled/anaphoric pronominal/pro (18)3:
(18)
position,
filled
by
am
rugat pe Ioni ca mine
s plimbe proi/eli cinele.
Lihim.acc have.1sg asked pe Ion that tomorrow sbj walk-3sg pro/he dog-the
I asked Ioni PROi to walk the dog tomorrow./that hei should walk.
(Alboiu 2006: 35-36)
Landau (2004: 826-827) analyzes verbs such as urge and ask on a par
with other subject control verbs like try and manage and claims that these
predicates select Controlled-subjunctives and, consequently, that their empty
subjects must be PROs. He quotes Comorovsky (1985) as well as Dobrovie
Sorin (2001) to support his claims:
(19)
(20)
L1- am ndemnat ca de mine

Pro1/*2 s mearg la coal cu bicicleta.
him-I.have urged
that from tomorrow Pro
prt go.3sg to school with the.bike
I urged him to ride his bike to school from tomorrow on
(Romanian: Comorovsky 1985, (3b), in Landau 2004: 827)
I1 krkova Pro1/*2 t recitoj nj poezi.
him asked.1sg Pro
prt recite.3sg a poem
I asked him to recite a poem.
(Albanian: Dobrovie-Sorin 2001, (4c), in Landau 2004: 827)
The notation used in the examples above (i.e. Pro) is Landaus: the author
chooses to notate the empty embedded subject as neither PRO, nor pro, but as a
combination of the two before actually proceeding to demonstrate that the
examples undoubtedly display OC properties and hence that the null subject of
the embedded subjunctives must be a formative of the PRO type. Regardless of
the notation adopted, the indices clearly show what the authors claim, namely the
fact that the object control verbs in question induce OC. In Landaus own words:
these constructions display all the typical properties of obligatory control. The embedded
subject must be null, lexical subjects excluded in this position; it must be coreferential
with a c-commanding matrix antecedent; control by a distant antecedent, or a discourse
referent, is impossible. Furthermore, Pro only permits a sloppy reading under ellipsis []
and supports a de se but not a de re interpretation just like PRO does. Thus, these are
3
The examples under (i), (ii) and (iii) support similar claims (see the subscripts):
(i) Mamai
a
obligat -ok
pe Anak [s mnnce prok toat supa]
mother-the have.3sg obliged her-acc pe Ana s eat-3sg pro all soup
Mother obliged Ana to eat up the soup.
(ii) proi Lk
-am
convins pe Mateik [s plece
elk primulk.]
pro him-acc have.1sg convinced pe Matei s leave-3sg he first-the
I convinced Matei to leave (himself) the first (himself).
(iii)
proi Ik
-am
ordonat servitoareik [s tearg eak masa, nu valetulm.]
pro her-dat have.1sg ordered maid-dat
s wipe-3sg she table-the not butler-the
I ordered the maid the maid to wipe the table herself, not the butler.

47
definitely OC constructions, with the peculiar property that the controlled subject occurs
in a subjunctive clause, rather than an infinitive or gerund. (Landau 2004: 827)
For Balkan languages, the advantages of the (controlled) pro analysis

come (traditionally) from government and case: [+Agr] governs the null empty
pro, but not PRO. Also, this empty subject seems to bear its own case, visible
on secondary predicates present in the complement. While these were tenable
arguments for pro (and against PRO) in GB, the distinction becomes less clear
according to more recent theories, which do away with the notion of
government (see above) and claim that PRO can bear standard case as well.
In the next section, we will discuss data which supports an analysis of the
empty subject of verbs of object control as (free/referential) pro, while
distancing ourselves from both of the previous trends. That is, while we bring
clear evidence that the empty subject of the subjunctive complements selected
by object control verbs does not exhibit the typical characteristics of PRO,
contra Landau (2004) and the authors quoted therein, we also depart from an
analysis that argues for a controlled pro. More precisely, we argue against OC
altogether in such constructions in Romanian and show that they actually
display NOC properties.
3. Object Control in Romanian revisited
3.1. The data
In this section we are going to consider the complements of some verbs
analysed in the literature as object control predicates, i.e. as triggering control
of the embedded subject position by the direct or indirect matrix object. The
verbs taken into account are a ruga/a cere ask/beg, a convinge convince, a
ncuraja encourage/urge, a ordona order (directive), alongside a obliga
oblige, which is classified as implicative.
As is the case with subject implicatives (verbs in the class of try and
manage), we will show that these object control verbs are not obligatory
control verbs. Compared to subject implicatives, these are less controversial or
problematic in that their complements have been analyzed as non-anaphoric CP
projections, hence phases (cf. Alboiu 2006: 35-36)
Our aim here is to show that the CP status of the complements of these
verbs allows for more than just a controlled nominative embedded position, i.e.
that this nominative position can also be occupied by elements other than
bound/co-referent pronominals or a controlled version of pro. That is, evidence
will be given that the embedded nominative position is available not only to
lexical DP subjects but also to disjoint null subjects retrieved as such by the
different phi-specification on the subjunctive verb. Moreover, just as with
subject implicative predicates, split control is an option that these dependents
48
allow quite frequently, further strengthening our claim that an obligatory

control analysis is not on the right track.
Semantically, the control reading mostly occurs with agentive controllers
who are interpreted as directly responsible for securing the truth of the
complement proposition. This, of course, only happens as long as the embedded
subject is null, so as to allow but not impose a control interpretation.
However, given the evidence below, a null embedded subject is not locally
bound by default. (hence the impossibility for it to be a PRO).
In the examples below, all taken from various literary and non-literary
sources found online, we have bolded and italicized the selecting verbs. Please
note the subscripts included, as well as the bolded notations in the glosses,
which emphasize the different markings on the matrix object and the embedded
(null) subject:
(21)
a.
b.
c.
d.
Bock ii a
ordonat lui Opreai ca avioanelem s zboare aa
Boc him-dat-3sg have.3sg ordered to Oprea that planes-the s fly-3pl like
cum merg tancurile in Irak
how go-3pl tanks-the in Irak
Boc ordered Oprea that planes (should) fly the same way as tanks roll in
Irak.
Mamak biatului s- a
rugat de profesorii s nu afle
tatalm-the
mother-the boy-gen have.3sg asked of professors s not find out-3sg fatherlui ce
s-a
ntmplat.
his what have.3sg happened
The boys mother begged the teachers that his father not be told about what
happened.
prsit
unitatea i miia
Colonelul Kemenicik a
colonel-the Kemenici have.3sg abandoned unit-the and me-dat has.3sg
ordonat ca nimenim s nu intre
n birou nainte de
a se ntoarce el
ordered that no one s not enter-3sg in office before comp come back he
Col. Kemenici left the unit and ordered me that nobody should enter the
office until he came back.
[Ai plecat la All Star Game cu gndul de-a participa la competiia de Slam
Dunk?]
[Did you go to All Star Game intent on taking part in the Slam Dunk
competition?]
La nceput nu, dar colegiik miau
ncurajat s fie
i un
at beginning not but colleagues me-Acc have.3pl ncurajat s be-3sg and one
romnm la concurs
Romanian at contest
Not at first, but my colleagues encouraged me that there should be a
Romanian, too in the competition.
In all the four examples above the complement features an overt disjoint lexical
subject. In (21c), this is the negative pronoun nimeni nobody, licensed by
embedded (local) negation.
Likewise, disjointedness can be signaled by the presence of a floating
quantifier whose features retrieve a lower subject different from the matrix

49
object (22a) or merely by the phi-features of the subjunctive. In (22c & d) the
embedded verb appears in a passive-reflexive construction which identifies a
null disjoint subject:
(22)
a.
b.
c.
d.
Colonelulk ii a
ordonat (locotenentuluii) s participe
colonel-the him-dat.3sg has.3sg ordered lieutenant-dat.3sg s participate-3pl
toim n raidul de a doua zi.
all in raid of second day
The colonel ordered the lieutenant that they should all take part in the next
day raid.
[mi-am desprit de elj pentru c au aflat pariniik mei] i mi[I broke up with him because my parents found out] and me-acc.1sg
au
obligat prok s nei+j desprim
proi+j
have.3pl obliged
s us.refl break up-1pl
I broke up with him because my parents found out and obliged me that we
should break up.
a
ordonat s nu se execute
foc prom
generalulk Milea miigeneral Milea me-dat have.3sg ordered s not se execute-3sg fire
asupra demonstranilor
onto demonstrators
Gen. Milea ordered me that no gun shots should be fired onto the
demonstrators.
D.S.S.k leia
ordonat s nu se deschid focul prom n
D.S.S. them-dat have.3sg ordered s not se open-3sg fire
in
interiorul cldirii
interior-the building-gen
D.S.S. ordered them that fire should not be opened on the premises.
Another important piece of evidence is that with a convinge convince

for example, Contol Shift is definitely a possibility, as attested by the following
example. Were it not for the secondary predicates in italics in the complement
(i ea she, too or singur alone-fem) given the homonymy of 3rd person
singular and plural subjunctive forms in Romanian the example could mean
that Mary convinced her parents to go. These predicates show that it is not the
direct object that controls into the complement, but the more distant subject:
(23)
Mariai ik a
convins pe prinik s se duc (singuri /i eai) la
Maria them.acc have.3sg convinced pe parents s go-3sg (alone-fem/and she) at
concert.
concert
Maria convinced her parents that she should go (alone/too) to the concert. not
Maria convinced her parents to go to the concert.
Finally, the following set of examples illustrates the extent to which these
verbs allow split control (see above for a brief account of this type of NOC),
such that the reference of the null embedded subject (syntactically marked for
50
plural, as shown by the plural phi-features on the subjunctive verb) includes that
of both the matrix subject and (direct or indirect) object.
(24)
a.
b.
c.
d.
e.
abia iiam
convins prok [s nu plecam prok+i
hardly them-acc-3pl have.1sg convinced
s not leave-1pl
n ziua
cnd avem
cazare
in day-the when have-1pl accommodation
I had a hard time convincing them we should not leave on the very day of
our check-in.
mai bine ar fi
s ii
convingi prok [s traversai prok+i Bulgaria
better
would be s them-3pl convince
s cross-2pl
Bulgaria
noaptea]
night
It would be better if yousg convince them that youpl should cross Bulgaria at
night.
Spre
sear, gazdak mea mi a
convins [s mergem
towards evening host my me-acc-1sg have.3sg convinced s go-1pl
prok+i la un vecin]
at a neighbour
Around dusk, my host convinced me that we should pay a visit to one of the
neighbours.
li am
convins prok [s facem
prok+i bebe]
him-acc-3sg have.1sg convinced
s make-1pl
baby
I convinced him that we should have a baby.
Vrciuk a
convins -oi
[s fac sex
prok+i n zece
Vrciu have.3sg convinced her-acc-3sg s make sex-3pl
in ten
minute!]
minutes
Vrciu convinced her in ten minutes that they should have sex!
Examples (24a-e) have a convinge convince in the matrix, with italicized

collective predicates, when available. In (24e), for example, the embedded VP
is a face sex have sex, which, lack of the PP cu X with X can be taken as a
(syntactically) plural predicate.
The next examples illustrate split control with a cere ask/beg (25a) and
a obliga oblige (25 b-e) in the matrix:
(25)
a.
b.
c.
[G]eneralul Nuk mii a

cerut [s organizm prok+i mpreun
general Nu me-dat-1sg have.3sg asked s organize-1pl
together
primirea unui camion de la Timioara.]
reception a-gen truck from at Timioara
General Nu asked me that we organize together the reception of a truck
from Timioara.
li am
obligat prok [s mergem prok+i mpreun acolo]
him-acc-3sg have.1sg obliged
s go-1pl
together there
I obliged him that we go there together.
cnd a
aflat
prok [c -s
nsrcinati proi] [mi when have.3sg found out
that be-1sg pregnant
me-acc-1sg
a
obligat prok [s ne
cstorim
prok+i]

d.
e.
51
have.3sg obliged
s us.refl get married-1pl
When he found out that I was pregnant, he obliged me that we get married.
oblig
prok [s ieim
prok+i pe cmp, n
i mi
and me-acc-1sg oblige-3sg
s get out-1pl
on field in
ntmpinarea lor]
welcoming-the theirs
and he obliges/obliged me that we go out into the fields to welcome them.
Eak mi
oblig
s nu m
lenevesc
i s mergem prok+i
she me-acc-1sg obliges-3sg s not me-acc get lazy-1sg and s go-1pl
dimineaa i seara
la plimbare
morning-the and evening-the at walk
She obliges me not to laze about and that we should go for walks in the
mornings and evenings.
A ncuraja encourage and a ordona order are no different in this

respect (see (26) below):
(26)
a.
b.
c.
d.
l iam
ncurajat prok s gteasc proi, s fac
curat proi
him-acc-3sg have.1sg encouraged s cook-3sg
s make-3sg clean
cu mine, s vorbim prok+i la telefon
with me
s talk-1pl
at phone
I encouraged him that he should cook, tidy up the house with me, that we
speak on the phone.
la nceputul
lui 2007 am
ncurajat prok un grup de tinerii
at beginning-the of 2007 have.1sg encouraged
a group of youngsters
s facem
prok+i o echip de jurnaliti cretini
s make-1pl
a team of journalists christian
At the beginning of 2007 I have encouraged a group of youngesters that we
should set up a team of Christian journalists.
Tot faniik miau
ncurajat s -i
facem
proi+m i
still fans-the me-acc-1sg have.3pl encouraged s it-dat make-1pl
and
un videoclip acestui nou single
a video
this-dat new single
It was again the fans who encouraged me that we should also have a video
done for this new single.
a
ordonat prok s ateptmi+m noi ordine de la elk
Miime-dat-1sg have.3sg ordered
s wait-1pl
new orders from him
He ordered me that we should await new orders from him.
In conclusion, the object control verbs considered directive/exercitive

verbs of command/order as well as the (more) implicative oblige have been
shown not to trigger obligatory control effects, being thus better analyzed as
NOC predicates. This is an important conclusion, since it goes against most
claims in the literature on both Romanian and the Balkan languages more
generally, see Dobrovie-Sorin (1994) and Kapetangiani (2010).
52
3.2. Further evidence: a questionnaire
3.2.1. Method
The aim of our questionnaire was to test whether disjoint lexical subjects
are accepted by speakers in the complements of subject implicatives and object
control verbs. In the present paper, we will focus on the data and results for
verbs of object control. Given the evidence above, our expectation was that this
should be easily accepted. Also, as far as object control verbs are concerned, we
expected that a obliga oblige, given its semantics (implicative), should yield
somewhat different results from its exercitive object control sisters. This is in
accord with more recent accounts in Landau (2012), who distinguishes between
desiderative object control verbs and implicative control verbs and associates
them with different syntactic behaviour (see the Discussion)
Our questionnaire consists of 24 sentences whose grammaticality
speakers were asked to decide on. The sentences featured five implicative verbs
of subject control (a ncerca try, a cuta (s) try/endeavour, a reui
manage, a izbuti succeed/manage, a risca risk) and five verbs of object
control (four directive: a ruga/cere ask/beg, a ncuraja encourage, a
convinge convince, a ordona order and the implicative a obliga
oblige/coerce). Next to these which were our target , two more matrix
verbs were included as distractors: the modal a putea can and the
implicative/psych verb a-i aminti remember. As previously stated, all the
sentences involving these matrix verbs featured subjunctive complements with
overt disjoint DP subjects. Therefore, a number of 12 verbs were tested, of
which we mainly focused on the five subject implicatives and the five object
control predicates just mentioned. The fact that our questionnaire actually
contains 24 sentences instead of 12 is easily accounted for: each of the 12 verbs
tested appeared twice, i.e. for each of them there were two distinct sentences on
whose correctness respondents had to make judgements. The two sentences
with the same selecting predicate differed in that one was a s-complement,
whereas the other a ca-complement (i.e. had embedded material displaced to the
complements LP sometimes the disjoint subject DP itself, other times
adverbial expressions). Notice that under our approach, s-subjunctives are no
different from ca-subjunctives as far as their CP status is concerned. As stated
above, we take the subjunctive complements in the control environments under
analysis as CP projections (either CFins with covert ca or CforcePs)4. Moreover,
the two sentences featuring the same type of matrix verb were not given one
4
Thus, s-subjunctives are CFinP projections, whereas ca-subjunctives would be CForceP
projections. Nonetheless, CFinPs (i.e. complements without lexicalization of ca) can also be phasal
(i.e. temporally independent, Nom. case position, etc).

53
after the other, but were jumbled with the others, for a better homogeneity of
the test. Next to keeping them apart, we tried to make them as different in
content as possible, so different matrix subjects/objects as well as different
embedded DPs and predicates were used each time.
Our expectation was that the presence or absence of ca should not in
principle influence the interpretation of the given sentence5 an expectation
borne out by the results. Another prediction was that the bulk of respondents
should allow for the possibility of disjoint subjects in the complements of these
verbs. More precisely, this was an expectation we put to the test.
To sum up, our questionnaire contains 24 sentences featuring (twice) all
the 12 above-mentioned predicates as matrix verbs selecting either s- or casubjunctives with overt disjoint subjects. Its aim is to show how felicitous such
disjoint subjects are in such contexts as far as our informants are concerned.
The respondents were 40 people of various ages roughly between 22 to
60 and professional backgrounds (students, colleagues, mainly people
activating in various fields). They were asked to pass grammaticality
judgements on the 24 statements by choosing either yes or no as an answer
to the requirement State whether you find the following examples
grammatical/interpretable or not. They were also given the opportunity to
make any observations they saw fit in a special rubric provided after each
sentence in part. Most significantly, next to choosing Yes or No, they were
asked to provide grammaticality scores from 1 to 5 for each sentence, such that
a score of 5 meant perfectly grammatical/correct, whereas a score of 1 stood
for ungrammatical/incorrect/unacceptable. Naturally, the choice of a Yes
triggered a score of 5, whereas a No answer could be associated with a score
on a 1-4 scale.
In the discussion of the results below, we provide the percentages for the
No answers as well, depending on the scores received. Importantly, as we
shall see, most of the No answers revolve around a score of 3 or 4, which
shows that even though the respondents did not consider a certain example
perfectly acceptable, they did not dismiss it as altogether ungrammatical. Below
are a few samples of the sentences in our questionnaire, involving predicates
like a ordona order and a convinge convince. The disjoint embedded DP
subjects appear in boldface, as are the relevant disjoint phi-features in the
corresponding glosses:
Namley, if one chooses to treat s-subjunctives as different from ca-subjunctives in that

the latter would allow disjointedness readings more readily than the former, we would expect to
have different interpretations for the same matrix verb with the two types of complement, which
has not been the case. That is, the bulk of respondents equally accepted as correct (or less than
correct/incorrect, for that matter) sentences with disjoint embedded subjects within both s- and
ca-subjunctives.
54
(27)
a.
b.
c.
d.
Boc i- a
ordonat
lui Oprea
ca avioanele armatei
Boc him.dat have.3sg ordered-3sg to Oprea-3sg that planes-the army.gen
s survoleze zona. din dou n dou ore
s fly over-3pl area-the from two in two hours
Boc ordered Oprea that the armys planes should fly over the area every two
hours.
Ia
ordonat secretarei
s nu intre
nimeni n
her.dat have.3sg ordered secretary.dat-3sgi s not enter-3sgj nobody in
birou pn nu se termin edina.
office till not finish
meeting-the
He ordered the secretary that no one was to enter the office before the
meeting was over.
Cu greu
am
reuit
s l
convingem pe Mihai
with difficulty have.1pl managed-1pl s him.acc convince-1pl pe Mihai-3sgi
ca masa festiv s aib
loc la Marriott
that dinner festive s have-3sgj place at Marriott.
We had a hard time trying to convince Mihai that the farewell dinner should
take place at Marriotts
ntr-un sfrit lam
convins pe colonel
s plece
i
in one end him.acc have.1pl convins pe colonel-3sg s leave-3pl and
soldaii
n permisie peste weekend
soldiers-the in leave
over weekend
We finally convinced the colonel that the soldiers, too should have a
weekend leave.
2.2.2. Results
Below we present the results of our questionnaire for the five object
control verbs. The Yes columns provide the percentages corresponding to
those answers which confirmed the possibility of control suspension, whereas
the No ones give the numbers corresponding to those answers which
disallowed this option to various degrees. As far as the total number of No
answers for each of these verbs is concerned, we also provide the percentages
relative to the 1 to 4 scores, to see to what extent the judgement of
ungrammaticality translates into totally unacceptable (i.e. a score of 1) or else
marginally acceptable or even quite acceptable (i.e. scores ranging between
3 and 4).
Let us then look at the results for the five object control verbs we have
chosen to test. With some of these verbs the results are lower than the ones
observed for subject control implicatives of the try and manage type. This
might be due to their semantics, i.e. to the fact that most of them are
directive/exercitive verbs (as well as the force type implicative) that signal
that a certain action (asking/ordering/convincing/encouraging, etc.) is
performed by an Agent (the matrix subject) onto a(n animate) Patient/Theme
(the matrix object), who should as such be the logical subject (agent) of the
event denoted by the embedded subjunctive. They are thus less permissive with

55
respect to the possibility of allowing disjointedness in their complement

clauses, especially if this is signaled by an overt disjoint DP in situ (see more in
the Discussion below). Nonetheless, the majority of ungrammaticality
judgements for object control verbs ranges around a score of 3 or 4.
Tables 1 and 2 below give the results for a ruga ask/(beg), first the
grammatical vs. the ungrammatical answers and then the percentages
corresponding to the scores of ungrammaticality. As seen in Table 2, the 7
choices of no (out of a total of 80) have been given scores of either 4 (4
answers) or 3 (3 answers).
YES (score: 5) NO (score 1-4)
73 (91.25%)
7 (8.75%)
Table 1: Grammaticality judgements for a ruga ask/beg
Score 4 4 (57%)
Score 3 3 (43%)
Score 2 0
Score 1 0
Table 2: Percentages of ungrammaticality scores for a ruga ask
Next, in Tables 3 and 4 we illustrate the results for another object

control verb, a ordona order. Like in the case of the previous one, contexts
with this matrix predicate received a lot of grammaticality judgements (almost
89%), and scores of 3 and 4 for the remaining ungrammaticality ones:
71 (88.75%)
9 (11.25%)
Table 3: Grammaticality judgements for a ordona order/command
Score 4 7 (77.7%)
Score 3 2 (22.2%)
Score 2 0
Score 1 0
Table 4: Percentages of ungrammaticality scores for a ordona order/command
The results for a convinge convince are given below, with slightly
decreasing scores:
57 (71.25%)
23 (28.75%)
Table 5: Grammaticality judgements for a convinge convince
56
Score 4 15 (65.21%)
Score 3 7 (30.43%)
Score 2 0
Score 1 1 (4.5%)
Table 6: Percentages of ungrammaticality scores for a convinge convince
Finally, the next two sets of charts (7 & 8 for a ncuraja encourage and
9, 10 for a obliga oblige/make) show that with these last two verbs
respondents have been more reserved in allowing control suspension. However,
in spite of the fact that the number of ungrammaticality judgements is higher,
the bulk of these scores still revolve around 3 or 4.
50 (62.5)
30 (37.5%)
Table 7: Grammaticality judgements for a ncuraja encourage
Score 4 15 (50.21%)
Score 3 12 (40%)
Score 2 3 (10%)
Score 1 0
Table 8: Percentages of ungrammaticality scores for a ncuraja encourage

39 (48.75)
41 (51.25%)
Table 9: Grammaticality judgements for a obliga oblige
Score 4 19 (47.5%)
Score 3 16 (40%)
Score 2 5 (12%)
Score 1 0
Table 10: Percentages of ungrammaticality scores for a obliga oblige:
To sum up the data regarding these object control predicates on the

whole, let us have a look at the last table, Table 11, which gives the general
percentage across all the five verbs considered. Out of a total of 400 answers,
290 confirmed the possibility of control suspension, whereas the remaining 110
disallowed this possibility while not dismissing it as downright
ungrammatical (see the percentages for the no choice).
290 (72.5%)
110 (27.5%)
Table 11: Grammaticality judgements for object control predicates (a ruga ask, a
ordona order/command, a convinge convince, a ncuraja encourage/urge, a obliga
oblige)

57
3.2.3. Discussion
The most important fact to be emphasized is that the percentages above
confirm our initial expectations, as well as the empirical data illustrated in the
previous sub-section, proving that the majority of speakers (72.5% for object
control verbs) allow disjoint subjects in their complements. Let us analyze the
results in more detail, concentrating on key issues, i.e. s-subjunctives vs. casubjunctives, (ii) the position and type of embedded subject, and (iii) type of
matrix verb.
It is interesting to notice that there were no significant differences in the
interpretation of these sentences depending on the introductory element (s vs.
ca). That is, generally speaking, speakers who accepted the possibility of
control suspension in the subjunctive complements of the ten verbs under
analysis did so regardless of whether these dependents were s- or casubjunctives. Similar observations apply to those cases where the respondent
dismissed such a possibility: the choice of no was roughly consistent across
the two types of complements.
There were cases, however, when some respondents mentioned in the
observation rubric corresponding to a s-subjunctive (with a disjont subject in
situ) that though the sentence may sound interpretable and correct as it was
a word order according to which the (disjoint) subject appears leftmost,
necessarily hosted by ca would make it sound even better.
Most likely, this has to do with parsing reasons: while reading through
the sentence from the matrix down, the parser first sees the matrix subject, the
verb and the (in)direct object, followed by the embedded subjunctive. At this
point, a point where s/he hasnt yet bumped into the lexical disjoint subject,
particularly if the phi-features of the subjunctive happen to match those of the
matrix tense the parser has most probably already identified the matrix object
as the agent of the action denoted by the subjunctive. Thus, the presence of a
disjoint subject might be perceived as disruptive, since it forces one to
overwrite an already formed (or at least preferred) interpretation.
Conversely, once the parser sees the complementizer, followed either by
the left-dislocated disjoint subject itself or by topicalized embedded material,
this might be taken as a (more) reliable cue that the complement domain is truly
self-standing. What is more, when the embedded DP subject appears
topicalized, it is actually the first embedded element available for interpretation.
In such cases therefore, the presence of an overt ca alongside the topicalized
embedded subject immediately following it allows the parser to understand that
somebody else is performing the embedded action even before knowing what
that action actually is.
58
Bearing in mind the above observations concerning the availability of ca

and s, the type of subject used and its position within the complement also
seems to have influenced speakers judgements. Thus, some of the lower
percentages in the tables above might find an explanation if we focus on this
particular criterion. Let us remember that all the 24 examples featured overt
disjoint subjects in the complement, some more agentive than others. That is,
we have chosen to test the possibility of control suspension the hard way, i.e.
by exposing respondents to the least user-friendly scenario.
Disjointedness can also be signaled by mere mismatch in phi-features
across the two verbs, therefore in the absence of overt disjoint subjects.
Actually, a quick look at the empirical data presented above will show that there
are more examples with a null pro embedded subject than with overt ones, and
these are quite permissive in allowing split control. Had we included such types
of complement clauses among the sentences in our questionnaire, the results
might have been stronger in some cases.
Obviously, certain configurations are preferred to others, depending on
both the type of subject (overt vs. null) and its position in the clause (post- or
pre-verbal with lexicalization of ca in the latter case). Structures with null
(disjoint) subjects (retrieved via subjunctive inflection) do not seem to pose
problems, generally. As for those with overt subjects, there seems to be a
preference for the post-verbal/topicalized position of the subject (with overt ca)
most likely for parsing reasons, as discussed.
As for object control verbs, a similar gradability obtains. Thus a ruga
ask seems to be the less restrictive of them, allowing for control suspension in
over 90% of the cases. It is closely followed by a ordona order/command,
with 88.75%. The next three, in decreasing order, are: a convinge convince
(71. 25%), a ncuraja encourage (62.5%) and finally a obliga oblige
(48.75%). In spite of the difference in results, these verbs share clear NOC
properties (see above).
It must then be their sematics that can account for this difference, as
documented in Landau (2012). Discussing the duality of (Obligatory) Control
(i.e. mainly the EC/PC cut in English and infinitive languages or (to a much
lesser extent) the OC/NOC one in subjunctive languages), the author argues
that EC complements are semantically extentional and untensed ([T]), whereas
PC complements are intensional (only they can introduce possible worlds with
diverging temporal coordinates) are hence tensed ([+T]). He further correlates
this distinction with the (in)visibility of implicit arguments as controllers, such
that EC verbs need an explicit controller (because EC control is syntactic
predication6), whereas PC verbs allow implicit controllers.
6
The condition on syntactic predication is that the argument predicated of must be
syntactically represented. Hence, because EC control is syntactic predication, it fails with an
implicit controller (Landau 2012: 14). Actually, it is precisely this unavailability of implicit

59
We do not expand on these here, mainly for space reasons and because
further investigation in required. The relevance of this analysis to the discussion
at hand about our object control predicates is the following: discussing object
control crosslinguistically7, Landau (2012) points out a cut within this class into
what he calls a) desiderative object control verbs (persuade/ask) and b)
implicative object control verbs (force-type predicates). Importantly, the
difference lies, again, in their ability to omit the object controller, such that a)
desideratives (order, command, entrust) freely allow object drop whereas b)
implicatives (compel, force, influence) resist it.
Before we (briefly) look at the behaviour of our verbs with respect to
object drop, let us notice that the proposed cut nicely reflects the gradability of
our five object control predicates above. From the proposed framework, we
understand that desiderative object control verbs pattern with/are PC verbs,
whereas implicative object control verbs pattern with EC ones. Hence, verbs
like ask or order/command are less restrictive on their complements than verbs
like force/compel or influence.
This is exactly what our results reflect: a ruga/a cere ask and a ordona
order/command (alongside a convinge convince/persuade, whith slightly
lower scores), as irrealis desiderative object control verbs not entailing their
complements, have shown themselves quite permissive as far as control
suspension is concerned, unlike a ncuraja encourage (to be likened, perhaps,
to Landaus influence) and a obliga oblige (= force/compel), which, as
implicative object control verbs have revealed themselves more restrictive with
respect to this possibility.
Of course, since all the five verbs display NOC behaviour in Romanian,
we are not arguing here that they evince different control constructions as
they may well do in control languages such as English (alongside Polish,
Hebrew, a.s.o.).
Interestingly, this cut seems to be coherent with respect to the
(im)possibility of object drop. Bearing in mind the above gradability from most/
more to least/less permissive (a ruga ask > a ordona order/command > a
convinge persuade > a ncuraja encourage/urge > a obliga oblige/force/
compel), the examples below confirm that while object drop is acceptable with
the first two, it is definitely less so with the other ones. Given that the data as
well as the results have confirmed the possibility of control suspension in these
complements, we represent the embedded subject (when empty) as ei/k:
(28)
a.
Am
rugat ei [s mi
aduc
ei/k o carte] / [s nu mai stea
have.1sg asked
s me.dat bring-3sg/pl
a book / s not more stay-3sg
lumeak aici]
control with EC verbs which defines the EC/PC dichotomy. For more details, see forthcoming
work of the author on this issue.
7
The examples are from Hebrew and Polish, languages that offer fertile testing ground for
object drop because they do not abhor object drop as mush as English (Landau 2012: 11)
60
b.
c.
d.
e.
folk
here
I have asked [that s/he/they should bring me a book]/[that people should not
cram the stairs]
Am
ordonat ei [s nu (se) mai trag
(ei/k)]
have.1sg ordered
s not
more shoot-3sg/pl
I have ordered [that s/he/they should cease fire]
ei/k]
??Am
convins ei [s nu mai vin
have.1sg convinced
s not more come-3sg/pl
*I have convinced [that s/he/they should not come anymore]
??/*Au
ncurajat ei/k [s plece
lumeak mai devreme]
have.3pl encouraged
s leave-3sg folk more early
*They have encouraged [that everybody should leave earlier]
*Am
obligat [s plece
cu toii (abia mine)]
have.1sg obliged s leave-3pl with all (barely tomorrow)
*I have obliged [that they should all leave only tomorrow]
To conclude, while both desiderative/irrealis object control verbs like

ask or order and implicative object control verbs like oblige/compel (if
we adopt Landaus (2012) terminology) allow control suspension in Romanian,
they seem to do so to different degrees. The stronger restrictiveness of the latter
is a side effect of their (more restrictive) semantics and it nicely correlates with
the exhaustive type of control relation that such predicates impose on their
dependents in control languages. This further supports the view that control is
ultimately a semantic phenomenon in Romanian, depending on favouring
conditions and, most importantly, on the semantics of the selecting predicate.
4. Conclusions
In our paper, we have brought conclusive evidence against the claim that
Romanian exhibits obligatory control with object control verbs or with any
type of control verb, for that matter. The conclusions we reach in this paper are
valid for control predicates throughout that is, for subject implicative verbs as
well. Consequently, Romanian is not a control language and it would seem that this
is valid for other languages of the Balkan Sprachbund as well, Greek in particular.
Evidence in favour of control obviation, as well as for semantic tense (the
possibility of disjoint (future-oriented/irrealis) temporal adverbials in the
complements of both subject implicatives and object control verbs has
prompted us to depart from the classical binary classification of (Balkan)
subjunctives in control environments and argue for a tripartite one in
Independent, Restricted and Anaphoric subjunctives, with the afore-mentioned
predicates selecting the second type of subjunctive complements.
As such, the data discussed above in section 3 have been crucial not only
in order to prove previous theories on object control wrong, but also and more
importantly, perhaps in order to re-asses the previous binary classification of
Balkan subjunctives in control configurations.

61
REFERENCES
Alboiu, Gabriela (2007), Moving Forward with Romanian Backward Control and Raising, in
W. Davies and S. Dubinsky (eds.), New Horizons in the Analysis of Control and Raising,
Springer, Dordrecht, pp.187-211.
Alboiu, Gabriela (2006), Are we in agreement?, in C. Boeckx (ed.), Agreement Systems, John
Benjamins, Amsterdam/Philadelphia, pp.13-39.
Chomsky, Noam and Howard Lasnik (1993), The theory of Principles and Parameters, in J.
Jacobs, A. von Stechow, W. Sternefeld, Theo Vennemann (eds.), Syntax: An International
Handbook of Contemporary Research, De Gruyter,Berlin, pp. 506-569.
Cornilescu, Alexandra. (2003), Complementation in English: A Minimalist Approach, Editura
Universitii din Bucharest.
Cotfas, Maria Aurelia (2012), On the Syntax of the Romanian Subjunctive: Control and
Obviation, PhD dissertation, University of Bucharest.
Dobrovie Sorin, Carmen (2001), Head to head Merge in Balkan subjunctives and locality, in M.
L. Rivero and A. Ralli (eds.), Comparative Studies of Balkan Languages, Oxford
University Press, Oxford, pp. 44-74.
Kapetangiani, Konstantia (2010), The Minimalist Syntax of Control in Greek, PhD dissertation,
University of Michigan.
Hornstein, Norbert (1999), Movement and Control, Linguistic Inquiry, 30, pp. 69-96.
Hornstein, Norbert (2001), Move! A Minimalist Theory of Construal, Blackwell Publishing,
Oxford.
Landau, Idan (1999), Elements of Control, PhD dissertation, MIT.
Landau, Idan (2004), The scale of finiteness and the calculus of control, Natural Language and
Linguistic Theory, 22, pp. 811 877
Landau, Idan (2012), Towards a dual theory of obligatory control, handout at Annual Conference
of the English Department 14, June 2012, University of Bucharest.
Landau, Idan (2013), Control in Generative Grammar: A Research Companion, Cambridge
Spyropoulos, Vassilios (2007), Finiteness and control in Greek, in W. Davies and S. Dubinsky
(eds.), New Horizons in the Analysis of Control and Raising, Springer, Dordrecht, pp.
159-183
LE VERBE FAIRE PLURI-FONCTIONNEL ET SES QUIVALENTS

EN HINDI ET EN ROUMAIN
SABINA POPRLAN*
THE PLURI-FUNCTIONAL FRENCH VERB FAIRE AND ITS EQUIVALENTS IN

HINDI AND ROMANIAN
Abstract
The French verb faire to do may function as a lexical verb, as a support for a noun and
as an auxiliary verb. As a support verb it appears in verbo-nominal constructions; as an
auxiliary, it is used in co-verbal structures imposing to the main verb a certain type of agentivity:
causative (directive), factitive or eventive/ instrumentative; to these values, we have added the
quasi-perlocutionary role. The agentive prototype may also be used as a full, lexical verb when
it functions as an anaphora (lexical or submissive) or as an autonomous, non-anaphoric lexeme.
The comparison of this verb to its Hindi and Romanian heteronyms (karn to do, bann to
make) and a face has shown that in both languages the verbs function as a support for a noun,
but only the Romanian a face may be used in co-verbal structures. In Hindi, the different values
are expressed by a basic verb (the eventive) or a derived one (for the factitive and the causative
meaning). Regarding their full verb status, karn/ bann and a face can be used as nonanaphoric items, but only karn and a face function also as lexical or submissive anaphoras.
Keywords: lexical and auxiliary verb, lexical and submissive anaphora, agentive values,
pragmatic interpretation, non-anaphoric use.
1. Introduction
Nous proposons, dans cette tude, une analyse des divers emplois du
verbe faire de mme que de ses htronymes hindi (karn et bann) et
roumain (a face).
Le verbe faire dont la signification est celle daccomplissement,
dexcution, reprsente le factitif explicite1 prototypique et, en tant que tel, il
permet, en vue de son analyse, une premire classification de ses emplois.
*
Universit de Bucarest, Facult des Langues et Littratures trangres, Section de

Langue Hindi, sabina_ioana@yahoo.com.
1
Les autres verbes daction, hyponymes de faire, pourront tre considrs, en change,
comme des factitifs implicites (cf. Theban et Theban 2005: 58). Quand le terme factitif est
64
SABINA POPRLAN
On mentionnera ainsi quil peut fonctionner comme verbe plein , aussi

bien quen tant que verbe support . Comme verbe daide ou auxiliaire, il peut
tre le support dun nom, ou dun adjectif, constituant avec ces parties du
discours des expressions verbo-nominales, verbo-adjectivales, ou le support
dun autre verbe ( linfinitif en franais) dans des structures coverbales.
Lemploi en tant que verbe plein de faire correspond, en premier lieu,
un usage simple (non-marqu, non-substitutif) recouvrant diverses valeurs
smantiques du verbe oprationnel, et, en second lieu, la fonction
anaphorique, marque.
Avant de passer lanalyse proprement dite de ce verbe, nous pensons
apporter une motivation pour la description ci-dessous.
Pour arriver aux emplois substitutif et semi-substitutif ou soumissif de
faire, nous avons d, en premier lieu, isoler son fonctionnement global en tant
que verbe plein (i.e. non co-verbal, non support ), qui concerne les
emplois mentionns aussi bien que les structures non-anaphoriques. ct du
verbe faire en tant que verbe plein , nous avons dabord distingu son emploi
de verbe support dans des expressions verbo-nominales (apeles ainsi de
manire gnrique) et co-verbales (quand il reprsente un auxiliaire).
Nous nous sommes galement arrte sur le premier type structural o
faire est verbe support vu que, par exemple, certaines expressions verbonominales comme faire signe/ ir karn en hindi ou, toujours en cette langue,
ravn karn, litt. envoy faire ont un sens causatif, important distinguer surtout
dans le cas des anaphores soumissives , o il sous-tend le segment antcdent.
Le second type structural, fond sur faire co-verbal, intresse en premier
lieu pour les structures dagentivit initiale et mdiane2, dont les profils
analyss dans ce chapitre nous ont aide aussi dans lanalyse de lanaphore
mronymique soumissive 3, vu que ces deux types de factitivit sous-tendent les
segments de cette anaphore: < fairei fairem fairem >, < dire de faire (et) faire>.
employ comme adjectif il pourra tre remplac aussi par factuel . Une seconde signification
de ce terme est celle dagentif (au sens gnral).
2
Ce type agentif occupe une position centrale dans la sphre de la factitivit, vu que la
plupart des actions humaines sont des accomplissements dune requte formule par un Causateur
distinct de lExcutant, ou bien des actions pratiques projetes ou inities par lAgent mdian
lui-mme, cas o il est simultanment Causateur (les situations dauto-causation). Dans cette
tude, nous avons suivi la thorie des actants (conus en tant que rles smantaxiques ou de la
smantique propositionnellle) de Theban (1980, 2007). Lauteur dcrit sept actants (lAgent
initial/ le Causateur, lAgent median/ lExcutant, lAgent final: la Force ou lInstrument, le
Patient, le Lieu initial/ la Source, le Lieu median/ le Trajet et le Lieu final/ le Bnficiaire), de
mme que sa description des trois types dagentivit (initiale ou causative, mdiane ou factitive et
finale: ventive/ instrumentative ).
3
Nous avons ddi un chapitre ce type anaphorique, que nous avons identifi partir
des theories tudies et des exemples analyss, dans notre thse de doctorat portant sur lanaphore
verbale. Lanaphore soumissive peut tre dfinie comme tant lexpression de la raction
factuelle, dobissance, la requte formule dans lantcdent; cet accomplissement reprsente,
au niveau lexical, une reprise semi-substitutive de lexpression causative gouvernante: <
causation + factitivit factitivit > (Poprlan 2010: 177). Nous prcisons aussi que faire est

65
Le mme tour co-verbal peut exprimer, en franais et en roumain, aussi le

troisime type dagentivit, finale, qui rend compte, pour ce qui est des
rfrents humains, des actions involontaires de ceux-ci, comme celle de
renverser un objet quelconque par manque dattention, mais il se rfre, en
premier lieu, aux actions des forces de la nature (comme le vent, la pluie, etc.).
Nous avons identifi, en dernier lieu, un fonctionnement quasi-perlocutoire de
lauxiliaire faire, apparent lagentivit initiale ou causative (fonction
illocutoire), dans les structures co-verbales faire croire, faire accroire.
Avant de commencer notre expos, nous prcisons quen hindi, la
diffrence des langues romanes o les verbes agentifs reprsentatifs sont faire et
a face, il y a deux factitifs explicites: karn et bann.
Le premier (karn) est le verbe support typique des structures
nomino-verbales, de mme que substitut segmental et soumissif ; il peut tre
aussi le support dun autre verbe dans la structure bi-verbale de lhabituel
(participe perfectif invariable de V1 + karn); le second, dont le sens spcifique
est celui de confectionner , complte laire smantique du premier; il peut
fonctionner comme verbe autonome et il apparat souvent dans des structures
nomino-verbales; dhabitude il est non-anaphorique4.
2. Faire en tant que verbe support
2.1. Dans des expressions verbo-nominales
Dans les expressions verbo-nominales, le verbe faire reprsente
lingrdient central et stable. Tandis quen roumain et en franais, le mme sens
reprsentatif pour deux des trois types de notre classification des anaphores verbales: les anapores
lexicales ou segmentales (en tant que substitut fidle, infidle ou rsomptif) et les vnementielles
soumissives (en tant que marqueur spcifique daccomplissement dune requte formule
dans lantcdent). Le troisime type est exprim par les anaphores temporelles; les anaphores
vnementielles connaissent aussi le type 1 ou endo-temporel les vnementielles proprementdites (Poprlan 2010: 51).
4
Le verbe karn apparat comme verbe support dans des structures nomino-verbales
comme intazr karn, attente faire ( attendre ), nt karn petit-djeuner faire
( prendre le petit-djeuner ) ou km karn, travail faire ( travailler ). Bann peut
fonctionner comme verbe plein quon emploie quand il sagit dun obiectus effectus, dans des
situations comme celles de construire une maison, ghar bann ou un nonc, vkya bnn;
selon nous, il reprsente plutt un verbe support dans des expressions comme khn bann
( prparer un plat, cuisiner ), citra bann ( faire un dessin, dessiner ) ou yojn bann
( faire un plan, planifier ). part les structures trans-existentielles quon vient dnumrer, ce
factitif correspond aussi aux transitions quationnelles (obiectus affectus), cf. infra, exemple 9,
Ma tumh manuy ke machue bang Je vous ferai pcheurs dhommes ou qualificatives,
cf. infra, exemple 8: yadi var ne mujhe kurp bany hot si Dieu mavait fait laide
(toujours obiectus affectus).
SABINA POPRLAN
66
lexical peut tre rendu par une expression analytique, ou par un autre verbe
autonome (factitif implicite), en hindi, on rencontre le plus souvent la variante
verbe support, ce qui nexclut pourtant pas les solutions monoverbales, plus
rares. Voil quelques exemples qui illustrent les variantes analytiques et les
solutions monoverbales en franais et en roumain: faire la traduction dun texte
/ traduire un texte, faire des achats / acheter; a face traducerea unui text / a
traduce un text, a face cumprturi/ a cumpra. En hindi les variantes nominoverbales, drives partir du verbe karn, sont les plus frquentes: kis ph k
anuvd karn (faire la traduction dun texte), bien quon rencontre, de manire
plus rare, aussi des variantes monoverbales: khardr karn / khardn, faire
des achats / acheter.
(1)
Or se trouvait l un homme qui avait une main paralyse; ils lui posrent une question:
Est-il permis de faire une gurison le jour du sabat? Ctait pour laccuser.
(Matthieu, 12.10)
M w , xM x . Dx w M E x ,
M c M M A ?
Vah ek manuya th, jisk hth skh gay th. s par do lagne ke lie log ne un se
pch, Ky virm ke din cg karne k j hai?
i iat un om avnd mna uscat. i L-au ntrebat, zicnd: Cade-se, oare, a vindeca
smbta? Ca s-l nvinuiasc.
lexpression verbo-nominale du franais faire une gurison quon peut

paraphraser, dans ce contexte biblique, par le factitif implicite gurir,
correspond en roumain le verbe curatif plein a vindeca, plus frquent dans cette
langue, tandis quen hindi la variante bi-morphologique est adjectivo-verbale:
cg5 karne (karn tant linfinitif oblique, forme marque par le suffixe -e
et reclame par la postposition k, au fminin, k, signifiant ici de ), dont le
sens, dans le contexte plus large: cg karne k j hai serait existe-t-il la
permission de faire ou rendre sain / guri [le jour du sabat] .
2.2. Dans des expressions co-verbales
Faire peut tre aussi le support dun verbe. On lappelera alors co-verbe,
verbe auxiliaire ou auxiliant. Du point de vue syntaxique, la structure typique en
franais pour faire coverbal est < faire + lauxili linfinitif >: faire crire.
cette structure correspond parfois en roumain le tour < faire + lauxili au
5
Le mme adjectif, accompagn du second verbe faire, bann, artefactuel, construit une
autre expression cg bann, toujours valeur trans-qualificative, qui signifie globalement
corriger, amliorer .

67
subjonctif >: a face s scrie, mais aussi (et assez souvent) la structure bi-verbale
< a pune + le subjonctif du verbe auxili >, a pune s scrie. En hindi, le sens
correspondant faire crire en tant que structure agentive initiale ou causative
(< fairei >): demander quelquun dcrire est rendu par un lexme
synthtique, intgrant, correspondant la signification directive, dans ce cas,
likhvn: likh-, racine du verbe likhn crire + -v, suffixe spcifiquement
causatif + la terminaison de linfinitif -n attache la nouvelle racine likhv-:
enjoindre qulquun dcrire .
Au cas o le mme syntagme faire crire et a face s scrie correspondrait
la signification dicter et a dicta , on aurait en hindi le premier driv
partir de la racine, ici transitive, likh-, likh- form laide du suffixe - qui
introduit non plus le sens factitif initial ou causatif (cf. lagentivit initiale dans
la thorie smantaxique de Theban), mais le sens factitif proprement dit ou
central, correspondant lagentivit mdiane (< fairem >), qui indique une
action pratique (non-discursive) spcifique, comme celle de dicter un texte
quelquun.
part les significations agentives initiale et mdiane quon vient de
mentionner, faire en tant que verbe auxiliaire peut tre aussi le porteur dun
sens agentif final (< fairef >), comme cest le cas pour une expression du type
faire tomber, si on se rfre, par exemple laction du vent qui, soufflant, fait
tomber quelques feuilles dun arbre.
Tandis quen roumain on peut rencontrer la structure factitive ci-dessus,
sous la forme dj mentionne < faire + lauxili au subjonctif >, a face s
cad, en hindi, comme prcdemment, on naura pas une variante structurale
analogue, mais une squence du type < cause du / par le vent quelques feuilles
de larbre tombrent >, hav se pe ke kuch patte gir gae.
Les expressions coverbales seront, dans ce qui suit, classifies selon un
critre fonctionnel smantaxique, compte tenu du type dagentivit ou factitivit
que faire en tant que verbe agentif ou factitif explicite impose au syntagme dans
lequel il apparat.
Nous allons inventorier et analyser trois types de faire coverbal: faire en
tant quauxiliant agentif initial, mdian et final, auxquels nous ajouterons le
faire coverbal perlocutoire, dans des structures comme faire (ac)croire quelque
chose quelquun. Larticle de Theban et Theban (2005) nous a inspire dans le
choix de cette organisation conceptuelle.
Dans lexemple ci-dessous, lauxiliant faire de lexpression faire
composer impose celle-ci une valeur agentive initiale correspondant lacte
directif par lequel le Causateur (Ai/ C, le matre de musique) a exig de
lExcutant (Am/ E, son lve) daccomplir laction en question.
(2)
Matre danser: Est-ce quelque chose de nouveau?
SABINA POPRLAN
68
Matre de musique: Oui, cest un air pour une srnade que je (C) lui (E) ai fait composer
ici, en attendant que notre homme ft veill. (Le Bourgeois gentilhomme, p. 25)
-vM : M D D c ?
x -vM : , x- M M x M , o M M
E .
Nritya-ikak: Ko na cz hai ky?
Sgt- ikak: H , preyas gt k ek tarz hai jise ma ne (C) yah par taiyr karvy
hai, jab tak ki rmn uh (p. 89)
Profesorul de dans: Ceva nou?
Profesorul de muzic: Da, o arie de serenad. Am pus s-o fac aici, pn s-o scula
jupnul. (pp. 377 - 378)
Vu quen hindi on ne rencontre pas de structure syntaxique quivalente,

le mme sens agentif initial ou causatif est rendu, de manire typique, par un
verbe causatif, ici karvn ( linfinitif marqu par la terminaison -n)
demander de faire driv dune racine en loccurrence transitive kar- du verbe

karn faire , laide de laffixe de causation exclusivement verbale -v. Le
causatif synthtique karvy hai joue ici le rle de support verbal pour ladjectif
taiyr prepar , achev , dans lexpression taiyr karvy hai, litt. prpar
jai demand de faire , de la squence preyas - gt k ek tarj hai jise ma ne
yah par taiyr karvy hai cest lair dune srnade que jai fait faire ici .
En roumain, lexpression faire coverbal du franais est rendue, dans cet
exemple, par la structure toujours coverbale am pus s-o fac dans laquelle
lauxiliaire a pune s est un support causatif de nouveau exclusivement verbal.
Au niveau de la configuration actancielle, lExcutant nest explicite
quen franais, dans la squence je lui ai fait composer, tandis quil est implicite
dans les deux autres langues; la composition mlodique dsigne par les verbes
composer et a face reprsente un vnement trans-existentiel (obiectus effectus).
Le fragment qui suit illustre le rle dauxiliaire agentif median de faire.
Dans celui-ci, lexpression coverbale faire comprendre indique une action
venir, assume par lExcutant thmatis, Don Quichotte, qui veut expliquer
quelque chose son cuyer, le Bnficiaire.
(3)
Je voudrais avoir assez dhaleine pour parler posment, et que la douleur dont je
souffre cette cte brise se calmt un peu, pour te faire comprendre, Panza! dans
quelle erreur tu es. (Don Quichotte, vol. I, p. 141)
xM ,
M
Ac
o M E . M O x
. E M Q Q M
x . (vol. I, p. 98)
Sako, yadi mujh m dam hot to ma tumhr bt k zordr uttar det. Mer ek cho
pasl m bhayyak dard ho rah hai. Vah yadi utn der ke lie mer p cho det to ma
tumhr bhl ko acch tarah samajh det.

69
- Mi-ar plcea s pot rsufla mai uor, ca s-i vorbesc mai n voie, i mcar de mi s-ar ostoi
durerile din coasta asta, s te pot face s pricepi, Panza, ct de greit judeci!
(vol. I, p. 180)
La structure est similaire en roumain, dans la squence s te pot face s

pricepi, tandis quen hindi lon rencontre le factitif spcifique, synthtique ou
driv laide du suffixe - de la racine samajh- comprendre , le sens du
radical nouveau samajh- faire comprendre , correspondant laction
pratique en question. Le factitif est intensifi, en loccurrence, par loprateur
verbal den, dans la structure prdicative au conditionnel samajh det je
texpliquerais , qui oriente laction vers le destinataire.
Le fragment ci-dessous a t choisi pour illustrer la fonction agentive
finale de lauxiliaire faire. La factitivit rudimentaire, non-humaine,
involontaire est typiquement associe aux actions des forces ou des entits de la
nature, lactant correspondant tant lAgent final6 Af (F / I).
La structure agentive finale, co-verbale en franais, et assume par des
agents naturels comme les mites et les vers qui font tout disparatre sur la terre,
est exprime dans les deux autres langues par des prdicats mono-verbaux, dont
les sujets agents sont la rouille , morc, rugina et les mites, ke, molia.
Lhtronyme hindi est bi-prdicatif: jah morc lagt hai, ke khte
ha, o la rouille attaque (litt. sattache), les mites mangent et celui du
roumain, mono-prdicatif: unde molia i rugina le stric.
(4)
Ne vous amassez pas de trsors sur la terre, o les mites et les vers font tout
disparatre. (Matthieu, 6.19)
A M , c , M Q
Prithv par apne lie pj jam nahkaro, jah morc lagt hai, ke khte ha
Nu v adunai comori pe pmnt, unde molia i rugina le stric
Le fragment qui suit illustre lemploi quasi-perlocutoire de lauxiliaire

faire: Orgon, le pre de Mariane, forme le projet dimposer celle-ci dpouser
Tartuffe. Dorine, la suivante de Mariane, qui entend cette intention, refuse de
sy fier. Dans la rplique dOrgon, la structure faire croire est paraphrasable par
(Je sais bien le moyen) de vous faire adhrer ce que je viens dannoncer/ de
vous en persuader .
6
Lagent final est le troisime actant agentif dans la thorie de Theban. Responsable de
lagentivit instrumentative ou ventive, il reprsente lInstrument laide duquel lagent
immdiatement suprieur, lExcutant, fait une certaine action (on crit avec un crayon), ou bien
les Forces de la nature comme le vent, la pluie, dont laction nest pas contrlable par les
humains.
SABINA POPRLAN
70
(5)
Orgon Dorine: Je sais

(Molire, Le Tartuffe, p. 93)
bien
le
moyen
de
vous
le
faire
croire.
A : A M M x x M .
Orgon: Aur ma jnt h ki kaise tumh vivs kary jy. (p. 34)
Orgon: Am leac eu la-ndemn, ca s te fac a crede. (p. 37)
La squence coverbale est rendue de manire similaire en roumain ca s

te fac a crede. En hindi, en labsence dune structure comparable, nous
rencontrons lexpression au passif tumh vivs kary jy, qu toi confiance
soit faite , la structure vivs karn, dont le sens initial est celui de
confiance faire faire quelquun (vu quelle se fonde sur le driv causatif du
verbe karn, faire ) remplit, en loccurrence, un rle quasi-perlocutoire.
3. Faire en tant que verbe plein
Nous avons suivi jusqu present le comportement du verbe faire en tant
que verbe support , premirement dans des expressions verbo-nominales, et
ensuite dans les structures co-verbales, o il est responsable, en tant
quauxiliant, des valeurs factitives ou agentives (initiale ou causative, mdiane
ou pratique et finale: instrumentative ou ventive), auxquelles nous avons
ajout lemploi quasi-perlocutoire (cf. Poprlan, 2010: 110-114).
Dans ce qui suit, nous proposons une analyse du mme lexme en tant
que verbe plein , en usage autonome ou non-anaphorique, tout dabord,
ensuite, en tant que substitut verbal, et finalement en emploi semi-substitutif ou
mronymique soumissif .
Nous commencerons notre analyse par la partie qui revient au verbe faire
en emploi non-substitutif.
3.1. Faire en emploi non-substitutif
En suivant la thorie de la smantaxe (ou smantique
propositionnelle ) de Theban (1980: 26-28), nous affirmons la possibilit de
lidentification dans lUnivers de quatre types de structures statiques (ou
dtats): existentielles, quationnelles, qualificatives, possessives et locatives,
auxquelles lauteur fait correspondre quatre types de transitions (qui se
droulent, de manire ncessaire, entre des tats de dpart et finals concatgoriels). Nous pouvons ainsi assister , et dcrire, des vnements transexistentiels, trans-quationnels, trans-qualificatifs et trans-locatifs/-possessifs.

71
Vu que la dimension dynamique ou transitionnelle de lUnivers est

gouverne par les divers types dagentivit ou factitivit, dont faire est le
marqueur explicite, nous organiserons la section sur faire autonome ou nonsubstitut selon les vnements ou les transitions quil indique.
3.1.1. Dans des structures vnementielles trans-existentielles
Dans un premier exemple, faire trans-existentiel, supposant le passage
dun tat de non-existence un tat existentiel (lvnement ou la transition se
droulant entre des tats de mme type) est rendu en hindi par lexpression
baser karn, habitation faire , o lon rencontre, de manire plutt
a-typique, le verbe karn, faire , dhabitude verbe support ou anaphorique,
la place de bann, construire, confectioner , frquent dans les contextes
artefactuels, constructifs.
(6)
elle [la moutarde] devient un arbre, si bien que les oiseaux du ciel viennent faire leurs
nids dans ses branches. (Matthieu, 13. 32)
x Q o M AM v M A M ExM Q ox M .
ais pe bant hai ki k ke pch kar usk liy m baser karte ha.
se face pom, nct vin psrile cerului i se slluiesc n ramurile lui
Lhtronyme roumain nest pas, en loccurrence, le verbe analogue a

face, mais directement le verbe dsignant laction vise dans la prparation
dune habitation (des nids, dans cette situation), a se sllui, demeurer .
Lexemple intresse aussi pour lquivalence < devenir - bann se
faire, devenir - a se face >, sous-tendue par le mme faire trans-existentiel.
3.1.2. Dans des structures vnementielles trans-quationnelles
la diffrence des vnements trans-existentiels, caractriss, comme
lon vient de voir dans les exemples ci-dessus, par lapparition ou la cration
dun objet/ Patient (lon aurait pu avoir aussi des disparitions; dans les deux cas,
il sagit dvolutions sortales ou ontologiques), les structures transquationnelles concernent les changements contingents (ou de type intra-sortal)
se droulant lintrieur de la mme catgorie ontologique.
Dans la citation ci-dessous, Jsus propose aux frres Pierre et Andr, qui
taient en train de jeter le filet dans la mer, de le suivre; il les fera, en change,
pcheurs dhommes. Lon peut remarquer que tout en restant des tres humains,
les deux frres changeront en fait de statut social, spirituel, devenant par la suite
des aptres. Le faire trans-quationnel du franais est rendu, de manire
attendue, par les htronymes bann en hindi et a face en roumain.
72
SABINA POPRLAN
(7)
Il leur dit: Venez ma suite et je vous ferai pcheurs dhommes.

(Matthieu, 4.19)
Dx E x M , c AA. w M o F .
s ne unse kah, Mere pche cale o. Ma tumh manuy ke machue ban g.
i le-a zis: Venii dup Mine i v voi face pescari de oameni.
3.1.3. Dans des structures vnementielles trans-qualificatives

Les structures trans-qualificatives concernent des changements qualitatifs
lintrieur, de nouveau, de la mme catgorie ontologique (de type intrasortal).
(8)
Le ciel, ce que vous dites, ma faite belle (). Mais, () si le ciel, au lieu de me faire
belle, met faite laide. (Don Quichotte, vol. I, p. 134)
D Ax x M (...) D M o
var ne mujhe asm saundarya k vardn diy hai () yadi var ne mujhe kurp
bany hot (vol. I, p. 92)
Vru cerul s m fac frumoas, aa cum zicei voi, () dac cumva, ntocmai aa cum
m-a fcut frumoas, m-ar fi fcut cerul urt (vol. I, pp. 169 - 170)
Les lexmes roumains sont des quivalents fidles des termes franais; en
hindi, le premier htronyme est une expression verbo-nominale, vardn den,
faire don , dans la squence var ne mujhe asm saundarya k vardn diy
hai, Dieu ma fait (le) don dune beaut infinite ; le second est exprim par le
verbe bann, symptomatique pour la fonction de verbe plein de faire/ a face.
3.1.4. Dans des structures vnementielles trans-locatives/ - possessives
Les vnements trans-locatifs ou trans-possessifs correspondent, en
essence, lide de dplacement dun Patient dun Lieu initial (Source) dans un
Lieu final (Bnficiaire/ But), en parcourant un Lieu mdian (Trajet) plus ou
moins implicite.
(9)
Ils neurent pas fait deux cents pas que leurs oreilles furent frappes par un grand
bruit deau (Don Quichotte, vol. I, p. 181)
Q M E - M A x D .
Ve tho dr gae hge ki unh jal - pratp k vz sun d. (vol. I, p. 134)
dar nu fcuser nici dou sute de pai, cnd le ajunse la ureche un vuiet mare, ca de
ape (vol. I, p. 240)

73
Si en franais et en roumain les structures trans-locatives sont

comparables, la version hindi propose un hyponyme factitif exprim par le
verbe jn se dplacer, aller dtermin par le nom valeur adverbiale dr
distance , le verbe tant au perfectif futur (cf. le futur antrieur du franais):
Ve tho dr gae hge ils auront parcouru une courte distance quand ,
indiquant, modalement, une prsomption.
Un dernier fragment extrait de la Bible illustre le faire trans-possessif:
(10)
Pour toi, quand tu fais laumne, que ta main gauche ignore ce que fait ta main droite.
(Matthieu, 6.3.)
o , o M M
.
Jab tum dn dete ho, to tumhr by hth na jnne pye ki tumhr dy hth ky kar
rah hai.
Tu ns, cnd faci milostenie, s nu tie stnga ta ce face dreapta ta.
La variante hindi est la seule diffrente lexicalement; elle exprime le faire

du franais par lexpression complment interne dn den faire don . Le
transfert de lobjet se passe entre un donateur (la Source) et un rcepteur (le
Bnficiaire) humains, en contraste avec les structures trans-locatives o les
actants Li (Lieu initial ou Source) et Lf (Lieu final ou Bnficiaire)
correspondent dhabitude des rfrents non-anims.
3.2. En emploi substitutif ou anaphorique
Quand il est substitut (anaphore lexicale), le verbe faire peut reprendre ou
bien un segment lexical noyau prdicatif unique plus ou moins complment
dont lextension peut correspondre une proposition (en tant quanaphore
fidle ou infidle) ou bien une squence phrastique (reprise pluri-verbale),
quand il reprsente une anaphore rsomptive, macro-segmentale ou phrastique.
Les htronymes habituels de faire substitutif sont a face en roumain et karn,
en hindi.
Lexemple qui suit illustre la valeur danaphore lexicale infidle de faire.
(11)
Si Dieu habille ainsi lherbe des champs, qui est l aujourdhui et qui demain sera jete
au feu, ne fera-t-il pas bien plus pour vous, gens de peu de foi! (Matthieu, 6.30)
A x ! M x A A M c M . Ex
D Cx M x , ?
Re alpavivsiyo! Khet k ghs j bhar hai aur kal clhe m jhk d jyeg. Yadi use bh
var is prakr sajt hai, to vah tumh ky nah pahanyeg?
SABINA POPRLAN
74
Iar dac iarba cmpului, care astzi este i mine se arunc n cuptor, Dumnezeu astfel
o mbrac, oare nu cu mult mai mult pe voi , puin credincioilor?
Les configurations des paires anaphoriques sont distinctes: tandis quen

franais cest le verbe faire qui remplit, comme attendu, cette tche, en hindi,
les membres verbaux en relation sont: sajn orner, embellir et pahann
habiller , quasi-synonymes, hyponymes de karn (faire): Yadi use bh var
is prakr sajt hai, to vah tumh ky nah pahanyeg? , Si Dieu la ainsi
orne [lherbe des champs (les lis)], alors pourquoi ne vous habillera-t-il [vous
aussi]? .
Le roumain prsente cette fois-ci une anaphore nulle, dont lantcdent
est le verbe a mbrca habiller .
3.3. En emploi semi-substitutif en tant quanaphore vnementielle
mronymique soumissive
Le verbe faire comme semi-substitut verbal correspond laction
soumissive par laquelle le futur Excutant sengage effectuer laction
propose par le Causateur ou procde directement son excution: < demander
de faire laction X sengager faire laction X (exprimer son
engagement)/faire ladite action >. Le type semi-substitutif se dfinit ainsi, en
premier lieu, un niveau smantaxique, vu quil correspond lacte que nous
avons appel soumissif par lequel lAgent mdian (Am / E) annonce
lexcution ou procde directement laccomplissement de laction que lAgent
initial (Ai / C) lui impose par sa requte. La semi-substitution concerne ainsi la
reprise, par la mise en pratique dun projet intimatif, de lingrdient factuel ou
factitif de laction initiale verbale: < dire de faire (et) faire >.
Vu que cette semi-substitution origine pragmatique (lanaphoris
directif) renvoie, en tant quaction soumissive, une partie du syntagme
antcdent, nous y reconnaissons aussi un cas spcifique de mronymie, les
anaphores semi-substitutives sappelant aussi soumissives mronymiques.
Lexemple qui suit illustre ce type de faire:
(12)
Cela fait, il commanda lune de ses dames de lui ceindre lpe, ce quelle fit avec
beaucoup de grce et de retenue En lui ceignant lpe
(Don Quichotte, vol. I, p. 67)
cM M o Ex M v M A v M M M M c
A M . Ex oQ Ax M x M M ExM M M
x
C M

75
Itn kar cukne ke bd usne ek vey ko de diy ki vah vr yoddh k kamar ke cr

aur talvr ko ghum de. Usne ba sth ke sth yah km kiy Usk kamar m
talvr ko ghumte samay (vol. I, p. 26)
Acestea fiind mplinite, dete porunc uneia dintre cele dou femei s-l ncing pe don
Quijote cu spada, ceea ce ea ndeplini cu mult dibcie i nelepciune ncingndu-l
cu spada (vol. I, p. 55).
Faire reprend ainsi en tant que semi-substitut mronymique

soumissif , le segment ceindre lpe de lantcdent il commanda ( lune
de ses dames) de lui ceindre lpe. Si lantcdent reprsente un acte directif
ou de causation, lanaphore correspond lacte soumissif par lequel lExcutant
accomplit laction impose par le Causateur. Dans la squence ce quelle fit, le
pronom dmonstratif ce que, en consonance avec loccurrence fit, reprsente
une anaphore de mme type.
En hindi, la structure coverbale franaise est remplace par une squence
hypotaxique, la proposition subordonne traduisant le composant factuel: de
diy ki vah vr yoddh k kamar ke cr aur talvr ko ghum de il donna
ordre quelle attacht sa taille lpe . Le prdicat anaphorique yah km kiy
il fit cette action comprend, comme en franais, le noyau verbal kiy
complment dun nom dtermin par un adjectif pronominal dmonstratif yah
km cette action partageant le mme rle semi-substitutif.
En roumain, la situation est similaire celle du hindi pour ce qui est de
lantcdent: dete porunc s-l ncing pe don Quijote cu spada, tandis que le
segment anaphorisant est plus proche de la squence franaise ceea ce ea
ndeplini. On remarque, en outre, dans la structure de ce prdicat mronymique,
le fait que, aux htronymes parfaits karn du hindi et faire du franais
correspond, en roumain, un verbe du mme champ smantique, le quasisynonyme a ndeplini accomplir .
Les segments anaphoriques comments ci-dessus sont continus par les
expressions de laction proprement dite: En lui ceignant lpe, Usk kamar m
talvr ko ghumte samay et ncingndu-l cu spada, hyponymes actionnels de
faire ou factitifs implicites spcifiques.
4. Conclusion
Le verbe faire dmontre ainsi sa capacit de fonctionner aussi bien
comme verbe plein (ou lexical) de mme quen tant que verbe support .
Quand il semploie comme verbe support , il apparat dans des structures
verbo-nominales, mais aussi dans des constructions bi-verbales, en tant
quauxiliaire dagentivit initiale, mdiane ou finale, rles auxquels nous avons
ajout lemploi quasi-perlocutoire. Le prototype agentif fonctionne aussi
comme verbe plein quand il remplit le rle danaphore (lexicale ou
mronymique soumissive ) ou en tant que verbe autonome (non-substitutif).
SABINA POPRLAN
76
La comparaison avec le hindi et le roumain nous a dvoil les faits

suivants: dans les deux langues, les htronymes karn / bann et a face
acceptent lutilisation en tant que verbes support pour un nominal, mais
seulement le verbe roumain peut apparatre comme auxiliaire agentif dans les
structures co-verbales. En hindi, les diverses valeurs agentives sont rendues par
un verbe basique ou un deriv factitif ou causatif. Les verbes karn / bann et
a face semploient galement comme verbes pleins non-anaphoriques (en
hindi, surtout le second) et comme anaphores karn et a face.
SOURCES DES EXEMPLES
Cervants, LIngnieux Hidalgo Don Quichotte de la Manche, traduction en franais de Louis
Viardot, Garnier - Flammarion, Paris, 1969.
Cervantes, Dn Kvigjo, traduction en hindi par Chhavinath Pandey, de lAcadmie de
Littrature, Sahitya Akademi, Delhi, 1983.
Cervantes, Don Quijote, traduction en roumain par Ion Frunzetti et Edgar Papu, Editura pentru
Literatur, Bucarest, 1969.
*** vangile selon Matthieu , in Le Nouveau Testament , La Bible, Alliance Biblique
Universelle Le Cerf, Paris, 1995.
*** Evanghelia dup Matei, in Noul Testament, Biblia, Editura Institutului Biblic i de
Misiune al Bisericii Ortodoxe Romne, Bucarest, 1988.
*** Sant Matt ke anusr Susamcr , in Nay Vidhn, diteur Sataprakan Scrakendra,
1991.
Molire, Le Tartuffe, ditions Sociales, Paris, 1970.
Molire, Le Bourgeois gentilhomme, Nouveaux Classiques Larousse, Paris, 1965.
Molire ke do nak, traduction en hindi par B. M. Vajpeyi, de lAcadmie de Littrature, Sahitya
Akademi, Delhi, 1982.
Molire, Teatru (Tartuffe sau Impostorul, Burghezul gentilom), traduction en roumain par A.
Toma, Editura Univers, Bucarest, 1973.
BIBLIOGRAPHIE
Austin, John (1970), Quand dire, cest faire, traduction par Gilles Lane, ditions du Seuil, Paris.
Charolles, Michel (1990), LAnaphore associative. Problmes de dlimitation , in Verbum,
XIII, 3, pp. 119-148.
Cornish, Francis (1999), Anaphora, Discourse and Understanding. Evidence from English and
French, Clarendon Press, Oxford.
Cuni, Alexandra (2006), De quelques formes verbales du roumain et de leurs emplois
injonctifs , in Travaux et documents, 32, pp. 203-217.
Iliescu, Maria (1993), Le factitif roumain , in Revue roumaine de linguistique, 4, pp. 297-305.
Kacr, Yamuna (1973), Hind rpntaratmak vykara ke kuch prakara, Kendrya Hind
Sansthn, Agra.
Kleiber, Georges (1993), Lorsque lanaphore se lie aux temps grammaticaux , in C. Vetters
(diteur), Le Temps, de la phrase au texte, Presses Universitaires de Lille, Lille, pp. 117166.
Kleiber, Georges (1994), Contexte, interprtation et mmoire: approche standard vs. approche
cognitive , in Langue Franaise, 103, pp. 9-22.

77
Kleiber, Georges (1996), Anaphores associatives mronymiques: dfinition et proprits , in

Lexikalische Analyse romanischer Sprachen, Max Niemeyer Verlag, Tbingen, pp. 51-62.
Poprlan, Sabina (2010), LAnaphore verbale. Domaine typologique: franais, hindi, roumain,
Editura Univerisitii din Bucureti, Bucarest.
Searle, John (1982), Sens et expression. tudes de thorie des actes de langage, traduction en
franais par Jolle Proust, Minuit, Paris.
Theban, Laureniu (1980), Pour une smantaxe roumaine , in Revue roumaine de linguistique,
XXV, 1, pp. 23-36.
Theban, Laureniu (2007), Kraka, (deep) case, theta-role, Actant. Des termes la notion , in
Alexandra Cuni (diteur), Concepts trans- et interculturels, Editura Universitii din
Bucureti, Bucarest, pp. 65-92.
Theban, Maria, Laureniu-Ioan Theban (2005), Smantique et syntaxe du verbe faire en franais,
roumain, latin et portugais (I) , in Revue roumaine de linguistique, L, 1-2, pp. 57-71.
Tuescu, Mariana (1979), Prcis de smantique franaise, deuxime dition revue et augmente,
Editura Didactic i Pedagogic & Librairie C. Klincksieck, Bucarest & Paris.
Tuescu, Mariana (2007), LAuxiliation de modalit. Dix auxi-verbes modaux, Editura
Universitii din Bucureti, Bucarest.
DYNAMIQUE DU FIGEMENT : LADJECTIF EN FRANAIS

DANIELA BORDEA*
Abstract
The adjective has preoccupied grammarians since the dawn of time, having been an object
of study even since antiquity; along the years, we have witnessed an evolution of the conceptions
and analyses of the adjective, so that nowadays this category benefits from the latest
developments in language studies. Our research aims at analyzing the behavior of the adjective
along the lines of its transition from free combinatorial abilities, where the adjective is incidental
to the noun, to blocked combinatorial dynamics, where the adjective is an integral part. The two
types of combinatorial dynamics (free and blocked) are presented comparatively. We analyze the
transition of the adjective from the dynamics of free combinations to blocked combinations by
explaining the way of establishing blockage and stating the conditions and elements taking part in
the blockage. The blocking mechanism is also explained and illustrated. We define, explain and
illustrate the parameters that underline the complexity of the blocking process. The structures and
values of the blocked phrases are also stated. The paper analyzes the properties of the adjective in
blocked structures in contrast with the properties of the adjective in free combinations on the
basis of blocking tests.
Keywords: adjective, blocked combinatorial abilities, blocking, blocking mechanism,
blocking parameters, blocking degree, blocking tests.
1. Introduction
Ladjectif a constitu une proccupation pour les grammairiens ds
lAntiquit, aussi assiste-t-on une volution de la conception sur ladjectif et
des modles de description de cette partie du discours le long de lhistoire,
jusqu prsent.
Lobjet de notre recherche a t la variation du comportement de
ladjectif lors de son passage de la combinatoire libre o il est incident un
nom aux squences figes o ladjectif est partie composante.
*
Universit de Bucarest, Facult des Langues et Littratures trangres, Dpartement des

Langues Modernes, daniela.bordea@yahoo.com.
DANIELA BORDEA*
80
2. Recherches prliminaires
2.1. Le figement dans les approches traditionnelles
Les grammaires traditionnelles envisagent les expressions figes du point
de vue formel et fonctionnel. Cest aussi le cas des expressions figes contenant
des adjectifs, quand on fait la prcision que les lments constituants sont
gnralement eux-mmes des mots de la langue, possdant leur autonomie
syntaxique et appartenant une catgorie grammaticale dtermine. Par
exemple les lments de petit-beurre peuvent semployer sparment, le
premier comme adjectif et le second comme nom (Riegel et al. 1994 : 547).
Un regard comparatif sur le comportement de ladjectif dans la
combinatoire fige par rapport celui de ladjectif dans la combinatoire libre
met en vidence un comportement atypique de ladjectif dans les structures
figes : la place inhabituelle de ladjectif par rapport au nom (vif-argent,
chauve-souris, plat-ventre, faire la sourde oreille, pleurer chaudes larmes) ;
absence des degrs dintensit et de comparaison :
(1)
a.
b.
c.
d.
e.
un coffre fort /vs/ * un coffre assez fort

une porte cochre /vs/ * une porte trop cochre
tourner rond /vs/ * tourner plus rond
des toffes bleu clair / vs/ * des toffes trs bleu clair
* des toffes bleu trs clair
marcher plat-ventre /vs/ * marcher trs plat ventre
2.2. Le figement dans les approches structuralistes, distributionnelles

et transformationnelles
Dans ce type dapproches la combinatoire fige de ladjectif est
envisage du point de vue de la complmentation. Par exemple, pour la
structure N0 est Adj de N1, le complment de N1 peut tre appropri lorsque N1
correspond un seul type de substantif ou mme un seul substantif (Picabia
1978 : 84, 85, 95 et 97) :
(2)
a.
b.
c.
d.
e.
Ce livre est franc de port

Jean est perclus de rhumatismes
Paul est imbu de sa personne
Pierre est mort de fatigue
Michel est paralys de bonheur.
81
2.3. Le figement dans les approches syntactico-smantiques

Le processus du figement est envisag notamment en smantique
structurale du point de vue de lintgration smantique et lexicale des
constituants1 du syntagme fig/de lexpression fige. Ainsi, les lments en
combinatoire fige sont des lexies saisies globalement, le passage de la
combinatoire libre la combinatoire fige tant d la lexicalisation.
Dans ce cas le sens dune expression fige est conu comme une unit
(pismme), mais autre que la somme des sens de ses constituants immdiats.
Les lments lexicaux qui composent une expression fige correspondent donc
un seul pismme, bas sur des traits connotatifs, mais non obligatoires,
dpendant surtout des acquis socio-culturels des interlocuteurs.
Comme le fait remarquer Cuni (1980 : 201), dans ce cas, les composs
se caractrisent par un degr de cohsion trs lev; ceci fait quon ne peut ni
sparer les composants par un dterminant de ladjectif (*sang plus froid, *sang
trs froid), ni coordonner ladjectif composant un autre adjectif (*chaise
longue et verte).
On peut en conclure que les grammaires traditionnelles ont une vision
descriptive, statique sur le figement et le prsentent surtout du point de vue
formel; le point de vue fonctionnel et le smantisme se manifestent faiblement,
tandis que dans les autres types de grammaires, la vision sur le figement devient
plus dynamique; le centre dintrt se dplace vers le point de vue fonctionnel
et smantique, vers les transformations et les comparaisons.
3. tude de cas: ladjectif en franais
3.1. Le processus de figement; de la combinatoire libre la
combinatoire fige
Selon Tuescu (1978 : 91), loppos de la combinatoire libre, propre
la syntaxe, la combinatoire fige est le phnomne par lequel une squence cooccurrente de lexmes forme une unit indissociable sur les plans fonctionnel
(formel ou morphosyntaxique) et smantique .
la diffrence des squences libres, dont le sens global se calcule en
additionnant le sens de chacun de leurs lments, les expressions figes
nobissent pas la rgle de compositionnalit :
(3)
a.
b.
Marie a achet un cordon bleu pour sa jupe blanche.

Pierre a achet un panier pain perc, il ira se faire rembourser.
Cf. Tuescu (1978 : 90, 91 et 94).
DANIELA BORDEA*
82
Ces phrases sont libres parce que leur sens peut tre dduit partir du sens de
chaque mot qui les forme. Au contraire, les phrases
(4)
a.
b.
Marie est un cordon bleu.

Pierre est un panier perc.
signifient Marie est une bonne cuisinire et respectivement, Pierre est un

grand dpensier . Dans ce cas leur sens est imprdictible et incomprhensible
pour un locuteur moins avis, qui narrivera pas le dduire, moins que le
contexte ne lui donne des indices dinterprtation.
Nous analysons quelques exemples:
(5)
a.
b.
c.
Jai achet une table ronde /vs/ Jai organis une table ronde
Marie a achet un cordon bleu /vs/ Marie est un cordon bleu
Pierre a achet un panier pain perc /vs/ Pierre est un panier perc
Nous proposons le schma :
On peut donc considrer que dans le cas de ladjectif les critres

minimales ncessaires pour quon puisse parler de figement sont respectes.
Ces critres sont : (i) la squence doit tre forme de plusieurs mots ; (ii) les
mots impliqus dans le figement doivent avoir, par ailleurs, une existence
autonome ; (iii) lensemble se caractrise par un sens global qui nest pas la
somme des sens des constituants lis (non compositionnalit du sens de
lensemble).
3.2. Ralisation du figement : lments, conditions et mcanisme
Nous considrons que la relation de figement suppose lexistence des
lments suivants: (i) un lment central, A, qui reprsente une condition
ncessaire, mais non suffisante; (ii) un lment secondaire, B, qui reprsente
83
une condition ncessaire, mais non suffisante; (iii) un savoir qui relie les
deux lments et qui est une donne pragmatique. Cette donne, la
mmorisation (Svensson 2004 : 42), reprsente une condition non
ncessaire pour le figement syntagmatique mais une condition ncessaire pour
les autres types de figement. Elle fonctionne comme un liant qui ralise la
cohsion plus troite des lments impliqus dans la relation de figement : une
carte bleue est une carte ayant la couleur bleue, mais dont on sait que cest une
carte accrditive et de paiement, ainsi on ne peut pas dire ma carte bleue pour
nimporte quelle carte de couleur bleue ; donner ( quelquun) un carton rouge
ne signifie pas seulement lui donner nimporte quel carton qui soit rouge, mais
lui donner un mauvais point, porter un jugement dfavorable son encontre.
Cette donne pragmatique a le rle de limiter ou mme dannuler la
possibilit dun paradigme tel :
(6)
a.
b.
organiser une table ronde

*organiser une table carre / ovale / rectangulaire
et de rendre impossible le remplacement dun lment par un synonyme :

(7)
a.
b.
Pierre est un panier perc

*Pierre est une corbeille perce
ou par une unit de la mme famille :

(8)
a.
b.
un cerf-volant
*une biche volante
Nous proposons le mcanisme suivant pour la relation de figement :
DANIELA BORDEA*
84
Ce mcanisme fonctionne en feed-back: llment central exige un

certain lment secondaire et llment secondaire ne se combine quavec un
certain lment central pour garder le sens (et non avec des synonymes de celuici).
On peut en conclure que le figement se ralise selon un mcanisme
cyclique, ferm sur lui-mme.
3.3. Paramtres qui caractrisent le figement
Nous proposons les termes paramtres du figement pour dsigner le
degr de figement (que nous appelons paramtre intensionnel) et la porte du
figement (que nous appelons paramtre extensionnel).
3.3.1. Le degr de figement
Les degrs de figement illustrent le continuum entre les squences libres
et celles qui sont entirement contraintes.
Ainsi, une squence est dautant plus libre que le nombre de relations
linguistiques entre les diffrents composants est lev. Quand il ny a aucune
relation syntaxique entre les diffrents lments, la structure est totalement
fige. Cest cette variabilit qui permet de parler du degr de figement dune
suite donne et de faire la diffrence entre composition et figement (Gross
1988). Par exemple, certains adjectifs sont constitus de plusieurs mots sans
85
tre figs, dans la mesure o les lments constituent un paradigme, comme

cest le cas de la squence lail (Gross 1996: 98):
(9)
Cette tarte est lail
Dans le paradigme du substantif ail on pourrait avoir: pommes, prunes, etc. ;

mais pour ladjectif cran dans Pierre est cran, il ny a pas de possibilit de
permutation pour le dernier lment; on a alors affaire un adjectif fig. Pour
mieux illustrer ce fait nous proposons un exemple similaire :
(10)
a.
b.
Une table de bois / de marbre / de pierre

Une table de guingois (sans possibilit de paradigme).
Mais labsence de paradigme ne constitue pas un phnomne indpendant

dautres proprits: le sens de ces suites nest pas compositionnel, mais global,
et les proprits habituelles des adjectifs pithtes sont bloques. Dans ce cas on
a affaire un figement total, quand la suite concerne fonctionne de faon
compacte, en bloc, exactement comme les catgories simples (Gross 1996 : 16,
28 et 38). Par exemple, un groupe nominal du type Dt N Adj. est compos de
trois constituants. Mais une suite comme un cordon bleu, bien quayant le
mme nombre dlments lexicaux, na que deux constituants: un dterminant
et un bloc unique soudant en une seule unit les mots cordon et bleu. Ici,
ladjectif ne joue pas le rle dun modifieur. Sur le mot cordon on peut donc
construire : (i) des groupes nominaux libres ordinaires : un cordon solide,
un cordon de trois mtres, etc. ; (ii) un groupe nominal semi-fig : un cordon
lectrique (nom compos) ; (iii) un nom compos fig : un cordon (-) bleu (une
bonne cuisinire). Gross (1988 : 63) en conclut que le figement nest pas une
valeur absolue, mais relve dune gradation correspondant des proprits
transformationnelles potentielles ralises des degrs diffrents . Il en
dcoule que ce sont les proprits syntaxiques du groupe ou de la squence qui
permettent de calculer le degr de figement; la relation est inversement
proportionnelle: plus le groupe est souple du point de vue syntaxique, moins il
est fig.
Les composs substantif-tte sont appels par la tradition linguistique
des composs endocentriques et tous les autres types de composs sont appels
exocentriques.
Lorsque ladjectif nest pas entirement fig du point de vue syntaxique
et smantique (comme cest le cas des composs endocentriques), il peut
dsigner, aprs un substantif-tte, une spcification dans le cadre dune
typologie: casque lger, accent aigu, accent circonflexe. La relation entre le
nom et ladjectif nest pas opaque (un accent aigu est un accent), (figement
transparent) ; ladjectif dsigne donc un trait caractristique de lobjet, qui
permet de lidentifier parmi dautres appartenant la mme famille. Ces
DANIELA BORDEA*
86
adjectifs sont considrs comme des tiquettes et non comme des qualits
(Gross 1996 : 51).
Les composs exocentriques sont figs du point de vue syntaxique et
smantique. Ce sont des suites geles qui ne diffrent dun mot simple que
par leur polylexicalit et les marques morphologiques spcifiques aux composs
(Gross 1996 : 35-36), comme par exemple panier perc (figement opaque) :
(11)
a.
b.
Luc est un (panier perc, * panier)

* Un panier perc est un panier
Selon nous le domaine du figement comporte trois divisions : (i) figement

faible ; (ii) figement transparent ; (iii) figement opaque. Le figement se ralise
selon le mcanisme prsent sous 2.2.
Dans le cas dun syntagme fig structure binaire Adj + Nom llment
central est un substantif et llment secondaire est un adjectif qui peut suivre
ou prcder le substantif :
(12)
a.
b.
a.
b.
(13)
cordon bleu
panier perc
rouge-gorge
chauve-souris
Pour mettre en vidence les diffrents degrs de figement pour ce type de

structure, nous proposons une grille de trois tests. La grille se prsente ainsi :
Test I () :
Test II () :
Test III (+) :
un A est / nest pas un A ;

un A a / na pas la qualit dsigne par B ;
il existe un savoir (une donne) pragmatique qui relie B A (la
mmorisation) ; lexistence de cette donne pragmatique
reprsente la condition obligatoire pour quil y ait figement2.
Cest pour cela que la rponse ce test doit tre toujours (+), donc seulement le
test I et le test II peuvent admettre des rponses variables. Dans ce cas, selon
une formule mathmatique de lanalyse combinatoire on a : 2n = 22 = 4 variantes
possibles. Les quatre variantes, notes : (1), (2), (3), (4) sont (Schma 3)
Voir les explications sous 3.2. et Schma 2.
87
Du point de vue linguistique la variante (4) nest pas possible parce que si la
rponse au test I est ngative :
Test I () :
Test II () :
un A nest pas un A, alors la rponse au test II doit tre toujours

elle aussi ngative ;
un A na pas la qualit dsigne par B.
Voici quelques exemples dapplication des tests (voir Schma 4):

(i) figement faible
Test I (+) :
un vin rouge est un vin ;
Test II (+) :
un vin rouge a une couleur proche du rouge ;
Test III (+) : on dit vin rouge pour dsigner un vin dont la couleur est
proche du rouge.
Dans ce cas de figement faible le sens du syntagme fig est quasi
compositionnel.
(ii) figement transparent
Test I (+) :
les vers blancs sont des vers ;
Test II () :
les vers blancs nont pas la couleur blanche
Test III (+) :
on dit vers blancs pour des vers qui ne riment pas.
DANIELA BORDEA*
88
(iii) figement opaque

Test I () :
un cordon bleu nest pas un cordon
Test II () :
un cordon bleu na pas la couleur bleue
Test III (+) :
on dit cordon bleu pour dsigner une bonne cuisinire.
Nous remarquons que le degr de figement augmente du figement faible au
figement opaque, donc le plus bas degr de figement (figement faible)
correspond trois rponses positives, cas o le sens du syntagme fig est quasi
compositionnel. Le figement faible reprsente donc un tat intermdiaire entre
la compositionnalit et le figement proprement dit. Au fur et mesure que le
degr de figement augmente, les deux premires rponses deviennent ngatives,
de sorte quau plus haut degr de figement (figement opaque) on a seulement la
troisime rponse positive. Nous remarquons que la division 3 reprsente le
plus haut degr sur laxe du figement.
3.3.2. La porte du figement
Il est possible quune chane donne soit totalement fige lorsque le
figement affecte lensemble de la squence (cordon-bleu, col-vert, panier
perc), ou partiellement fige, lorsque le figement affecte un seul sousensemble de la squence donne, tandis que le reste relve dune combinatoire
libre (chevaucher bride abattue, rouler tombeau ouvert).
Ltendue, calcule en nombre de mots, de la squence soude,
reprsente la porte du figement (Gross 1996 : 38) : blanc comme neige (3
units); marcher plat ventre (4 units).
Le figement partiel, qui ne concerne pas le substantif-tte, peut tre
considr comme priphrique par rapport au noyau du groupe nominal.
3.4. Structures des expressions figes contenant des adjectifs
Les adjectifs peuvent entrer dans des expressions figes aux structures varies.
Dans ce qui suit, nous allons en donner un inventaire qui ne se veut pas exhaustif.
(14)
(15)
(16)
ADJECTIF + ADJECTIF
a.
un fruit aigre-doux
b.
un enfant sourd-muet
ADJECTIF + NOM COMMUN
a.
blanc-bec
b.
un livre bon march
NOM COMMUN + ADJECTIF
a.
racine carre
b.
col-vert
c.
avoir une peur bleue
d.
chaise longue
(17)
(18)
(19)
(20)
(21)
(22)
(23)
89
NOM PROPRE + ADJECTIF (M. Grevisse; 1988: 533):

a.
Philipe le Bel
b.
Charles le Tmraire
c.
Ivan le Terrible
d.
Marne-la-Coquette
e.
Brive-la-Gaillarde
f.
Noisy-le-Grand
ADJECTIF + ADJECTIF ADVERBIALIS
a.
des cheveux coups court
b.
une moustache coupe ras
VERBE + ADJECTIF ADVERBIALIS
a.
se faire fort de
b.
se porter fort pour
c.
penser juste
d.
tenir bon
e.
tenir ferme
f.
marcher droit
g.
voir clair
h.
tourner rond
ADJECTIF + PRPOSITION + NOM
a.
bleu de froid
b.
rouge de colre
c.
mort de fatigue
d.
paralys de bonheur
ADJECTIF + PRPOSITION + VERBE
a.
bte pleurer
b.
laid hurler
c.
fou lier
ADVERBE + ADJECTIF
a.
des gens malintentionns
b.
un enfant bien portant
ADJECTIF + COMME + GN
a.
aimable comme une porte de prison
b.
bavard comme une pie
c.
blanc comme neige
d.
clair comme le jour
e.
doux comme un agneau
f.
mchant comme un ne rouge
g.
pauvre comme Job
h.
sage comme une image
3.5. Valeurs des expressions figes contenant des adjectifs

Les exemples ci-dessous illustrent les valeurs des expressions figes
contenant des adjectifs :
(24)
valeur nominale :
a.
cordon bleu
b.
col-vert
DANIELA BORDEA*
90
(25)
(26)
(27)
c.
panier perc
d.
chaise longue
e.
cerf-volant
f.
court-circuit
valeur adjectivale adjectifs composs :
a.
b.
un fruit aigre-doux
c.
une blouse jaune paille
d.
une jupe bleu fonc
valeur adjectivale locutions adjectivales :
e
de bas tage
f.
de bon poil
valeur verbale (locutions verbales) :
a.
en avoir le cur net
b.
faire la sourde oreille
valeur adverbiale
petit petit
3.6. Proprits des expressions figes contenant des adjectifs. Tests de

figement
Nous allons analyser les proprits des expressions figes par rapport aux
proprits des adjectifs en combinatoire libre.
Les constructions libres ont des proprits transformationnelles qui
dpendent de leur organisation interne. Dans le cas des expressions figes,
lopacit smantique est corrle une absence de proprits
transformationnelles et il y a une relation de proportionnalit inverse entre le
figement dun groupe et le nombre de proprits transformationnelles
observables (Gross 1988 : 69). Quand une suite donne se prte toutes les
modifications envisages, le sens est totalement compositionnel et lon parlera
dun groupe ordinaire. Inversement, si aucune des proprits nest ralisable,
alors il est lgitime de parler de figement.
Dans une squence fige aucun des lments lexicaux constitutifs ne peut
tre actualis individuellement, mais ils ont une dtermination globale:
(28)
a.
b.
c.
d.
un rouge-gorge
*un rouge la gorge
*un rouge cette gorge
*un rouge sa gorge
Entre les diffrents lments qui sont dans la porte du figement il ny a

pas de relation prdicative:
(29)
a.
b.
un panier perc
*ce panier est perc
(30)
(31)
a.
b.
a.
b.
91
un rouge-gorge
*cette gorge est rouge
un blanc-bec
*ce bec est blanc
Cependant, lorsque le sens nest pas opaque, le compos est moins fig,
ladjectif dsigne un type particulier, une varit par rapport aux autres de la
mme catgorie:
(32)
(33)
a.
b.
c.
a.
b.
c.
un accent aigu
*cet accent est aigu
ceci est un accent aigu
un vin blanc
*ce vin est blanc
cest du vin blanc
Il est remarquer quun groupe nominal ordinaire, form dun nom et dun
adjectif est en fait le rsultat dune phrase, tandis quune expression fige ne
lest pas (Gross 1996 : 51):
(34)
(35)
a.
b.
a.
b.
une arme dangereuse

cette arme est dangereuse
une arme blanche
*cette arme est blanche
Ladjectif affect par le figement ne peut pas tre nominalis:

(36)
(37)
a.
b.
a.
b.
un panier perc
*le perc de ce panier
un bas-bleu
*le bleu de ce bas
Les expressions figes (dans leur totalit) et les adjectifs impliqus dans
le figement ne reoivent pas de gradation ou dadverbe dintensit:
(38)
(39)
(40)
(41)
a.
b.
c.
d.
a.
b.
c.
d.
a.
b.
c.
d.
a.
cause dun court-circuit la lumire sest teinte.

* cause dun trs court circuit la lumire sest teinte.
* cause dun circuit trs court la lumire sest teinte.
* cause dun circuit particulirement court la lumire sest teinte.
Cet enfant est doux comme un agneau.
*Cet enfant est trs doux comme un agneau.
Cet enfant est trs doux, doux comme un agneau.
Cet enfant est extrmement doux, doux comme un agneau.
Jean est laid faire peur.
*Jean est trs laid faire peur.
*Jean est laid faire trs peur.
*Jean est laid faire beaucoup de peur.
Ce garon est bte pleurer.
DANIELA BORDEA*
92
b.
c.
*Ce garon est assez bte pleurer.

*Ce garon est bte assez pleurer.
Le syntagme adjectival bte pleurer est en voie de lexicalisation; pleurer

nest presque plus senti comme un complment de bte, mais il revt une valeur
de superlatif (selon Goes 1999 : 177). Mais on peut dire : de leau distille, de
leau bi distille, sans pouvoir dire *de leau trs distille. Cependant, dans le
cas dune expression comme de bonne humeur, cest la relation entre de et
humeur qui est fige; la position adjectivale (qui est obligatoire) fait lobjet dun
paradigme de (bonne, mauvaise) humeur et permet ce niveau linsertion
dun quantifieur: de trs bonne humeur (Gross 1996 : 19).
Dans les squences figes linsertion dlments nouveaux est trs
rduite. Les expressions figes sont des suites bloques, que le locuteur ne peut
pas modifier :
(42)
a.
b.
c.
d.
un compte-rendu
*un compte vite rendu
*un compte bien rendu
*un compte correctement rendu
Lordre des lments qui composent une expression fige ne peut pas tre
chang:
(43)
a.
b.
c.
d.
e.
f.
Pierre a calcul la racine carre de ce nombre

*Pierre a calcul la carre racine de ce nombre
Il sest produit un court-circuit
*Il sest produit un circuit-court
Ils sont rentrs sains et saufs
*Ils sont rentrs saufs et sains.
Goes (1999: 223) fait la prcision que dans ce cas les adjectifs fonctionnent
simultanment comme attributs par rapport au syntagme nominal et comme
adverbes par rapport au verbe :
(44)
a.
b.
Cet enfant est sourd-muet

*Cet enfant est muet-sourd.
En ce qui concerne la motivation de lordre des adjectifs, Noailly (1999: 47)

considre que la priorit matrielle de lun des adjectifs a une incidence sur leur
poids relatif dans la smantique de leur relation et confre au premier des deux
un sens prioritaire, que le second vient seulement modifier. Par exemple, dans
le cas de un enfant sourd-muet on comprend que la maladie de la mutit vient
sajouter celle de la surdit, comme une consquence. Cependant, la
qualification bleu-vert applique un objet donn peut tre conteste, en
montrant que lobjet est plutt vert-bleu.
93
Dans le cas des expressions formes dun adjectif suivi dun complment,
Wilmet (1997 : 214) remarque la solidarisation de ladjectif avec son
complment :
(45)
a.
b.
un tableau beau couper le souffle

*un beau tableau couper le souffle
Le degr de figement peut tre mis en vidence par la formule un X est un X :

(46)
a.
b.
c.
d.
*un panier perc est un panier

*un cerf-volant est un cerf
*une toile cire est une toile
un vin rouge est un vin
e.
une lettre recommande est une lettre
(fig) (figement opaque)

(sens quasi compositionnel)
(moins fig)
(sens quasi compositionnel)
(moins fig).
Les expressions figes nadmettent pas la relativisation:

(47)
(48)
(49)
(50)
(51)
(52)
(53)
a.
b.
a.
b.
a.
b.
a.
b.
c.
a.
b.
c.
a.
b.
a.
b.
c.
une chaise longue

*une chaise qui est longue
un pont-levis
*un pont qui est levis
des dpouilles opimes
*des dpouilles qui sont opimes
des fentres grandes-ouvertes
*des fentres grandes qui sont ouvertes
*des fentres ouvertes qui sont grandes
une soie gris de lin
*une soie grise qui est de lin
*une soie de lin qui est grise
une chauve-souris
*une souris qui est chauve
des fruits aigres-doux
*des fruits aigres qui sont doux
*des fruits doux qui sont aigres
Cependant, lorsque les deux adjectifs formant un compos (suite Adj +

Adj) dsignent des qualits indpendantes, on peut avoir:
(54)
a.
b.
c.
un enfant sourd qui est aussi muet
un enfant muet qui est aussi sourd
Nous remarquons que si ladjectif dsigne une qualit inhrente dun

substantif, alors la relative doit tre mise en apposition pour viter le plonasme
:
(55)
a.
b.
*rouge comme une tomate qui est rouge

rouge comme une tomate, qui est rouge
DANIELA BORDEA*
94
c.
d.
e.
f.
*rapide comme lclair qui est rapide

rapide comme lclair, qui est rapide
*rond comme une bille qui est ronde
rond comme une bille, qui est ronde
Mais les expressions moins figes peuvent admettre la relativisation :

(56)
a.
b.
une lettre recommande

une lettre qui est recommande
tant donn que la relation entre le nom et ladjectif est restreinte, il ny a

pas de possibilit de coordination avec un autre adjectif:
(57)
a.
b.
c.
d.
e.
f.
un cerf-volant
*un cerf-volant et beau
*un cerf beau et volant
une toile cire
*une toile cire et blanche
une toile blanche et cire
Mme remarque pour les suites V + Adj (Tuescu 1978: 92) :

(58)
a.
b.
c.
voir clair
voir dun bon il
*voir clair et dun bon il
On ne peut pas reprendre la base seule, comme substitut gnrique, ni la

pronominaliser (Tuescu 1978: 92) :
(59)
a.
b.
*Elle a achet une chaise longue et elle a mis cette chaise dans sa chambre
*Il a fait la sourde oreille, mais toi, tu ne las pas faite.
Cependant, pour les expressions moins figes on peut avoir: Il a achet du vin
rouge et il la mis sur la table.
Etant donn le fait que le sens global dune expression fige correspond
un concept existant dans la langue (Gross 1996: 42), lexpression fige peut
commuter avec un quivalent synonyme: une fine bouche commute avec
gourmet. Cependant il est remarquer qu lintrieur des suites figes la
possibilit de substitution synonymique ou par des units de la mme famille est
exclue :
(60)
(61)
(62)
a.
b.
a.
b.
a.
un court-circuit
*un bref-circui
clair comme le jour
*clair comme le matin
un cerf-volant
b.
95
*une biche volante
Une squence en combinatoire fige a rarement des correspondants

contraires terme terme :
(63)
(64)
(65)
(66)
Combinatoire libre :
a.
une grande fentre /vs/ une petite fentre
b.
une fentre ouverte /vs/ une fentre ferme
Combinatoire fige :
a.
une fentre grande-ouverte
b.
*une fentre petite-ferme
a.
clair comme le jour
b.
*sombre comme la nuit
a.
fort comme la mort
b.
*faible comme la vie
Les expressions figes nacceptent pas la passivation :

(67)
a.
b.
Jean a fait la sourde oreille

*La sourde oreille a t faite par Jean.
4. Conclusions
Notre tude sur le figement dans la classe de ladjectif franais a mis en
vidence quelques conclusions : (i) le figement est un processus dynamique qui
se ralise selon un mcanisme cyclique, ferm sur lui-mme, qui fonctionne en
feed-back ; (ii) le figement se caractrise par un paramtre intensionnel (le
degr de figement) et par un paramtre extensionnel (la porte du figement) ;
(iii) le figement peut tre mis en vidence par des tests de figement.
Quelle quelle soit, lexpression fige rend le message plus clair, surtout
quand elle sert illustrer des concepts abstraits, et ainsi elle amliore et facilite
le processus de communication. Linterlocuteur saisit plus facilement les
images mentales que les abstractions et cela justifie le fait que les expressions
figes utilisent des lments constitutifs pris au vocabulaire fondamental de la
langue, le plus concret et le plus susceptible veiller des reprsentations
mentales.
BIBLIOGRAPHIE
Cuni, Alexandra (1980), La formation des mots. La drivation lexicale en franais
contemporain, Editura Didactic i Pedagogic, Bucarest.
Goes, Jan (1999), Ladjectif. Entre nom et verbe, Duculot, Paris.
96
DANIELA BORDEA*
Grevisse, Maurice (1988), Le bon usage, 12e d., (revue par Andr Goosse), Duculot, Paris.
Gross, Gaston (1988), Degr de figement des noms composs , in Langages, 90, Larousse,
pp.57-72.
Gross, Gaston (1996), Les expressions figes en franais, noms composs et autres locutions,
Ophrys, Paris.
Noailly, Michle (1999), Ladjectif en franais, Ophrys, Paris.
Picabia, Llia (1978), Les constructions adjectivales en franais. Systmatique
transformationnelle, Droz, Genve.
Riegel, Martin, Jean-Christophe Pellat, Ren Rioul (1994), Grammaire mthodique du franais,
Presses Universitaires de France, Paris.
Svensson, Maria-Helena (2004), Critres de figement, Ume Universitet, Ume,.
Tuescu, Mariana (1978), Prcis de smantique franaise, Editura Didactic i Pedagogic,
Bucarest.
Wilmet, Marc (1997), Grammaire critique du franais, Duculot, Paris.
GAIRAIGO WITHIN JAPANESE LANGUAGE:

LANGUAGE SUICIDE OR CASUAL CULTURE?
FRANCESCO VITUCCI*
Abstract
This paper stems from didactic module conducted in the fall semester of the academic
year 2012-2013 in the course of Japanese Philology in the Department of Foreign Languages and
Literature of Alma Mater Studiorum Bologna University, Italy. It aims at introducing the
Japanese scholars different positions on the issue of loanwords within Japanese language from a
historical and sociolinguistic perspective. Successively, contemporary issues concerning garaigo
will be analysed from the following perspectives: (i) Can gairaigo be considered as an example
of language suicide and as an outcome of the English language imperialism?; (ii) Can gairaigo
be considered as the expression of a casual bilingualism that enriches the Japanese lexical
panorama and augments mutual intelligibility among intercultural speakers?; (iii) Are gairaigo
useful to the lexical enrichment of Japanese language if analysed in a mere Japanese
communication context? The above issues will be analysed in order to understand the current
position and further developments of katakanago within Japanese language.
Keywords: Japanese, gairaigo, katakanago, wasei eigo, sociolinguistics.
1. Introduction
According to the sociolinguist Sachiko Okamoto (2008), Japan can be
defined as a monolingual country since, leaving aside the language policies
followed by the government in this last century, it is geographically in a
position of isolation that accentuates these characteristics. However, the
increase of loanwords together with the lack of their standardization in
transcription over the years, has raised many discussions among scholars and
within the Japanese government so much as to talk about aGairaigo
Hanran foreign words flood. According to the dictionary by Kno et al.
(1996), one should consider gairaigo all the terms entered in the archipelago
during the Muromachi period after the first contacts with European merchants
and missioners and, secondly, loanwords imported from Europe and America
*
School of Languages and Literature, Translation and Interpretation, Alma Mater
Studiorum Bologna University, Department of Asian and African Studies, Ca Foscari University,
Venice, fvitucci@gmail.com.
FRANCESCO VITUCCI
98
starting from the Meiji period (mainly from English). Conversely, are not to be
considered gairaigo, the kango of Chinese origin. Unlikewasei eigo
terms, gairaigo mainly designate loanwords imported from foreign
languages and conventionally written in katakana. The first garaigo reached
Japan in the late medieval period, brought by the Portuguese and the Dutch1.
The use of katakana was due to the fact that in Japan since the Heian period this
syllabary had been utilized for the practice of kanbun kundoku in order to insert
auxiliary glosses within Chinese classics ().
2. A brief history of garaigo
Today, scholars are still arguing on the issue of garaigo transcription2.
The current mode of gairaigo transcription was sanctioned in 1991 by the
ministerial document Gairaigo no Hyki. However, since no fixed
rules for transcription can be found in this document, the reader is often left
confused. Obviously, given the absence of a clear framework and having to rely
on subjective transcription choices, it is clear that the probability that these
terms will be reproduced incorrectly is no doubt very high. In fact, as described
above, the use of katakana for transcription had already begun during the
Muromachi period with the translation of Dutch and Portuguese lexicon, even
if, at that time, most of these words were still transcribed through kanji or
hiragana. It is only during the Edo period that katakana are officially
formalized for the transcription of loanwords. Arai Hakuseki () and
Sugita Genpaku () set the rules for transcription in their works
(Seiy Kibun, 1715) and in (Kaitai Shinsho, 1774), the translation of
Ontleedkundige tafelen, from Dutch. Among the transcription parameters
suggested by Arai Hakuseki, it is mentioning, as an example, the line of
elongation for long vowels ().
Among these one can mention: tobacco, soap,

coloured sugar candies, from Portuguese; shovel, chalk, music
box, from Dutch; clinical records, compressed gas cylinder,
neurosis, ideology, from German; summary, fianc,
parfait, encore, from French.
2
The following list contains a number of wrong gairaigo frequently encountered in
Japanese conversation: instead of bed; instead of dog
food; instead of tea bag; instead of
news;instead
of simulation;
instead
of
narcissist; instead of Budapest; instead of
badminton; instead of Cupid.

99
Although during the Edo period one normally resorted to the use of kanji
with katakana rubi for writing foreign words, during the period of the socalledBunmei Kaika, all gairaigo terms were finally transcribed
through katakana trying to follow the phonetics rules of the languages which
they came from. Between the beginning of the Meiji period and Taish period,
due to an increase of foreign lexicon imported from several research branches
often translated throughshinkango, Japanese speakers started to feel
uncomfortable with the ideographic writing and began to replace kanji
transcriptions with katakana. During the Second World War foreign words
were banned, but they soon returned to prominence with the American
occupation of Japan. Thanks to the so-called policy of language
renewal, Latin characters were reintroduced in the elementary schools together
with the teaching of English in junior high school. In the daily life of the
Japanese, the habit of dealing with road signs, place names, station names
transcribed with the Latin alphabet became widespread, with billboards written
in English so as not to feel the presence of English and rmaji as intrusive and
alien. It is in this climate that wasei eigo terms permeated the Japanese
language due to the dominance of English in Japan (Okamoto 2008).
After the economic boom reached by Japan in the Fifties, Western
products and ideas invaded the Japanese archipelago. From this point on, the
term katakanago was coined to indicate the transcription not only
of loanwords, but also of onomatopoeia, dialect expressions, oral and written
language produced by foreign speakers, hybrids terms, names of animals and
plants, technical jargon, obscene language, proper names of people and states,
toponyms entirely transcribed with katakana (see e.g. Kaneya 2002).
Nevertheless, the increase of gairaigo was perceived in Japan as a threat to the
survival of the Japanese language. In the Fifties, it comes
togairaigo rany gairaigo misuse, in the Sixties to
gairaigo kzui gairaigo flood, and in the Seventies
togairaigo hanran gairaigo overflowing. In the Sixties,
theKokuritsu Kenkyjo began their investigation on gairaigo
entitled (Gairaigo no rikaichsa) which revealed garaigo
intelligibility disparities among Japanese due to generational gaps and lifestyle
differences between the town and the countryside (Jinnouchi 2007). In the
Eighties, the myth of thekokusaijin spread all over the country and the
first gairaigo dictionaries began to be published (Tanaka and Tanaka 1996).
According to them3, the percentage of loanwords, if compared to late
nineteenth century Japanese dictionaries, had risen from 1.4% to 10%.
3
Statistical surveys are also being conducted through television and reveal how the
presence of gairaigo is strongly influenced by the type of TV program broadcast. Compared to
programs that do not use loanwords (such as jidaigeki, sum matches, programs with a
FRANCESCO VITUCCI
100
3. Japanese surveys and anti-gairaigo measures
In 1989 Junichiro Koizumi founded theYogo Tekiseika

Iinkai committee in order to limit loanwords in public offices and in 1997 as
Minister of Health, Labour and Welfare added vigor to the
Katakanago Tsuiky Und movement in order to replace
those gairaigo, which were not understandable to senior citizens, with Japanese
terms. All this, however, was unable to contain the increase of katakanago so
that the phenomenon was brought to the fore by a New York Times article
(2/21/1995) that made fun of wasei eigo terms, failing to understand the
function they play within Japanese language and disregarding the fact that in
Japan they are not considered as English terms (Kristof 1995). After
theGairaigo no hyki was promulgated in 1991 in order to clarify
the transcription of loanwords in Japanese, in 1997 a survey by the Japanese
Ministry of Culture entitledKokugo ni kansuru yoronchsa
showed that 90% of respondents, especially, the older segment of the population
located in rural areas, had difficulty in understanding garaigo (Figure 1).
Figure 1. Data from the survey

Kokugo ni kansuru yoronchsa (from Okamoto 2008)
In 2002, Koizumi, at the time Prime Minister of Japan, founded

theGairaigo Iinkai committee in order to draft a document aimed
pedagogical background), there are others who use gairaigo in large quantities such as: formula 1
races (25.8%), music programs with Japanese idols (25.7%), and fashion programs (23.2%).

101
at finding viable solutions to the problem. The results of the surveys and
research
were
first
collected
in
2003
in
a
dossier
entitled
Gairaigo Iikae Teian wakarinikui gairaigo wakariyasuku
wo suru tame no kotobazukai no kuf.
The dossier was meant to replace
gairaigo with Japanese terms by adding explanatory comments where
necessary. It also insisted that the comprehensibility of gairaigo is lowered by
generation gaps, sex, type of profession and place of residence in the country .
This statement is based on the fact that gairaigo normally enter the Japanese
language from technical jargon and that they then gradually filter downward
into the common language. Therefore, according to the dossier, it would be
necessary to distinguish the various contexts in which garaigo are utilized
introducing Japanese translations and explanations when possible. Even though
the activity of the committee continued until 2006, in the end, not a lexical item
of those examined was replaced (Figure 2).
Figure 2. Data from the Gairaigo Iikae Teian dossier
Experiments of garaigo replacement were also carried out by the

newspaper Yomiuri through an activity entitled Shin nihongo no
genba (Hashimoto 2003) which was aimed at fostering the digital literacy of the
elderly bands of population. Another pamphlet entitled
Jijibaba no tame no pasokon shinan was published after this
experiment. In 2007, the Japanese Ministry of Culture conducted its
investigation on gairaigo entitled Gairaigo ni kansuru
FRANCESCO VITUCCI
102
yoronchsa that revealed interesting trends in Japanese society. In fact, one

began to notice less aversion towards loanwords by older generations and, in
particular, by women. These data revealed that with the passing of generations,
intransigence towards gairaigo was diminishing.
4. Yji Suzuki and the Casual Bilingualism
The linguist Yji Suzuki (2003) states that the reason for the lack of
English skills in Japan would lie in the fact that despite the huge presence of
gairaigo from English and wasei eigo, English is only offered as a formal
second language within the Japanese school curriculum and always in a
monolingual environment4. Nevertheless, thanks to mass media, nowadays the
new generations of Japanese are increasingly exposed to a greater number of
gairaigo living often in a situation of casual bilingualism. Although this
situation has been stimulated in the twenty-first century by the need to launch
Japanese products on the international market, in fact, even during ancient times
and the Meiji period, Japanese used to import loanwords but always
katakaniz-ing them without introducing English as a real vehicle of education.
Suzuki suggests that the introduction of gairaigo into Japanese, did not affect
the structure of the language itself. This is because there is too much structural
distance between the Indo-European language families and the Ural-Altaic one
to which Japanese belongs (Palmer 1979). For this reason, the increase of
gairaigo should be considered as a chance of lexical enrichment rather than as a
threat. To solve communication problems that gairaigo often produce in daily
conversation with foreigners, Suzuki proposes to pronounce them according to
their original language pronunciation and not to transcribe them in katakana,
but rather with the alphabet. This position seems to exceed that of Gairaigo no
hyki, which offers transcriptions based solely on Japanese phonetics.
Moreover, these statements are backed up by a research carried out by Suzuki
himself which showed how a sample of Japanese adults could not read English
texts written in alphabet, but only versions transcribed in katakana in which
they were able to track a greater number of gairaigo. Surely, many of the
gairaigo introduced from English have served the needs of the Japanese
commerce. Suzuki suggests that terms such as TV, air
conditioning, antenna, brake, tyre have been
imported from English in order to avoid confusion on the market. However, this
phenomenon also presents an active side whenever Japanese lexicon is exported
abroad without being translated (let us think of the so-called gaikgo
According to Suzuki, in the case of former colonies, the situation would be different.

103
phenomenon)5. Given this ongoing lexical globalization, Suzuki puts in contrast

the so-called casual cultures with frozen cultures (see Hall 1976). On the
one side, the casual culture would contain all the English lexicon shared on an
international level (for example: burgers, jeans, rock) together with all those
linguistic habits born within oral language and gradually penetrated into the socalled frozen culture. In particular, the latter refers to the formal culture that
receives new linguistic habits from the latest linguistic trends from generation
to generation. According to Suzuki (2003: 102), this antonymy would reveal the
possibility of exporting lexical items from one source language to a given target
language whenever a lexical gap occurs. This process would foster the so-called
casual bilingualism thanks to which each speaker in the world could share a
common lexicon base composed by international terms coming from several
languages. From a Japanese perspective, Suzuki suggests that, thanks to the
massive presence of gairaigo, this quasi-bilingual condition should already be
part of the Japanese language environment and that it could be further exploited
to boost English language learning in Japan and to reach an almost complete
bilingual environment such as those of former colonial countries.
5. Fumio Inoue and language suicide
Unlike Suzuki, the sociolinguist Fumio Inoue (2001) argues that the
gradual invasion of gairaigo into the Japanese language has increased the
absolute level of complexity of Japanese. This would be due to the increase of
synonymic terms. Nevertheless, even though absolute complexity increases, as
in the case of expressions like substituted by mail, replaced
by address, instead of hotel, or with terms with the
same root yielding different gairaigo, e.g. titanium Titan
Titanic, the relative complexity of Japanese language
decreases in the case of foreign speakers (especially for English-native
speakers). However, as already mentioned, there are several problems of
semantic shift between Japanese loanwords and their original terms in English.
This would force even native speakers to use bilingual English-Japanese
dictionaries to understand the many neologisms expressed through gairaigo.
According to Inoue (2001), this problem is due to the fact that from the Meiji
period new technical terms imported from foreign languages were no longer
translated into kango. Today, this problem would be further exacerbated due to
5
Let us think of terms such as walkman, stereo

headphone, but also of other culturally specific terms such as sushi, teriyaki,
geisha, and the like.
FRANCESCO VITUCCI
104
generation gaps among Japanese speakers. Nevertheless, as already mentioned,

young people seem more prone to the introduction of gairaigo. In fact, on the
occasion of the above mentioned 1997 survey carried out by the Japanese
Ministry of Culture (), to the
question what do you think about the increase of gairaigo?, more than 60% of
respondents answered it is not a problem if they increase. In particular, the
highest approval was expressed by young women with an 80% rating.
According to Inoue, these data should be considered as the proof of a linguistic
globalization which reveals, in turn, a lack of sensitivity of Japanese speakers
towards this issue. Moreover, Suzuki argues that with the increase of gairaigo
more and more new pronunciations have permeated the phonetic Japanese
system. This results in a lowering of the absolute difficulty of the language, but
also in an increase of the relative difficulty for Japanese speakers caused by
generation gaps. In fact, although new transcriptions of series such as [va],
[vi], [vu], [ve], [vo] have been introduced, older speakers still
have difficulty in utilizing them, since they cannot distinguish the consonants
[v] and [b] (just as they cannot distinguish [r] and [l]). In his essay suggestively
entitledNihongo wa ikinokoreruka, Inoue points out that
the presence of gairaigo has also increased the translation of movie titles from
abroad. If during the period before the second World War titles were literally
translated, with the time English has gradually replaced Japanese. At first, only
lexical substitutions occurred, but eventually entire sentences were transcribed
in katakana to slowly drift towards rmaji. According to data collected by Inoue
(2001), from the Fifties until the late Nineties, the percentage of movie titles which
includes gairaigo, e.g. in Figure 4, increased from 10 to more than 50%.
Figure 4. Translation of movie titles in katakana

105
This increase is certainly due to marketing issues. Katakanago titles, in

fact, make a better grip on the audience. That is why nowadays English tout
court is preferred to Japanese translations. The same phenomenon is also
observed for TV program titles (which use more and more gairaigo and
transcriptions in rmaji) and for J-pop music lyrics where rmaji lexicon
appear both in titles or within texts interspersed with Japanese syntax.
According to Inoue, as early as the European colonization of the Americas and
the late nineteenth century European colonization of Africa and Oceania, Latin
characters found themselves in a situation of dominance against local writing
(Inoue 2001: 181). This was because in terms of cognitive investment, the
rmaji proved very fruitful given the little investment of time and the fact that
one could obtain the maximum efficiency with a few signs. Also in Japan,
today one is witnessing the phenomenon of the so-called furi
romaji which is the habit of transcribing native terms not in kana, but directly
with alphabetic characters. This habit involves, in particular, station names
(Shinjuku, Ikebukuro) and names of magazines and periodicals be it foreign
(Focus, Friday) or native (Nonno, Hanako). Actually, the indiscriminate
increase of rmaji seems to be linked with a more cosmopolitan image sought
for Japanese products as one can note from Japanese logos such as those of
Toyota or Subaru. Nevertheless, the use of rmaji remains problematic, since
the Japanese are not used to including them in their written habits. Moreover,
although the Hepburn system remains the most credible solution for the
transcription of foreign languages, many uncertainties remain. In the end,
Inoue compares the evolution of the Japanese language to that of a
castella cake which is compressed from above by gairaigo and is going to
gradually wear down the body of language. This phenomenon, suggests Inoue,
would represent a form of linguistic suicide which the Japanese seem unable
to cope with. Certainly, states Inoue, the future survival of many languages will
be increasingly linked to their exchange value just like that of currencies on the
stock market. The higher the market value, the better the chances of survival.
Yet, there are sociolinguists who cannot agree with Inoue. According to
Okamoto (2008), importing gairaigo and producing wasei eigo would just
reflect two of the main processes that set in motion whenever two cultures
come in contact: namely, the absorption of linguistic elements from the foreign
culture () and the consecutive change within the indigenous one
(). In fact, once a loanword is imported, it is often dropped in new
linguistic contexts within the target culture. This would allow loanwords to
become closer to speakers diluting the sense of alienation that distinguishes
them in the beginning. Loanwords, suggests Okamoto, reveal the high degree
of lexical dynamism of languages. From a historical point of view, Okamoto
compares the impact of gairaigo on Japanese language to that yielded by
kango, shinkango and rmaji at the time of their introduction in Japan.
FRANCESCO VITUCCI
106
Accordingly, loanwords prove to be an instrument of refined lexical

enrichment that acts as a stimulus to the entire social development. Unlike
Suzuki, who suggests exploiting the potential of garaigo for English language
learning, and Inoue, who considers all loanwords as a sign of English language
imperialism, Okamoto overcomes these positions illustrating how language can
be continuously regenerated through the contact with the outside world while
maintaining, at the same time, its own lexical identity and its own dignity. This
would happen since, once placed in a given socio-cultural system, loanwords
develop a semantic life of their own that is detached from their source languages
contributing to the maintenance of the target society in which they develop.
6. Conclusions
As Jinnouchi suggests (2007), nowadays loanwords are spreading more
quickly due to the necessity of conveying new information into the Japanese
society. As a consequence, in order to understand the contemporary role of
gairaigo, one should analyze the Japanese language as a whole taking into
consideration keigo, youth slang, dialects and the role played by the media
rather than isolating gairaigo from the rest of the linguistic life of the country.
Okamoto (2008) also states that penetrating from technical jargons,
loanwords are often distributed in further areas of Japanese, enriching the
general overview of the language with linguistic expressions which did not exist
before. Moreover, being loanwords adapted to the Japanese phonetics
(contributing to the creation of wasei eigo), they should be considered as a
physiological outcome of Japanese lexical life. According to Okamoto (2008),
terms such as stove,connection, infrastructure,
level up, service spirit, make a
last minute cancellation reveal the great dynamism of the Japanese language
which is able to create terms accessible to the entire society. In fact, a big
mistake would be to consider these terms as English or aimed at the
conversation in that language. As a consequence, Okamoto (2008) recognizes in
these terms an exquisitely Japanese citizenship and use. Given these
premises, it seems unlikely that one can limit the spread of loanwords in Japan
since the use made in advertising, as an example, plays more and more with
puns created, not by chance, with gairaigo. From this perspective, one cannot
help but admit that reflecting the trend of the times, loanwords need to be
analyzed for the role they play within the society (Jinnouchi 2006 and 2012).
Accordingly, the spread and the acceptance of gairaigo should not be seen
solely as a linguistic matter, but rather, as an issue that directly involves the
realm of politics (with language policies), society (which spreads the language
reflecting certain fashions and trends), economy (which increasingly globalizes
foreign goods and services), and finally, marketing (which advertises products
and services, see Figure 3).

107
Figure 3. Examples of gairaigo used in advertising
On the basis of the above considerations, it will be interesting to continue

monitoring the lexical evolution in Japan given the dynamic complexity of its
writing system. Far from considering the use of garaigo as an act of language
suicide or a ploy to strengthen a situation of bilingualism, perhaps it would be
better to focus on the life cycle and the contexts in which gairaigo are dropped
(Suginamiku yakusho 2005). Not surprisingly, Japan has always depended in
writing on foreign contributions and it is thanks to these contributions that it has
been able to develop an extremely complex language from the lexical point of view.
As has been demonstrated through the data from several surveys, even
the direct government intervention through specific language policies at the
beginning of the twenty-first century proved to be unsuccessful since the
Japanese society itself has been able to develop an intrinsic capacity of lexical
absorption refuting any hypothesis of linguistic colonialism (Kotoba to shakai
henshiinkai 2000, Gottlieb 2005). We, scholars, must continue to follow
without prejudice the development of this story going beyond any form of reaction
or political subjection under way in the Country. Of course, the changing relations
with China and the import of gairaigo from languages other than English will
bring additional novelty into the Japanese language and society.
REFERENCES
Gottlieb, Nanette (2005), Language and Society in Japan, Cambridge University Press,
Cambridge.
Hall, Edward T. (1976) Beyond Culture, Anchor Books, New York.
Hashimoto, Gor (ed.) (2003), Shin nihongo no genba, Chokronsha, Tokyo.
Inoue, Fumio (2001), Nihongo wa ikinokoreruka, PHP Shinsho, Tokyo.
108
FRANCESCO VITUCCI
Jinnouchi, Masataka (ed.) (2006), Gairaigo to gendaishakai, Kokuritsuinsatsukyoku, Tokyo.

Jinnouchi, Masataka (2007), Gairaigo no shakaigengogaku Nihongo no gurbaruna
kangaekata, Sekaishissha, Tokyo.
Jinnouchi, Masataka (ed.) (2012), Gairaigo kenky no shinhatten, f, Tokyo.
Kaneya, Toshihiro (2002), Katakana Shingojitsuyou Jiten, Gakken Kenkyjo, Tokyo.
Kno, Rokur et al (1996), Gengogaku Daijiten, Sanseid, Tokyo.
Kotoba to shakai henshiinkai (ed.) (2000), Kotoba to shakai, vol. 4, Sangensha, Tokyo.
Kristof, Nicholas (1995), Japans favorite import from America: English, The New York Times
(February 21st).
Okamoto, Sachiko (2008), Shakaigengogaku, ALC, Tokyo.
Palmer, Leonard T. (1979), Linguistica descrittiva e comparativa, Einaudi, Turin.
Suginamiku yakusho kuchshitsu smuka (ed.) (2005), Gairaigo yakusho kotoba iikaech, Gysei,
Tky.
Suzuki, Yji (2003), Katakanaeigo de kajuaru bairingaru, Seikatsushinsho, Tokyo.
Tanaka, Harumi and Sachiko Tanaka (1996), Shakaigengogaku e no shtai, Minerva Shob, Tokyo.
Weblinks
Monbukagakush: Gairaigo no hyki, http://www.mext.go.jp/b_menu/hakusho/nc/k1991062
8002/k19910628002.html, retrieved on February 2nd, 2013.
Okamoto, Sachiko: Gairaigo no juy to kanri: gengo seisaku no shiten kara, http://libro.dobunkyodai.ac.jp/research/pdf/treatises05/05OKAMOTOa.pdf, retrieved on February 2nd,
2013.
ON THE DIACHRONIC DEVELOPMENT AND PRAGMATIC

FEATURES OF CHINESE PUBLIC SIGNS
JING DENG*
Abstract
This paper investigates the linguistic and pragmatic features of public signs in three
distinct historical stages of the Chinese society. It then focuses on the new pragmatic features of
public signs, in order to shed light on the characteristics of public language and social life in
modern China.
Keywords: social pragmatic analysis, public signs, linguistic and pragmatic features.
1. Introduction
Language is a mirror of society. It reflects a societys structure, its beliefs
and values as well as the transformations it undergoes. Among all varieties of
language public language, or namely, words and phrases frequently used on all
kinds of public occasions, is one of the most direct and obvious linguistic
evidence. It is an open and dynamic system closely related to social
development and peoples daily life, manifesting cultural qualities, moral
attainment and spiritual features of the whole society.
In different historic stages public signs prevail in every corner of cities in
China to provide information, give warnings or advocate specific social norms.
As social slogans directed to the general audience the technique and art of
designing public signs embodies the connotations of Chinese culture and the
progress of the Chinese society. This paper aims to analyze the historic
development of public signs and explore the linguistic features and pragmatic
strategies of public signs in todays China to shed light on characteristics of
public language in China and the social life of the Chinese people.
College of Foreign Languages and Literatures, Fudan University, and School of Foreign
Languages, Nanjing University of Science and Technology, dengjingchina2005@aliyun.com.
JING DENG
110
2. Definition and characteristics of public signs
Websters Third New International Dictionary (2000) defines the public

sign as a lettered board or other public display placed on or before a building,
room, shop or office to advertise business there transacted or the name of
person or firm conducting it. In Longman Dictionary of Contemporary English
(1978), the public sign is defined as a piece of paper, metal, etc. in public place,
with words or drawings on it that gives people information, warns them not to
do something (such as road signs or no-smoking signs). Public signs include
land post, advertising board, shop and slogans in tourist attractions and the other
public places.
From the above definitions of public signs, we can infer some basic
characteristics. First, as a mode of communication, it is a kind of one-way
communication with the general public in which the speaker conveys the
information while the audience receives it without possibility of negotiation.
Second, they tend to be simple and brief in form: people seldom spend much
time reading public signs; hence the designers have to convey their message in
the most direct and prominent way to attract the publics attention. The speaker
always tries to convey the largest amount of information within the limited
space of a small sign. Third, the main communicative purpose of public signs in
essence is to persuade, forbid, or warn the public to enact/prohibit the intended
action of the speaker in the interest of public welfare. With their illocutionary
force of directives, public signs often impose some face threat on the public;
therefore, the speaker often endeavors to alleviate the face threat through
strategic use of language.
3. Diachronic development of public signs in China
As a type of social slogans serving the public, the designing of public
signs is marked by times since they conform to specific national conditions and
social mentality at that time. From the evolution of public signs, one could
witness the political, ethical, and psychological factors involved and the
changes in peoples life. Roughly the development of public signs in China
could be divided into the following three phases in terms of their respective
linguistic and pragmatic features.
3.1. Phase one (before the reform and opening-up policy)
In this period, China pursued the planned economy and political
propaganda was prevalent throughout the country. In the cultural arena
thousands of years of feudalism still had great impact on peoples ideology, for
instance, the notion of social class hierarchy still existed in peoples mind.
ON THE DIACHRONIC DEVELOPMENT AND PRAGMATIC FEATURES

OF CHINESE PUBLIC SIGNS
111
Public signs in this phase generally were mainly didactic by nature to discipline
peoples behavior and maintain the social order. The tone of public signs at that
time was rigid, distant or even threatening with the speaker giving commands
on behalf of the administrative institutions concerned. Words like jnzh
is forbidden, ynjn is strictly forbidden or bx Do not
frequently appeared. For example, in many directive public signs such as
Ynjn jihu kich. Drunk driving is strictly forbidden. and
Bx jint copng, fuz jing chy
fkun! Trampling on the lawn is not allowed. Otherwise you will be fined!,
the tone was serious and authoritative with an unequal power relationship
between the interlocutors. Consequently although the illocutionary forces of
these public signs were asking the public not/to take the intended actions, it was
very possible that the very opposite perlocutionary act would occur because the
public was offended by the cold and blunt tone of the speaker.
3.2. Phase two (from the reform and opening-up program to the early
1990s)
Since the reform and opening policy in 1978, China strengthened its
exchanges with the outside world and gained great momentum in its economic
drive. And the material and cultural life of the Chinese people became richer in
this period. The designers of public signs began to pay attention to the propriety
issue by adopting a more friendly and polite tone. The most typical example is
the frequent use of the politeness marker please and some explanatory
remarks in directive public signs. For example public signs Qng
boch njng. Please keep quiet. and Wile
nn h trn de jinkng, qng byo xyn. For the sake of your and others
health, please do not smoke. These public signs indicate that the speaker
realized the face threat to the public incurred by the tough and overbearing
language of the public signs in the past and began to consider the audiences
emotions and feelings by showing them respect.
3.3. Phase three (the early 1990s - present)
During this period the economy of China continued to develop rapidly
and China enhanced its exchanges with the outside world. Peoples living
standard at that time was further improved and they craved for a more
meaningful social life. Public language in this period became much more
individualized and diversified with the integration of modern culture and
foreign cultures into traditional Chinese culture with the advent of the
information age. Public signs became humorous, diversified, humanized,
marking a more civilized Chinese society. For example, in a shopping center,
JING DENG
112
the label on clothes writes: Bi mw, w p zng. Dont

touch me. I hate dirtiness. This sign persuades the customer not to touch the
new clothes. With the personification in it, it achieves some humorous effect.
The public sign Jirn pnwng nn nqun guli. Your
family look forward to you to come back safely. reminds the audience to drive
carefully by mentioning the family wishes. The honorific pronoun nn you
shows the consideration and respect of the speaker towards the driver coming
back from their work.
From the above analysis, we could conclude that some problems existed
in the design of public signs before the reform and opening up of China. First,
the language of some public signs at that time, such as Ah copng
Take care of the lawn, Jiyu yngshu Save Water and
Boch njng Keep silence were too bland and tasteless to catch the
audiences eye, so it is very likely that the audience overlooked them and thus
they failed to achieve the communicative purpose. Second, some directive
public signs such as Bx chos xngsh Overspeeding is not
allowed. or Jnzh hngchun ml Jaywalking is forbidden.
are compelling and speaker-centered, impinging on the audiences freedom,
therefore posing a serious threat to their negative face.
4. Pragmatic features of public signs in todays China
In the above section, the characteristics of public signs in three distinct
historical stages of China were explored. With the changes happening in todays
Chinese society, public signs exhibit some new linguistic and pragmatic
features. The investigation of public signs was conducted in different cities in
China which witness great changes with the process of urbanization. Public
signs regarding topics of environmental protection, transportation safety and the
like, in locations such as streets, parks, residential complex and universities,
were collected so as to shed light on different walks of social life in China. In
this section, some major new linguistic features of these public signs as well as
the pragmatic strategies behind them will be discussed.
4.1. Politeness
Politeness is a symbol of human civilization present in any culture in the
world. In traditional Chinese culture, the concepts of politeness and rituals were
highly valued throughout its thousands years of history. Moreover, in modern
times, under the influence of democratic ideas from the western culture, people
demand more dignity and equality in communication. Since most public signs
are directives with face threat to the audience, the speaker will employ many
linguistic and pragmatic strategies to reduce the face threat. Besides the

113
pragmatic markers mentioned above like qng please, the speaker also often
resorts to indirect speech acts or a sequence of speech acts to adjust its
illocutionary force.
4.1.1. Indirect speech acts
According to Searle (1975) and Leech (1983) people use indirect speech
acts out of politeness. Brown and Levinson (1987) also hold that indirect speech
act is a strategy of avoiding face-threatening acts. Many public signs today
contain indirect speech acts to increase the degree of politeness. Consider
example (1).
(1)
a.
b.
Gnxi n du huco de ix.

Thank you for taking care of the plants.
T bdi guw sh yzhng shshng.

It is in vogue to use cloth bag when shopping.
In example (1a), an act of thanking is employed to replace the original act

of request. The speaker thanks the audience even before the intended act is
performed so the audience is subject to the performance of the intended act
because of the presupposition embedded in it. Thus it increases the possibility
of the intended act while maintaining the audiences face. Actually this strategic
usage of / gnxi/xixi Thank you has conventionalized in public
sign designing today. Many public signs today end with Xixi hzu
Thank you for your cooperation as in Qng byo
dshng xunhu, xixi hzu Please do not speak loudly. Thank you for your
cooperation. Example (1b) is an assertion that encourages the fashion of using
cloth bags to restrain the pollution caused by plastic bags. The audience could
hardly sense any threat to their negative face because of the indirect speech act
of the assertion.
4.1.2. Extended speech act
Ferrara (1980) holds that in communication people do not always use one
speech act; sometimes a sequence of speech acts may be used to realize the
communicative purpose. Wood and Kroger (1994) point out that a speech act is
generally composed of a central speech act, an auxiliary speech act and a microunit. Among these complicated speech acts one of them is primary and the rest
JING DENG
114
are auxiliary speech acts that help to enhance the acceptability of the intended
act. In many public signs today, there is more than one speech act. Consider the
examples in (2):
(2)
a.
Shu sh shngmn zh yun, qng jiyu yngshu.

Water is the source of life, so please save water.
b.
Tshgun shj sh qunxio shshng gngtng de jngshn cif, qng

byo schng, qiq, wsn!
Books in the library are common spiritual legacies of all faculty and students,
so please do not hide, steal or spoil them!
In the above examples, two speech acts are combined: an assertion and a
request. In example (2a) the first speech act is a statement that emphasizes the
importance of water to human life, while in example (2b) the significance of
books to faculty and students. In these auxiliary speech acts that support the
main speech act (i.e. the request), the speaker provides some surplus
information which violates the maxim of quantity assuming that we are concise,
brief and to the point in communication (Grice 1975). In this way the speaker
highlights the importance of water and books and reinforces the illocutionary
force of the main speech act.
4.2. Bonding
In a typical eastern culture like China, collectivism and closeness among
people is very important. In public sign designing, the speaker often tries to make
the audience feel warm and tender by underscoring the bonding between the
interlocutors, thus enacting the intended behavior unconsciously. One of the most
frequent devices of showing bonding is the unconventional usage of personal deixis
or shift of personal deixis, which emphasizes the solidarity between the
interlocutors by vitalizing the empathetic effect, as in the examples in (3):
(3)
a.
b.
Yng wmn de yfn nl, hunli chngshde ypin lntin.

With effort from each of us, we could have a blue sky in our city.
W sh Bijng rn, w zu hunbo sh.

Im from Peking, and I will do what I can to protect the environment.

115
According to Levinson (2001) deixis is organized in an egocentric way,

that is, the speaker is the central person, but in the some derivative usages of
personal deixis the deictic center is shifted to other participants. In example (3a)
the sentence is organized from the perspective of the audience (the inclusive
pronoun we) as if the speaker himself were one of them so as to shorten the
psychological distance between him and the audience. Hence the audience
would find it easier to accept and perform the intended act. Example (3b) dates
from the 2008 Beijing Olympic Games. The speaker expresses the social
identity of being a citizen of Peking, who is supposed to be more cosmopolitan
and broadminded than those in other regions, since capital Beijing is the
political, economic and cultural center of the country. The I-perspective of the
whole sentence is achieved by the use of w I as the subject, activating a sense
of belonging of the audience by deeming the speaker as a member of the same
group as the audience.
4.3. Sophistication
Although the primary purpose of public sign is practical by nature, the
designers of public signs nowadays attach more attention to the aesthetic value
of public signs themselves. The language of the public signs also demonstrates
the unique artistic taste and distinctive personality of the speaker. Thus the
speaker would adapt his style to the tastes of the audience to satisfy peoples
aspiration for a loftier spiritual life. To this end, various rhetoric devices are adopted
to increase the readability of the text within the limited words of public signs.
4.3.1. Quotation
Traditional and modern Chinese culture is a continuous source in the
designing of public signs. Many public signs in todays China are direct or
indirect quotations from Chinese literary works, which embodies a rejuvenation
of traditional Chinese culture to some extent. For example for many public
signs on college campus traditional Chinese cultural elements such as the
Chinese ancient poems or lyrics are often involved to form an elegant style
tailored to the audience with higher education. Consider the examples in (4).
(4)
a.
b.
Qngqng de w zu le/Zhngr w qngqng de li

Very quietly I take my leave/As quietly as I came here
Shi zh pn zhng cn/ll ji xnk

Look at the food on our plate/Every grain of which is from hard work.
JING DENG
116
Example (4a) is from the university library and tells the readers to keep
quiet while they study in such public places. These are the first two lines of the
modern Chinese poem Saying Goodbye to Cambridge Again by the modern
Chinese poet Xu Zhimo, familiar to and welcomed by most college students. It
creates a lifelike image in the audiences mind and gives them a sense of
beauty. It caters to the audience of young college students and contributes to
build a civilized atmosphere on the whole campus. Example (4b) is found in
student canteens of many universities as a reminder for college students not to
waste food. These are the original lines of the noted classical ancient Chinese
poem Chuhe [= Toiling Farmers] by Li Shen of the Tang Dynasty. With the
antithesis of balanced structure and symmetric rhythm it impresses the audience
and conveys the maximum of information in an economical way. In addition,
the striking size and regular font of the words on the red slogan reinforce the
cautioning effects on the audience.
4.3.2. Personification
One of the other prevalent rhetoric devices adopted in public signs in todays
China is personification, which gives personal attributes to inanimate objects which
makes public signs more vivid and lifelike, as in the examples under (5):
(5)
a.
b.
Xioco zhng jnr tintin de mimng, qng bi jngxng t de homng

The grass is in a sweet dream, so please do not disturb.
Go ti gushu, qng byo gi kzhu wnshn

Please spare your hands and do not tattoo the desk.
In example (5a), the grass is referred to as being an animate object with

human sensations and emotions. By personalizing the grass as a helpless being
in pain if we step on it, the speaker arouses the compassion of the audience to
protect the weak. Example (5b) is found in universities where scribbling and

117
carving on desks is commonplace. Here the desks are personalized as human

beings suffering from the pain of tattoos on the body, so the audience might be
aware of the inappropriateness of their behavior.
4.3.2. Metaphor
Another frequently used rhetorical device is the metaphor, which turns
the abstract into concrete and the bald into interesting. It enhances the
readability of the text by introducing images that trigger the audiences
imagination, as in (6):
(6)
a.
b.
Xngf sh k sh/Anqun sh wt
Happiness is like a tree; safety is the fertile soil.
Co sh shji de dtn/Sh sh dqi de jngmi

Grass is the worlds carpet and the tree is the earths veins.
In (6a), happiness is compared to a tree while safety to the fertile soil in

which the tree grows. By resorting to metaphor, the speaker vividly depicts an
image of a green tree and the soil in the minds of the audience, thus reminding
the audience of the inseparability of happiness and safety safety is the
foundation of happiness. Example (6b) compares grass to a carpet and the tree
to veins, which portrays a picture of plants on the earth in the audiences mind
and stimulates their imagination of a beautiful world.
4.4. Humor
Humor is also an important feature of public signs today. One of the
principles of designing public signs is to attract their attention and stimulate the
public interest. By humor the speaker could convey their intentions in an
implicit and tactful way which brings them amusement and makes an
impression. Particularly in our modern society people live in more and more
cramped spaces and undertake great pressure. A humorous public sign could
release the pressure in their life and is conducive to establishing good
interpersonal relations. One way for the speaker to achieve a humorous and
novel effect in public signs is by deliberately violating the maxims of
Cooperative Principle, as in (7):
(7)
a.
Bizhu le, bnrn yhn.
JING DENG
118
b.
Dont chase me, Im married.
Lko wc b jg
Five failures in road test.
Today in many big Chinese cities, transportation problems are increasingly

severe with more and more automobiles on the road. People are prone to get
tense when stuck in traffic jams. The above two examples are signs on the rear
of cars reminding drivers behind to keep the distance. Example (7a) involves a
pun: the Chinese character zhu has a double meaning: chase literally and
court (a girl) metaphorically. The existence of the two meanings of zhu
violates the Maxim of Manner, that is, avoid ambiguity. Example (7b) contains
a hyperbole exaggerating the drivers defects in the road test, which violates the
Maxim of Quality, namely, do not say what you believe to be false. By
deliberately violating the Cooperative Principle, these entertaining and creative
signs successfully convey the implicature of not overtaking the vehicle ahead.
The cartoon-shaped words or even the image of two pet cats contribute to create
the lighthearted and humorous effects.
5. Conclusions
Public signs as a kind of social managerial language are omnipotent in
peoples daily life to persuade, warn, advocate, or even entertain and enlighten
people. In China, public signs present different linguistic and pragmatic traits in
different historical periods. The language of public signs is a window to
understand the beliefs and values, cultural and historical traditions, and the
aesthetic taste of the entire Chinese society. This paper discussed the linguistic
and pragmatic features of public signs in different historical stages of Chinese
society. The main idea is that the public language, such as reflected through
signs, changes in order to keep up with the shifts in democracy, in literacy, and
in the standard of living.

119
REFERENCES
Brown, Penelope and Stephen C. Levinson (1987), Politeness: Some Universals in Language
Usage, Cambridge University Press, Cambridge.
Ferrara, Alessandro (1980), An extended theory of speech acts, in Journal of Pragmatics, 4, 3,
pp. 233-252.
Grice, Paul (1975), Logic and conversation, in P. Cole and J. L. Morgan (eds.), Syntax and
semantics, vol. 3, Speech Acts, Academic Press, New York, pp. 41-58.
Leech, Geoffrey (1983), Principles of Pragmatics, Longman, London.
Levinson, Stephen C. (2001), Pragmatics, Foreign Language Teaching and Research Press, Beijing.
Websters Third New International Dictionary (2000), Merriam-Webster, Springfield.
Longman Dictionary of Contemporary English (1978), Longman, Harlow.
Searle, John (1975), Indirect speech acts, in P. Cole and J. L. Morgan (eds.), Syntax and
semantics, vol. 3, Speech Acts, Academic Press, New York, pp. 59-82.
Wood, Linda A. and Rolf O. Kroger (1994), The analysis of facework in discourse: Review and
proposal, in Journal of Language and Social Psychology, 13, 3, pp. 248-277.
LINGUISTIC LANDSCAPE, MICROTOPONYMY AND

UNCONVENTIONAL USE OF ANTHROPONOMY
ON THE BORDER: VAMA VECHE, ROMANIA
ALINA BUGHEIU*
Abstract
The village of Vama Veche (meaning the old customs) is located on the coast of the
Black Sea in southeast Romania, on the border with Bulgaria. It is a small fishermens settlement
that in summer turns into one of the most important seaside resorts in the country. In the context
of geographical and social borders, the present paper proposes a sociolinguistic approach to local
microtoponyms, starting with the identification and analysis of the grammatical, lexical and
semantic structure of the names. The aim is to delineate, from the perspective of commercial
onomastics (especially names of pubs and accommodation locations), the sociocultural profile of
the tourists that visit Vama Veche. The linguistic material analysed consists of approximately two
hundred names, collected by the author of this paper mostly by means of field research in the
area. In the ever-globalising public space of the resort, the present paper highlights the way in
which microtoponyms have become a key means of asserting group identity.
Keywords: commercial names, microtoponyms, linguistic landscape, sociolinguistics,
onomastic behaviour.
1. Introduction
The present paper1 analyses commercial names in Vama Veche, currently
one of the most important Romanian seaside resorts, from a sociolinguistic
perspective. My study aims at delineating and describing the onomastic trends
that are promoted in the contemporary public space of the settlement and
performing a commercial onomastic characterisation of the linguistic landscape
(LL) of Vama Veche, by looking at the types of consumers that are targeted by
the name choices. Thus, by considering the trade names in Vama Veche in
relation to the tourists that visit the resort (regularly or occasionally), two main
*
North University Centre of Baia Mare, Technical University of Cluj-Napoca, Romania,

alina.bughesiu@gmail.com.
1
This study pertains to a broader-scope research on Unconventional Romanian

Anthroponyms in European Context: Formation Patterns and Discursive Function, a
project funded by CNCS, code PN-II-RU-TE-2011-3-0007, contract number 103/2011
(project manager: Daiana Felecan).
ALINA BUGHEIU
122
onomastic behaviours can be distinguished; they refer to the communicative

relationship that business owners try to establish either with a general
consumership (the commercial naming patterns are not specific to the resort in
question, as they can be found in other public spaces or touristic sites) or a
specific group of visitors (i.e. the vamaioi, the supporters of the values that
were initially attached to the place named Vama Veche: freedom, love,
individuality, purity and untarnished natural beauty).
My approach to the microtoponyms of Vama Veche starts by taking into
account the particular geographical and sociocultural configuration of the
locality, in the attempt to highlight its effect on the sociolinguistic development
of the settlement and commercial onomastic behaviour implicitly. Based on the
structural (i.e. grammatical, lexical and semantic) analysis of trade names in the
contemporary public space of Vama Veche, my paper underlines that
commercial names function as key means of asserting group identity (in
connection with the aforementioned vamaioi).
Methodologically, this research rests on the theoretical framework of
onomastics, sociolinguistics and linguistic landscape theory, as well as on
postcolonial studies (in the description of Vama Veche as a border space,
literally and symbolically). The corpus investigated consists of nearly 200
names of eating or drinking houses and accommodation locations, most of
which were collected by means of field research by the author of the paper in
the summer of 2011. In order to note the evolution of the two above-mentioned
types of naming trends (targeting a general or a specific touristic audience),
some specialised websites were also consulted: Booking.com, Plaja.ro,
Pubbing.ro, apte Seri, Turist Info.ro and viaRomania.
2. Identifying and defining the landscape in linguistic landscape
Vama Veche is one of the four hamlets of Limanu, a commune in
Constana county, in southeastern Romania. It is located on the coast of the
Black Sea, being the last settlement before the shoreline border with Bulgaria.
Vama Veche was founded at the beginning of the 19th century by some families
of Gagauz migrants (Romanian gguzi, a Turkic group of Orthodox faith that
allegedly resulted from the mixture of several nomadic tribes from the Eurasian
steppes)2. The hamlet was originally known as Yilanlk and later as Ilanklc, a
2
These settlers came from the Gagauz communities in Bessarabia and the
southern extremity of the Republic of Moldova, after the two former Romanian
territories were conquered by the Russian Empire. The number of Gagauz ethnics
severely decreased over the years, to such an extent that the results of the census of
2011 do not make any particular reference to the Gagauz minority, implying, however,
that there exist either less than three members of this community or that they are extinct
(Recensmntul populaiei i al locuinelor 2011).
LINGUISTIC LANDSCAPE, MICROTOPONYMY AND UNCONVENTIONAL USE OF

ANTHROPONOMY ON THE BORDER: VAMA VECHE, ROMANIA
123
name coined by the Gagauz settlers, meaning a place full of snakes (the
toponym was related to the faunal peculiarity of the region) (Primria comunei
Limanu, Judeul Constana).
Under the communist rule in Romania, Vama Veche was not visited by
tourists mainly because of its geographical position. However, for the very
same reason, starting with the 1990s, this poor fishermens hamlet with its
rather obscure beach became a favourite haven especially for nudists. Along
with the end of the 1990s and in the early 2000s, the village started being
promoted by a local rock band, eponymously called Vama Veche, who depicted
the settlement in the lyrics of their songs as an idyllic place, praising its pristine
landscape and the almost utopian freedom that it seemed to offer. Thus, tourists
began to pour in, eager to discover a site like none other on the Romanian coast
and, basically, like nowhere in Romania. A rock music festival was established
in 2003, called Stufstock (a portmanteau word coined from the Romanian neuter
appellative stuf reed + the second component of the English proper name
Woodstock), in line with the hippie atmosphere that pervaded the settlement
every year in summer. Three years later, another music festival was set up,
suggestively and symbolically named Folk You! (one of the logos of the event
was the profane middle finger whose nail resembled the headstock of a guitar).
Whereas Stufstock kept its name and tradition (the only significant change being
the recent focus on Romanian bands), Folk You! was subsumed under Micarea
de rezisten (the resistance movement), whose broad aim was to fight against
the gross behaviour and lack of education promoted by most of the popular
culture in contemporary Romanian public space.3 Therefore, in an almost
overnight experience, Vama Veche became a fully-fledged seaside resort, and
year after year its Edenic aura keeps fading, as a result of tourist services
providers attempt to satisfy more diverse visitors. October to May, however,
Vama Veche goes back to being a poor hamlet, a derelict and almost abandoned
place, with most of the inhabitants past their middle age, commuting to nearby
towns and bigger villages to work or study.
By the time Micarea de rezisten appeared, the tourists of Vama Veche
were already divided into two main subgroups. Perhaps it would be more
accurate to state that Micarea de rezisten occurred because of the two
communities of tourists that could be noticed to visit the resort, as a reaction of
one of the communities (the original one) to the consolidation of the others
existence (the newcomers):
(1) The vamaioi are the people that love Vama Veche for what it used to
be: a place where one could listen to jazz, rock or folk, sleep on the beach and
3
The frequently used term is manelizare, derived from the appellative manele, a
music genre that mixes several influences especially Turkish, Arabic, Serbian,
traditional Romanian folk music and traditional Gypsy music , recurrently dealing with
topics such as money, reciprocated or unrequited love, family and enemies, and
generally associated with Romanian mass culture.
ALINA BUGHEIU
124
enjoy the untarnished scenery. The term is a slang appellative derived from the
first component of the toponym Vama Veche with the addition of the suffix iot,
of Greek origin (also found in appellatives like fanariot Phanariot or cipriot
Cypriot, the latter recorded as borrowed from French), which is used to form
names of inhabitants. The suffix is fairly productive in colloquial Romanian
and, according to Zafiu (2000), it often evokes pejorative or humorous
connotations. This explanation tallies with the image that is triggered by the
employment of the term vamaiot by the mass media: to the outsiders, the
prototypical Vama Veche tourist is (euphemistically speaking) an unreliable
young person, whose sole aim in life is to lead a bohemian existence, with
disregard for any kind of order. Nonetheless, the online dictionary of urban
Romanian language, 123urban.ro, records another definition of the term in
question, highlighting in a jocose, manifesto-like note the positive qualities of
the prototypical Vama Veche tourist and aiming at distinguishing him/her from
the prototypical mass tourist: 1. Someone who goes to the seaside, to Vama
Veche on a regular, yearly basis. A vamaiot loves: the beach, the sea, a beer,
more beers, fried fish, a bed sheet laid down on the beach, tents, guitars, the
blues, Expirat [the name of a well-known pub, literally meaning stale, out of
date]. A vamaiot does not love: bathing suits, Jeeps, chaise longues, fads, raids,
buoys, Jet Skis, forbidden camping, leased beaches, plastic (123urban.ro, s.v.
vamaiot, orig. Romanian, my translation). The last part of the definition
identifies, by means of negation, the Other tourist, with whom the vamaiot
wishes not be mistaken. All the items or aspects listed in the latter section of the
definition are seen by the vamaiot as being the negative effects of globalisation,
perceived as a neutralising phenomenon that leads to the amputation of individuality.
(2) The mainstream tourist is a person that is more rule-bound and
follows a set of widely acknowledged and promoted standards in his/her pursuit
of a holiday experience. These expectations demand that the natural landscape
and the public space in a certain place be altered so that they can fulfil the
tourists needs. With a place like Vama Veche, alterations of this kind are
considered detrimental by the vamaioi, who believe that their (natural, social
and cultural) place of refuge should escape the influence of globalisation
unscathed. This does not mean that the mainstream tourist is the Wicked Witch
of the West in Romanian contemporary tourism. It is simply that, as a
representative and promoter of mass culture, s(he) stands for the very thing that
the vamaioi zestfully reject: swimming in the direction of the rivers flow.
3. Commercial names as markers of group identity
This sociocultural configuration of Vama Veche is also visible in the
linguistic landscape (LL) of the resort. The commercial signage of the hamlet is
suggestive of the two communities of tourists that pertain to Vama Veche. The
trade names that define the commercial onomastic dimension of the public

125
space are indicative in this respect on the level of (a) non-verbal configuration
(mostly visual cues, i.e. typeface and chromatic properties, and so on, but aural
cues can also occur: for instance, the music played in some pubs or bars can
enable one to associate that business entity with a certain referential categorial
subtype) and (b) lexical-grammatical and semantic construction. While the nonverbal elements may be used to capture the attention of prospective customers
and help them to perform the first step in the identification of a given
commercial entity in relation to either of the aforementioned touristic groups,
the name content can confirm or invalidate the categorial presupposition and
even trigger new associative and emotive meanings; moreover, the latter kind of
meanings, established by the non-verbal dimension of a business entity, may
sometimes be reinforced or even contradicted by the associations conveyed via
its trade name.
The LL in Vama Veche is determined by the business owners wish to
meet the tourists demands, trying to keep up to date with the newer wave of
tourists, while also preserving the loyal consumership. According to Kallen
(2009: 275), there are four types of tourists needs that can influence the
shaping of the local LL: (1) the need for an authentic experience of place, to
see the real foreign land; (2) the need to feel secure, ensuring that what is
different is not so different as to be threatening or in some way repugnant; (3)
the need to break away from normal routines; and (4) the need to return from a
journey of transformation, i.e. to create a memory of the experience of travel
that stands out from other experiences. While the third and fourth needs are
prominent as regards the vamaioi, it is the first and second needs that are
manifested by the mainstream tourists.
In the LL of Vama Veche the discourses that correspond to the two
touristic communities are often juxtaposed or even overlapped (Kallen 2009:
274). The latter situation can be found with business establishments that adjust
their marketing strategies in order to satisfy a wider range of tourists, albeit
their names are still indicative of the community of the vamaioi: e.g. Elgas
Punk Rock Hotel (the first component is an acronym containing the truncations
of the owners first names, Ella and Gabi), Expirat (Rom. adj.4 expirat stale,
out of date, obtained from the past participle form of vb. a expira to expire),
La Canapele (Rom. prepositional noun phrase < prep. la at + pl. c.n. canapele
sofas, indicating the type of furniture that can be found in the pub), La Galerie
(Rom. prepositional noun phrase < prep. la at + c.n., f., galerie gallery,
referring to art galleries), and so on. Thus, one and the same name can
4
The following abbreviations were used in the lexical-grammatical analysis of

the names: adj. = adjective; c.n. = common noun; En. = English; f. = feminine; It. =
Italian; Lat.= Latin; m. = masculine; pl. = plural; prep. = preposition; Rom. =
Romanian; Sp. = Spanish; vb. = verb.
126
ALINA BUGHEIU
frequently have a polyphonic functionality (Felecan 2011: 9), as it may

simultaneously be directed at both types of tourists (Kallen 2009: 274). It is the
sum of these discourses that makes up the representation of the country or
region (Kallen 2009: 274).
The microtoponyms that one can come across in the LL of Vama Veche
follow two naming orientations, corresponding to the two touristic groups that
they are aimed at. Therefore, from this point of view, one can talk about marked
names (which, through the associations they trigger, define the community of
the typical Vama Veche tourists) or unmarked names (which are neutral on the
semantic level; although commercially-minded, they follow a well-established
pattern in the field of touristic onomastics). I will present them in the reverse
chronological order of their establishment in the public space of the resort (i.e.
starting with the unmarked names), as in this way the contrast will be more salient:
(1) Unmarked names, the latest group of commercial names to become
coherent in the LL of Vama Veche, comprise especially names of
accommodation locations, as well as some names of shops or bars. Simple
names are scarce: Lyana (the name of a restaurant and shop, derived from the
female first name Liana, a variation of the Ileana and Elena, anglicised by the
spelling of -i- as -y-), Primavera (It. or Sp. c.n., f., primavera spring), Salsa (a
pub whose name is coined from the name of a Latin-American type of music
and dance), Scoica (Rom. c.n., f., scoic seashell, with the enclitic form of the
definite article a).
Most of the constructions are compound, and they generally develop the
pattern of what Soames (2002: 88) calls partially descriptive names, which
consist of a categorial component (denoting the basic level category of referents
to which a name bearer pertains) (see also Van Langendonck 2007: 6) and a
noncategorial component (which can be proprial anthroponymic or
hodonymic or nonproprial appellative or phrasal). The role of the expressed
categorial belonging is first and foremost practical: the widely acknowledged
and even internationally recognisable asserted lexical meanings of categorial
markers like hotel (the most easily spotted categorial identifier by non-speakers
of Romanian), casa de vacan (Rom. compound appellative, f., cas de
vacan holiday house, used with the enclitic form of the definite article a,
often occurring with the short form casa), pensiunea (Rom. c.n., f., pensiune
boarding house, with the enclitic form of the definite article a), and vila
(Rom. c.n., f., vil villa, with the enclitic form of the definite article a)
facilitate a tourists choice of a favourable location, costs- and amenities-wise.
In the name phrase, the categorial marker comes first, following the pattern of
Romanian partially descriptive names. The noncategorial component
(occupying the final position in the onymic expression) can be derived from
anthroponyms, especially first names (full forms or hypocoristics). The majority
of these first names are feminine and Romanian: Casa Ana, Casa Ana Maria,

127
Casa Gabriela, Casa Nicoleta, Casa Veronica, Casa de vacan Violeta,

Pensiunea Ana Emilia, Pensiunea Anita (of Hungarian origin), Pensiunea
Lucia, Pensiunea Mary (an anglicised variant of the owners first name, Maria,
or of the corresponding hypocoristic Mari), Pensiunea Mary Elena, Pensiunea
Ramona, Vila Anda (a hypocoristic form from Andra or Alexandra), Vila
Cosmina, Vila Flori (a hypocoristic of Florica or Florina), Vila Mady (an
anglicised hypocoristic form of Mdlina), Vila Simina. Occasionally, female
first names or hypocoristics can be preceded by forms of address: Casa
Domnia Ralu (Romanian c.n., f., domni young lady, a poetic form of
address; as an archaism, the term was used to designate princesses or rulers
daughters), Vila Miss Alina (En. c.n. miss). Some commercial names also
include male forenames: Casa AlexAndre (from the male first name Alexandru,
made to resemble the French cognate Alexandre), Casa Cornel, Casa Rzvan,
Casa Tudor, Casa Vlad, Vila Ctlin, Vila Toader, Vila Yanis (a Greek male
first name that frequently occurs in the contemporary Romanian onomasticon as
a result of the numerous Romanian temporary emigrants that went to Greece
and Cyprus to work). Sometimes, names of accommodation locations include
neuter hypocoristics, in that they can be derived from both male and female
anthroponyms (e.g. Vila Alex, in which Alex may be a truncation from the
Romanian male first name Alexandru or its female equivalent Alexandra).
Similarly, more opacity is also obtained when owners first names appear as
initials: Casa Double M. The owners name can be joined by the and symbol,
&: Vila Aty & Michelle (Aty, a hypocoristic form of the Hungarian male first
name Attila + the English or French female first name Michelle, instead of the
owners Romanian first name Mihaela). At the same time, the first name can
often be preceded by the Romanian preposition la at: Hotel La John (the
English male first name John, instead of the owners Romanian name Ion),
Pensiunea La Mihi (a hypocoristic form of the Romanian male forename
Mihai, the counterpart of the English Michael), Vila La Mariana (Romanian
female first name Mariana) Vila La Maricica (Maricica, a hypocoristic form of
the Romanian female first name Maria). Some names of accommodation
locations can also comprise surnames: Casa Coman, Casa Stanciu, Vila Iancu,
Vila Mari, Vila Neagu, Vila Prelipceanu. In some cases, full anthroponymic
designations occur, with a hypocoristic instead of the official first name: Casa
Nicu Becleanu (Nicu < Romanian male first name Nicolae).
The noncategorial component may be derived from hodonyms: Casa
Pescru (< Strada Albatros, but there was already a microtoponym derived
from this hodonym), Casa Nikita (< Strada Nichita Stnescu), Vila Albatros (<
Strada Albatros albatross street), or cultural names related to films (e.g.
Complex Sunset Beach < Romanian appellative, neuter, complex complex,
from complex turistic touristic complex + Sunset Beach, the name of an
American soap opera that was aired in Romania in the late 1990s; the typeface
128
ALINA BUGHEIU
in which the name of the hotel is written imitates the one of the name of the
soap opera) or mythology (e.g. Vila Polux). Displaying a semantic status that is
in between proper names and appellatives (as a result of antonomasia),
noncategorial elements in accommodation locations may refer to types of
alcoholic drinks: Hotel Bourbon (Bourbon, a type of American whiskey; the
owners even used the corresponding Bourbon typeface in writing the hotel
name; see Fig. 1), Vila Martini (Martini, a type of Italian vermouth).
Fig. 1: Hotel Bourbon, Vama Veche (Booking.com, Hotel Bourbon)
Compound names containing categorial markers may also comprise

Romanian appellatives that are semantically related to flora. The appellative
may post-modify the categorial marker: Casa de Flori (< Rom. c.n., f., cas
house, with the enclitic form of the definite article a + Rom. prep. de of,
referring to content + Rom. c.n., f., pl., flori flowers), Casa Florilor (< Rom.
c.n., f., cas house, with the enclitic form of the definite article a + Rom.
c.n., f., pl., flori flowers, in the genitive case florilor of flowers). Some
names relate to the sea world or, associatively, to characteristics of the summer
seaside climate: Casa Sunshine (Rom. c.n., f., cas house, with the enclitic
form of the definite article a + En. c.n. sunshine), Hotel Golden Sea (Rom.
c.n., neuter, hotel hotel + En. noun phrase golden sea < En. adj. golden + En.
c.n. sea), Hotel Laguna (Rom. c.n., neuter, hotel hotel + Rom. c.n., f., lagun
lagoon, with the enclitic form of the definite article a), Hostel Sea Star (En.
c.n. hostel, adapted into Romanian following the model of hotel + En. c.n. sea
star), Vila Laguna Paradis (Rom. c.n., f., vil villa, with the enclitic form of
the definite article a + Rom. c.n., f., lagun lagoon, with the enclitic form of
the definite article a + Rom. c.n., neuter, paradis paradise). Some
appellatives are meant to trigger associative meanings related to prestige, often
underpinned by the language choice itself: Hotel Amphora (Rom. c.n., neuter,
hotel hotel + Lat. or En. c.n. amphora), Hotel Victory (Rom. c.n., neuter, hotel
hotel + En. c.n. victory), Pensiunea Opera (Rom. c.n., f., pensiune boarding
house, with the enclitic form of the definite article a + Rom. c.n., f., oper
opera, with the enclitic form of the definite article a). The appellatives or

129
appellative noun phrases may also suggest the high standards of the touristic
experience: Hotel Jakuzzi (Rom. c.n., neuter, hotel hotel + *jakuzzi, perhaps
from the French borrowing jacuzzi, although it might also come from En.
Jacuzzi), Pura Vida Beach Bar & Hostel (Sp. adjectival noun phrase pura vida
pure life + En. compound, specialised categorial marker beach bar & hostel,
indicative of the development and increase of basic level categories in the field
of commerce; while the phrase pura vida could be considered highly suggestive
of the type of experience that Vama Veche is supposed to offer its tourists, it is
rendered neutral by the numerous accommodation locations throughout
Romania that bear this very name). The quality of the touristic experience is
also implied by the only commercial name that comprises an adjective
alongside the categorial marker: Casa Rustic (Rom. c.n., f., cas house, with
the enclitic form of the definite article -a + Rom. adj. rustic, male form; the lack
of gender agreement between the categorial term and the characterising
adjective shows that the latter is not treated as an adjective, but functions as
proper names usually do in partially descriptive names of this kind). There are
many rustic accommodation locations in Romania, as a result of the
development of agritourism.
Some microtoponyms do not include categorial markers but are
indicative of the type of commercial establishment (e.g. Quick Food, coined
after En. fast food), or are semantically related to the seaside landscape: Sea
Temple. Prepositional phrases can also be found, especially with the Romanian
preposition la at attached to other proper names: La Dinamo (Rom. prep. la
at + Dinamo, the name of a Romanian football club).
(2) Marked names make up the oldest coherent group of names in the
LL of Vama Veche, consisting of microtoponyms that refer to the community
of the vamaioi. On the level of lexical and grammatical constructions,
commercial names that can be included in this onomastic orientation may be
simple structures, derived from anthroponyms that display a significant cultural
and historical weight (e.g. Zapata, the surname of Emiliano Zapata, a key figure
in the 1910 Mexican Revolution), resonating with the ideals of the community
of the vamaioi, especially with what they see themselves to be in the
contemporary Romanian public space. Appellatives that are used as commercial
names may trigger associative meanings originally related to
- places or establishments: e.g. Cherhana (< Rom. c.n., f., cherhana
fishery, of Turkish origin), Gulag (< Rom. c.n., neuter, gulag, a Russian
borrowing), Shire (< En. c.n. shire, written in Blackletter typeface, also known
as gothic script; see Fig. 2), Taverna (< Rom. c.n., f., tavern tavern, with the
enclitic form of the definite article -a);
ALINA BUGHEIU
130
Fig. 2: Shire, Vama Veche (Pubbing.ro, Shire)
- mythical characters: Goblin (< goblin);

- people: Corsaru (< Rom. c.n., m., corsar corsair, with the enclitic
form of the definite article u(l)); Desperados (< En. c.n. desperado, probably
under the influence of the 1995 action film directed by Robert Rodriquez and
starring Antonio Banderas);
- parts of the human body: Hand (< En. c.n. hand, perhaps a truncation of
handmade, meant to be suggestive of the many arts and crafts workshops that
were often held there; this was one of the most important cultural and artistic
venues in the resort);
- drugs: Shrooms (< shrooms, one of the colloquial terms used to refer to
psychedelic mushrooms, or magic mushrooms);
- weapons: Molotov (< Molotov cocktail or cocktail Molotov, with the
word order that the phrase records in Romanian, figuratively used to refer to
any deadly combination);
- religious objects (Rom. c.n., neuter, crucifix crucifix; see Fig. 3).
Fig. 3: Crucifix, Vama Veche (Photo: Alina Bugheiu, 2011)
Commercial names in Vama Veche may also be derived from simple

adjectives: Expirat (Rom. adj. expirat stale, out of date, expired). In this case,

131
as in several others that will be shown further on, the associative and emotive
meanings triggered by the name form and/or content are not crucial in
connecting the commercial name to the community of the vamaioi; however,
this link is made based on the associative and emotive meanings established in
relation to the name bearer, which is an old, well-known bar in the resort.
Some commercial names can be derived from acronyms, which constitute
the basis of cultural names: e.g. M*A*S*H (< M*A*S*H, a famous American
television series that was aired in Romania in the 1990s and early 2000s; the
actual establishment is a big old army tent, see Fig. 4).
Fig. 4: M*A*S*H, Vama Veche (Pubbing.ro, M*A*S*H)
Compound commercial names that can be considered indicative of the

community of the vamaioi may be derived from bynames of cultural figures: El
Comandante (Sp., referring to Che Guevara, a major figure in the Cuban
Revolution, who was promoted by Fidel Castro to the rank of Comandante).
They may contain categorial identifiers, but these establishments are by far
fewer than in the case of unmarked names and mainstream tourists respectively.
Some are attached to anthroponyms, which appear by themselves (Bungalow
Peters Ana < surname + first name) or preceded by forms of address (Camping
Nea Ion < Rom. c.n. camping, an English borrowing, designating a camping site
+ Rom. nea, obtained by means of truncation from the colloquial form of
address nenea mister, used to designate especially elderly men) and
prepositions (also conjoined by the coordinating conjunction: Camping La
Misha i Bogdan < Rom. camping + Rom. prep. la at + Misha < Mia, from
Miu, a hypocoristic of the Romanian male first name Mihai + Rom.
coordinating conjunction i and + Rom. male first name Bogdan). Although
rarely so, an anthroponymic structure may also precede the categorial identifier,
often a compound construction, characterising the basic level category to which
the name bearer pertains: Elgas Punk Rock Hotel (< Elga, an acronym based
on the truncations of the owners first names, Ella and Gabi, with the synthetic
genitive s + compound, specialised categorial term punk rock hotel).
132
ALINA BUGHEIU
Anthroponyms may occur with appellatives that are suggestive of the basic
level category of the named entity, without mentioning it clearly: Papa la oni
(Rom. c.n., f., papa food, baby-talk + Rom. prep. la at + oni, according to
the approximate pronunciation of the male hypocoristic Sanyi, from the
Hungarian male first name Sndor). Similarly, in some instances, the categorial
identifier can be missing altogether: La Barb Neagr (Rom. prep. la at +
Barb Neagr, the Romanian cognate of Blackbeard, a famous early 18th
century pirate), La Mexicanu (Rom. prep. la at + Mexicanu the Mexican,
the owners nickname, which may be related to the type of food served, derived
from the Romanian appellative and corresponding adjective mexican
Mexican, with the apostrophe marking the colloquial pronunciation of the
word, without the ending consonant sound of the enclitic form of the male
definite article u(l)); La Theo Vamaiotu (Rom. prep. la at + Theo, a
hypocoristic form of the male first name Teodor, with the variant Theodor +
Vamaiotu the vamaiot, here functioning as a byname). In other cases,
categorial terms co-occur with appellatives or appellative constructions: Bazart
Hotel (< a portmanteau word obtained from the merger of Rom. c.n., neuter,
bazar + Rom. c.n., f., art), Casa Dor de Vam (Rom. c.n., f., cas house,
with the enclitic form of the definite article a + Rom. c.n., neuter, dor
longing + Rom. prep. de for, indicating the object of longing + Vam, the
short form of the toponym Vama Veche), Casa La Meduza (Rom. c.n., f., cas
house, with the enclitic form of the definite article a + Rom. prep. la at + Rom.
c.n., f., meduz jellyfish, with the enclitic form of the definite article a), Terasa
Povestea Ceaunului (Rom. c.n., f., teras patio, garden, with the enclitic form
of the definite article a + Rom. c.n., f., poveste story, with the enclitic form
of the definite article a + Rom. c.n., neuter, ceaun cauldron, in the genitive
case ceaunului of the cauldron), Vila No Name (Rom. c.n., f., vil villa, with
the enclitic form of the definite article a + En. determiner no + En. c.n. name).
There are trade names in the LL of Vama Veche that contain English
appellative noun phrases with a proclitic definite article: The Jack, The Shot
(referring to a shot of tequila), The Stage. Some appellative noun phrases may
be post-modified by prepositional phrases: Csua cu minuni (Rom. c.n., f.,
diminutive csua the little house + Rom. prep. cu with + Rom. c.n., f., pl.,
minuni wonders, miracles). The post-modifier can also be an adjective: Colu
Vesel (Rom. c.n., neuter, col corner, with the enclitic form of the definite
article ul and the apostrophe marking the elision in pronunciation of the ending
consonant l + Rom. adj., m., vesel happy).
Many commercial names in the LL of this resort contain the Romanian
preposition la (at) in initial position. The other components may be:
- appellatives: La Canapele (Rom. c.n., f., pl. canapele sofas, indicating
the type of furniture that is typical of this place), La Epav (Rom. c.n., f., epav
shipwreck), La Frontier (Rom. c.n., f., frontier frontier, border), La

133
Galerie (Rom. c.n., f., galerie gallery), La Pirai (Rom. c.n., m., pl. pirai
pirates), La Stuf (Rom. c.n., neuter, stuf reed);
- adjectives obtained from past participles: La Culcat (Rom. adj. culcat
laid down);
- prepositional phrases with an initial temporal adverbial phrase (La Fr
un Sfert < Rom. fr un sfert a quarter to) or a colloquial, euphemistic
expression functioning like an adverb of manner (La Pe Sub Mn < Rom. pe
sub mn secretly, stealthily).
4. Conclusions
Curiously, although not so numerous, marked names are more prominent
in the LL of Vama Veche. On the one hand, this happens because they occur in
the core area of the resort (not necessarily the geographically central one). On
the other, there is a financial motivation behind this situation: most of the
commercial establishments (guesthouses in particular) that bear unmarked
names are small businesses and they refrain from putting on view their name.
Thus, in most cases, what one finds displayed outside the establishment is just a
sign reading Cazare lodging. Nevertheless, these names do occur on
specialised websites aimed at tourists, contributing to the virtual LL of the
resort (cf. Puzey 2011: 22).
The significant representation of the marked names in the LL of Vama
Veche can also be accounted for by their brand-like function in relation to this
location, whose own name has developed the qualities of trademark use.
Marked names define especially the brand image of the resort (what the tourists
deem representative of it). As the commercial names that became marked
names were established by the community of the vamaioi, it is only natural that
they became associated with this community, as an expression of its existence
in the LL. Therefore, the preservation of this microtoponymic stock suggests
the solid structure of this group and of the peoples feeling of belonging to a
certain community (Helleland 2009: 503).
Whether marked or unmarked, microtoponyms in Vama Veche are
mostly based on anthroponyms. Nevertheless, a difference in their use in these
two contexts needs to be pointed out. Anthroponyms in unmarked names do not
have an identifying function, as it is common for trade names throughout
Romania (and not only) to be derived from first names (full forms or
hypocoristics) and, sometimes, family names. Therefore, such commercial
names convey familiarity and, when they comprise female first names in
particular (as most unmarked names in Vama Veche tend to do), they are also
related to feminine stereotypes of cosiness and good care. The only situations in
which first names help to achieve differentiation are when they are spelled
ALINA BUGHEIU
134
unconventionally (see oni for Hungarian male hypocoristic Sanyi in Papa la

oni), when they lead to acronymic formations (Elga < Ella and Gabi in Elgas
Punk Rock Hotel) or when they consist of bynames and nicknames (La
Mexicanu, Theo Vamaiotu). Therefore, one can state that anthroponyms in the
microtoponymy of Vama Veche function unconventionally, as their main role is
to ensure the association of a business with a certain sociocultural pattern,
rather than its individualisation in the linguistic and socio-economic landscape
of the resort.
The currently booming development of the unmarked group of
commercial names confirms Kallens (2009: 272) statement: Rather than
focusing primarily on territory and tradition, therefore, the model of the LL that
takes account of tourism must incorporate transience and diversity as an
essential part of the social environment. In other words, it is precisely the
stereotyping of the commercial landscape of Vama Veche that ensures its
existence as a resort, while also creating a context for the marked names to
stand out and actually be perceived as a language code defining group identity
(Coulmas 2010: 182). This twofold onomastic orientation in the LL of Vama
Veche highlights the symbolical border on which the resort is situated. By being
in between commodification and individualisation, globalisation and
localisation, mass culture and underground culture, Vama Veche proves to
include both directions of sociolinguistic and cultural development in its
existence, liminally following the logic of both/and (Ooiu 2003: 88).
REFERENCES
Coulmas, Florian (2010), Sociolinguistics: The Study of Speakers Choices, Cambridge
Felecan, Daiana (2011), Aspecte ale polifoniei lingvistice, 2nd edition, revised, Editura Mega,
Cluj-Napoca.
Helleland, Botolv (2009), Place names as identity markers, in W. Ahrens, S. Embleton and A.
Lapierre (eds.), Names in Multi-Lingual, Multi-Cultural and Multi-Ethnic Contact.
Proceedings of the 23rd International Congress of Onomastic Sciences, August 17-22,
2008, York University, Toronto, Canada, York University Toronto Canada, pp. 501-510,
http://yorkspace.library.yorku.ca/xmlui/bitstream/handle/10315/3986/icos23_501.pdf?Seq
uence=1.
Kallen, Jeffrey (2009), Tourism and representation in the Irish linguistic landscape, in E.
Shohamy and D. Gorter (eds.), Linguistic Landscape: Expanding the Scenery, Routledge,
Abingdon, pp. 270-283.
van Langendonck, Willy (2007), Theory and Typology of Proper Names, Mouton de Gruyter,
Berlin New York.
Ooiu, Adrian (2003), An exercise in fictional liminality: The postcolonial, the postcommunist,
and Romanias Threshold Generation, in Comparative Studies of South Asia, Africa and
the Middle East 23 (1&2), pp. 87-105, http://www.cssaame.com/issues/23/15.pdf.

135
Primria comunei Limanu, Judeul Constana (n.d.), Istoricul comunei, http://www.primaria

limanu.ro/portal/portal.nsf/AllByUNID/00000AD6?OpenDocument.
Puzey, Guy (2011), New research and directions in toponomastics and linguistic landscapes, in
Onoma, 46, pp. 211-226.
Recensmntul populaiei i al locuinelor 2011 (n.d.), http://www.recensamantromania.ro/
rezultate-2/.
Soames, Scott (2002), Beyond Rigidity: The Unfinished Semantic Agenda of Naming and
Necessity, Oxford University Press, Oxford.
Zafiu, Rodica (2000), Iaiot, mangaliot, sloboziot, in Romnia literar 17, http: //www.
romlit.ro/iaiot_mangaliot_sloboziot..
ONLINE SOURCES
123urban.ro, http://www.123urban.ro.
Booking.com, http://www.booking.com/city/ro/vama-veche.ro.html.
Plaja.ro, http://www.plaja.ro/vama_veche.
Pubbing.ro, http://www.pubbing.ro/localuri/vama-veche.
apte Seri, http://www.sapteseri.ro/ro/locuri/cluburi-si-baruri/vama-veche.
Turist Info.ro, http://www.turistinfo.ro/vama_veche/cazare-hoteluri-vile-pensiuni-vama_veche.
html.
viaRomania, retrieved from <http://cazare-vamaveche.viaromania.eu>.
RECENZII
Lise Gauvin, Aventuriers et sdentaires. Parcours du roman qubcois,

Honor Champion diteur, Paris, 2012, 243 pp.
Le livre de Lise Gauvin intitul Aventuriers et sdentaires. Parcours du roman qubcois
est en premier lieu un ouvrage de critique littraire, qui traite de la littrature franaise
qubcoise, surtout du roman qubcois du 20e sicle.
Cet ouvrage comprend une introduction, Post ou pri-colonialisme : ltrange modle
qubcois (pp. 7-16) et 7 chapitres : le chapitre 1 Questions de langue : variantes et
variations (pp. 17-43), le chapitre 2 Le romancier et ses doubles : crire, disent-ils (pp.
45-79), le chapitre 3 Aventuriers et sdentaires : lhritage du conte (pp. 81-107), le chapitre
4 Comment peut-on tre Montralais : une ville et ses fictions (pp. 109-129), le chapitre 5
Il tait une fois dans louest : les road novels qubcois (pp. 131-157), le chapitre 6
Thories-fictions, autofictions, romans-pomes et territoires du fminin (pp. 159-179), le
chapitre 7 Ces trangers du dedans : lcriture dite migrante (pp. 181-219). Les sept
chapitres sont suivis dune conclusion ( Une culture et une littrature comme rfrences pp.
220-225), dune bibliographie gnrale de langue et de littrature qubcoises et dune
bibliographie spcifique du roman qubcois (pp. 226-232), dun index des noms propres des
lettrs mentionns au cours de louvrage (pp. 233-236) et dun index des romanciers et des
romans mentionns (pp. 237-242).
Nanmoins, en plus dtre une histoire de lvolution du roman qubcois, cet ouvrage est
aussi un commentaire critique de certains lments considrs comme typiques de cette
littrature. Le livre constitue une prsentation gnrale concise du roman qubcois du 20e sicle.
La matire du livre est organise en fonction de quelques grands thmes et de symboles
importants, choisis par lauteur, qui correspondent aux chapitres mentionns ici. Ces thmes sont
accompagns dexplications incluant certains facteurs historiques, politiques, sociologiques et
anthropologiques qui ont influenc la littrature francophone du Qubec.
Les aspects les plus importants qui sont relevs par Lise Gauvin dans son ouvrage se
rfrent lutilisation des voix narratives par des romanciers qui attribuent la cration de leurs
uvres des personnages inclus dans le schma narratif des romans (v. le chapitre 2), mais aussi
aux sources dinspiration de certains romans qubcois, attribues par la critique littraire la
littrature orale de la Nouvelle-France, tels les contes populaires (v. le chapitre 3). Lise Gauvin
explique aussi le choix du titre de son livre (dans le chapitre 3) : Aventuriers et sdentaires. Ce
titre sappuie sur le dualisme primaire de la culture et de la socit franaises qubcoises, depuis
les commencements de la colonie dont le nom tait la Nouvelle-France jusqu la fin du 19e
sicle, puisqu cette poque-l il y avait effectivement deux modes de vie et dexploitation du
territoire pour la survie des communauts franaises : dune part, lexistence rurale, appuye sur
le travail agricole (pratiqu par les paysans, les sdentaires), dautre part, le mode de vie oppos,
pratiqu par les aventuriers, qui taient reprsents dabord par les coureurs des bois, cest--dire
par les chasseurs et par les trappeurs professionnels franais canadiens du nord des tats-Unis et
du Canada (les commerants de fourrures), remplacs, partir du milieu du 18 sicle jusquau 19
sicle, par les voyageurs ou les engags (qui possdaient des congs de traite mis par le roi et qui
gagnaient un salaire), remplacs ensuite, au 20e sicle, par ceux qui aimaient voyager travers les
routes et les chausses pour se faire embaucher ( court terme pour des travaux agricoles ou pour
dautres types de tches), remplacs, leur tour, par ceux qui aimaient vivre dans les grandes
villes. Dans le chapitre 4, Lise Gauvin se rfre justement aux romans urbains, qui trouvent leurs
138
sujets dans les villes franaises canadiennes, y compris dans des villes importantes, tels Qubec et
Montral, par exemple : en 1945, Gabrielle Roy inaugure, avec Bonheur doccasion, ce
nouveau cycle de fictions. (p. 109), [] la suite de Gabrielle Roy, dans les annes 50 et 60,
Montral devient un cadre romanesque privilgi, o voisinent le naturalisme dun Bessette (La
Bagarre), les sarcasmes et gauloiseries dun Jean-Jules Richard (Ville rouge, Faites-leur boire le
fleuve, Carr Saint-Louis, Centre-Ville), la prose critique dun Girouard (La Ville inhumaine)
[] (p. 112), mais cest surtout avec Michel Tremblay et ses Chroniques du Plateau MontRoyal [que] le roman qubcois prend dfinitivement possession de la ville [] [l]e Montral
romanesque de Tremblay commence dans les annes 1940 et se dploie jusquaux annes 80 (p.
116). Le chapitre 5 est consacr aux romans du voyage, qui traitent des dplacements des
personnages travers des espaces gographiques immenses. Ainsi, Lise Gauvin montre que
bien que la ville soit le cadre romanesque privilgi par les romanciers qubcois au cours des
dernires dcennies, lexploration du continent nord-amricain a galement sduit les crivains
qui, partir des annes 1980, ont promen leurs hros dest en ouest et de Gasp San Francisco
[...] ; [d]ans ce contexte, le roman du voyage ou le road novel peut tre considr comme la
mtaphore de tout rcit dans la mesure o il vient doubler, par son sujet mme, le parcours narratif
sur lequel se fonde le genre romanesque (p. 131). Lise Gauvin mentionne quelques exemples de ce
type de romans, o apptit despace et dsir dcriture sont intimement lis [...] , ce qui se vrifie
[...] avec les romans qubcois qui ont comme sujet la traverse de lAmrique, parmi lesquels
tout dabord Volkswagen Blues de Jacques Poulin, inspir du modle On the Road de Kerouac
[...] ; [d]eux autres romans, Une histoire amricaine de Jacques Godbout et Copies conformes de
Monique Larue, situent leur intrigue San Francisco, dernire halte du rcit de Poulain [...] ; [d]e
leur ct, les voyageurs dUn train pour Vancouver de Nicole Lavigne et du Joueur de flte de
Louis Hamelin parcourent le Canada dest en ouest (p. 132). Le chapitre 7 est consacr
lcriture dite migrante , cest--dire la littrature crite par les immigrants vivant au Qubec.
Ainsi, tel que lexplique lauteur : au cours des annes 1980, la littrature qubcoise est
engage dans diverses pratiques de dcentrement, pratiques dont la revue Drives, fonde par
lcrivain dorigine hatienne Jean Jonassaint, tmoigne loquemment [...] [c]e lieu dchanges
interculturels de premire importance est bientt suivi par la cration de la revue Vice Versa
(1984) par les Italo-Qubcois Fulvio Caccia et Lamberto Tassinari, ainsi que par la fondation par
Ghila Benestry Sroka de la Parole Mtque (1986), qui se dfinit comme magazine du
renouveau fministe [...] [l]a notion de littrature migrante est ainsi apparue pour dsigner les
textes dcrivains de diverses origines [hatienne, italienne, juive polonaise, asiatique dExtrmeOrient, etc.] ayant choisi de publier en franais [...] [b]ien avant cette dcennie, des auteurs
avaient immigr au Qubec sans que leur prsence ne soit souligne par une dsignation
particulire (p. 181). Nous ne mentionnerons ici que deux des exemples prsents par Lise
Gauvin parmi ces auteurs immigrs : Dany Laferrire, dorigine hatienne (tabli en 1976
Montral) (Comment faire lamour avec un Ngre sans se fatiguer ; Cette grenade dans la main
du jeune Ngre) et Marco Micone, dorigine italienne (qui avait immigr avec ses parents en 1958
Montral), auteur dramatique (ses pices de thtre prenant pour sujets des aspects de la vie de
plusieurs gnrations dimmigrs, ainsi que le souligne Lise Gauvin, telles les pices Gens du
silence, Addolorata et Dj lagonie), auteur dun manifeste (Speak What) et dun rcit (Le
Figuier enchant), mais aussi de traductions pour le thtre.
Par ailleurs, dans son ouvrage, Lise Gauvin relve galement des aspects linguistiques qui
caractrisent le franais qubcois. Nous prsentons ici les traits essentiels de ces aspects
linguistiques qui sont abords par Lise Gauvin :
le statut de la francophonie dans cette rgion du monde, son tat jeune et rcent,
puisqu on donne gnralement la date de 1837 et LInfluence dun livre ou Le Chercheur de
trsors de Philippe Aubert de Gasp fils comme premier roman. (p. 13) ;
le terme de pricolonialisme, choisi par Lise Gauvin pour dsigner ce statut de la
culture et de la littrature franaises qubcoises par rapport au franais de la Mtropole, mais
139
aussi langlais : Aussi me semble-t-il que le terme le plus adquat pour dcrire ltrange
modle qubcois, sa complexit et son originalit, est celui de pricolonialisme, car on indique
que cette littrature reste priphrique dans lensemble de la francophonie, mais aussi par rapport
laxe colonialiste ou postcolonialiste, comme toute pense dualiste qui ferait lconomie des
nombreux rseaux dappartenances et dinfluences qui la traversent et en font la spcificit ; ce
concept rejoint ce que Rownald Smith appelle le side-by-sidedness (p. 12) ;
laffirmation, dans une littrature qui cesse dtre traditionnelle au cours de la deuxime
moiti du 20e sicle, du vernaculaire qubcois (de ses particularits locales, comme des traits du
joual) et du multilinguisme, par exemple, lutilisation, par les mmes locuteurs appartenant la
communaut linguistique qubcoise, du franais, de langlais et de litalien, mais aussi dautres
langues ; voir la reprsentation de la pice de thtre les Belles-Surs de Michel Tremblay, qui
intgre la langue populaire au rcit et qui produit un effet joual du texte (pp. 26-27) ;
ladoption lcrit, dans les romans qubcois de la deuxime moiti du 20e sicle, de
plusieurs registres de langue franaise (le franais standard utilis en France, mais galement un
franais standard qubcois, le franais populaire qubcois, etc.) (v. le chap. 1) ;
la fminisation du franais qubcois par les auteurs fminins de la
littrature qubcoise : par exemple, Lise Gauvin prcise que si les annes 1960 ont t
marques, au Qubec, par une nouvelle donne politique, accompagne dune rflexion soutenue
de la part des crivains sur le rle et la fonction sociale de la littrature, il est gnralement admis
que les annes 1970 ont t surtout remarquables par les prises de position et les mouvements
fministes (p. 159) ; si ce courant littraire devient intressant sur le plan linguistique, selon
Lise Gauvin, cest parce que les femmes ont voulu penser dabord la langue, articulant leur
thorie des pratiques transgressives et provocatrices (p. 159) et qu elles se sont nommes
crivaines, fires de cet e que lon disait muet (p. 159) : lauteur explique le fait que ce sont ces
questions [qui sont] poses la langue et au langage [...] [et qui sont] formules dans des
noncs de type programmatique, mais aussi dans une srie de textes mixtes qui empruntent
tantt au manifeste ou lessai, tantt au rcit ou au pome en prose, tantt lautobiographie et
lautofiction les formes de leur agencement (p. 159) ; lune des figures importantes dcrivaines
cites est Nicole Brossard, qui crit des thories-fictions telles que LAmr ou le chapitre effrit,
o elle se propose de fminiser le langage en franais qubcois.
Deux conclusions du livre nous semblent importantes. Nous les illustrons ici par les
propos mmes de Lise Gauvin :
(i) Une culture et une littrature comme rfrences, telle tait la caractristique que je
signalais comme une nouveaut de la littrature qubcoise en 1983 dans Trajectoires, ouvrage
publi en collaboration avec des collgues de la Belgique francophone (pp. 223-224) 1 ;
(ii) Depuis les premiers crits, aventuriers et sdentaires sont reprsents dans le roman
qubcois. Entre le dsir denracinement et le besoin de nomadisme sinscrivent les tapes dun
parcours qui passe du roman paysan au roman urbain, de la traverse de la ville celle du
continent amricain. Le franais qui sy affiche renvoie une langue dcomplexe, arrime une
culture marque par lhtrogne et le non-conventionnel. Une culture porte par la forte
prsence des crivaines et riche des apports des nouveaux arrivants, ces trangers du dedans
devenus peu peu les tmoins privilgis et les porte-parole dune collectivit en mutation. Une
culture marque galement par un questionnement constant quant la lgitimit et la fonction
du littraire dans lespace social. Lcrivain au Qubec est celui qui, sans abdiquer son devoir de
vigilance concernant le statut accord au franais dans lensemble de la collectivit, rend compte
dune variance infinie des potiques. (p. 224) [nos soulignements].
Bien que Lise Gauvin fasse des rfrences, tout au long de chaque chapitre, de
nombreux auteurs franais qubcois fameux, tels le romancier Jacques Ferron, le pote Gaston
1
V. Lise Gauvin et Jean-Marie Klinkenberg (1985), Trajectoires. Littrature et

institutions au Qubec et en Belgique francophone, Labor, Bruxelles.
140
Miron et dautres, mais aussi aux titres des oeuvres crites par ceux-ci, louvrage Aventuriers et
sdentaires. Parcours du roman qubcois ne traite pas uniquement de lhistoire du roman
qubcois, ainsi que nous lavons vu : selon les propos de Lise Gauvin dans sa conclusion, cet
ouvrage sest voulu une invitation lire le roman qubcois davantage quune histoire de son
volution (p. 221). Dautre part, tel que nous avons essay de le dmontrer, les observations de
lauteur concernant le franais qubcois savrent remarquables pour le champ de la linguistique.
Par consquent, si louvrage de Lise Gauvin retient en premier lieu lattention des lecteurs
intresss par la littrature, lutilit du livre pour le domaine de la linguistique est loin dtre
ngligeable. De plus, la prsentation de lauteur est ralise dans un langage qui est accessible au
grand public. Ainsi le livre Aventuriers et sdentaires. Parcours du roman qubcois de Lise
Gauvin est-il destin toutes les catgories de lecteurs : aux spcialistes de la littrature , de la
linguistique romanes et des tudes francophones, mais galement aux amateurs des belles-lettres.
ANTONIA CIOLAC*
Liliana Soare, Din nceputurile terminologiei tiinifice romneti, vol. I, Medicin;

tiine ale naturii (botanic, zoologie, geologie) Studiu introductiv, selecie, transcriere,
comentarii i glosar, Paralela 45, Piteti, 2013, 273 pp.
La lectura del primer volumen del estudio Din nceputurile terminologiei tiinifice
romneti [Sobre los principios de la terminologa cientfica rumana] realizado por Liliana Soare
puede resultar fascinante, sin lugar a dudas, para los terminlogos o para los que se dedican a los
mbitos abarcados por la autora: medicina, botnica, zoologa, geologa. Tambin puede interesar
a los fillogos y los apasionados por la lengua rumana en general y por su evolucin en especial,
acorde con el progreso de la ciencia en el mundo, con la necesidad de adoptar nuevos trminos y
tambin con la capacidad del rumano de representar igualmente un instrumento de comunicacin
a nivel cientfico.
El presente estudio concentra la informacin que podra servir como punto de partida para
otras investigaciones o incluso para creaciones literarias con propsito reconstitutivo, cautivante
para un contemporneo, fillogo o no, que necesita una gua que le acerque a la interpretacin de
los textos antiguos, percibidos en general como menos accesibles y comprensibles (como las
crnicas de los historiadores moldavos y valacos del siglo XVII o la Istoria ieroglific de
Dimitrie Cantemir). Adems, el volumen de Liliana Soare es un estudio cientfico riguroso que
contina la investigacin de unas obras de referencia bastante recientes de la misma rea como
Formarea terminologiei tiinifice romneti [La formacin de la terminologa cientfica rumana]
de N.A. Ursu, mprumutul lexical n procesul modernizrii limbii moderne [El prstamo lxico en
el proceso de modernizacin de la lengua rumana] de N.A. Ursu y D. Ursu, Terminologie i
terminologii [Terminologa y terminologas] un estudio en dos volmenes coordinado por
Angela Bidu-Vrnceanu.
El estudio Din nceputurile terminologiei romneti presenta y comenta textos cientficos
(una manera de rescatarlos del olvido y de situarlos en los focos del anlisis lingstico)
pertenenciendo unos al mbito de la medicina (los textos de Amfilohie Hotiniul, Petre Maior,
Alexandru Teodori, tefan Vasile Episcopescu, Pavel Vasici-Ungurean) y otros a las ciencias de
la naturaleza (Amfilohie Hotiniul, Iacob S. Cihac, Iulius Barasch, Dimitrie Iarcu y D. Brndz).
Otro mrito de este estudio es el esfuerzo de haber transcrito en el alfabeto latino los textos
*
Institut de Linguistique Iorgu Iordan Al. Rosetti de lAcadmie Roumaine,

Bucarest, antonia_ferihan@yahoo.com.
141
impresos en el alfabeto cirlico, segn los sistemas propuestos en los estudios de I. Fischer, M.
Avram, Fl. Dimitrescu, Finua Asan, I. Gheie y Al. Mare. El volumen incorpora el estudio sobre
los textos cientficos escritos entre finales del siglo XVIII (Gramatica de la nvtura fizicii de
Amfilohie Hotiniul, del ao 1796) hasta la segunda mitad del siglo XIX (Curs elementar de
istoria natural. Zoologia de D. Brndz, del ao 1872). Esta franja temporal incluye, segn las
especificaciones ofrecidas por la autora, dos etapas: la primera, entre 1780 y 1860-1870, de
absorcin y de formacin del estilo cientfico rumano, y la segunda, despus de 1860, que marca
un principio de autonoma de la investigacin cientfica rumana, representada a nivel
internacional por Victor Babe, Nicolae Teclu, Anghel Saligny, Dimitrie Brndz, entre otros.
Adems del Argument [Argumento] (pp. 6-7) y Not asupra ediiei [Prlogo] (pp. 811, que contiene una tabla til de las correspondencias entre las letras del alfabeto cirlico y las
letras latinas, ya que cinco textos estn escritos en cirlico, mientras que uno est escrito con
alfabeto ruso), el estudio incluye tres captulos introductivos.
Din nceputurile studiului tiinific n limba romn repere istorice [Sobre los
principios del estudio cientfico en rumano referencias histricas] (pp. 12-17) se refiere a la
actividad cientfica desarrollada antes del perodo comentado en el presente volumen, dividido a
su vez en dos etapas: antes de 1640 (caracterizado por un nmero reducido de obras) y despus de
1640 (cuando se puede constatar un aumento del inters por las ciencias en los Pases Rumanos).
En Schi de istorie a medicinii i a tiinelor naturale n limba romn [Esbozo de la
historia de la medicina y de las ciencias naturales en la lengua rumana] (pp. 17-22) se constata
que hasta el ao 1830, la ciencia mdica era ms desarrollada en Transilvania comparativamente
con los Pases Rumanos bajo la influencia del Iluminismo, para que despus de este ao, el centro
de gravedad se desplazara hacia los Principados de Moldavia y Valaquia; el inters para las
ciencias de la naturaleza, que, igual que la medicina, conoce un perodo de florecimiento sobre
todo despus de 1830, se va a manifestar pronto en primer lugar en Transilvania, donde se
convierte en asignatura para estudiar en las escuelas aun desde el siglo XVII; adems, un listado
de nombres de plantas se incluye en el primer diccionario de la lengua rumana - Dictionariul
valachico-latinum [Diccionario valaco-latino];
Didacticismul trstur definitorie a primelor texte tiinifice [El didacticismo
rasgo caracterstico de los primeros textos cientficos] (pp. 22-33) muestra que, al nivel de la
enunciacin (que concierne la relacin locutor/interlocutor: raportul locutor/interlocutor i
poziia acestora fa de obiectul cunoaterii, p. 22), el texto cientfico es de tres tipos:
especializado, didctico o de divulgacin (popularizacin). El didctico se manifiesta a travs de
los textos estudiados por las formulaciones dirigidas directamente al interlocutor con intencin
pedaggica, creando la apariencia de un dilogo con el lector, por las estructuras interrogativas,
por las definiciones, sobre todo con propsito de divulgacin, a travs de las cuales se intenta
establecer una relacin entre la nocin nueva y los conocimientos anteriores de los lectores, por
las indicaciones etimolgicas, los ejemplos, las explicaciones, las taxonomas y las ilustraciones.
A estos captulos introductivos les sigue la parte ms amplia del estudio que incluye los
textos cientficos, ordenados cronolgicamente y comentados desde una perspectiva
terminolgica, ya que se aade tambin un listado de trminos y sintagmas terminolgicos del
mbito mdico o de las ciencias de la naturaleza, de origen popular o prstamos.
Para el mbito de la Medicina (pp. 34-115), representado por numerosos textos y
caracterizado por un lxico especializado de formacin temprana, se mencionan y se analizan los
siguientes autores:
Amfilohie Hotiniul, Gramatica de la nvtura fizicii [Gramtica sobre la enseanza de
la fsica] (1796), pp. 34-52, representante del iluminismo en Moldavia, conocedor de varios
dominios cientficos, cuyo lenguaje se caracteriza por prstamos lxicos, inserciones de etimones
populares, igual que por la utilizacin de sinnimos para algunos trminos cientficos, prctica a
evitar para no dar lugar a ambigedades, pero justificable por aquellos principios de las escrituras
con carcter cientfico en el espacio rumano;
142
Petre Maior, nvtura pentru ferirea i doftorirea boalelor [Enseanza sobre la
prevencin y el tratamiento de las enfermedades] (1816), pp. 53-72, autor dedicado a varias reas
cientficas, a la traduccin y a la adaptacin de textos extranjeros, que recurre tambin a calcos, a
la terminologa popular, a prstamos del italiano, pero tambin del aromuno;
Alexandru Teodori, Scurt artare despre om i despre ntocmirile lui [Corta
presentacin del ser humano y de su constitucin] (1825), pp. 73-85, autor al que se debe la forma
final del Lexicon de la Buda [El diccionario de Buda], introduce en su obra trminos anatmicos
internacionales, especificando para la mayora de ellos su equivalente griego;
tefan Vasile Episcopescu, Oglinda sntii i a frumuseii omeneti [Espejo de la
salud y de la belleza humana] (1829), pp. 85-102, que sienta las bases de la terminologa mdica
en Valaquia, caracterizada por la frecuencia de los trminos populares y de los calcos;
Pavel Vasici-Ungurean, Antropologia sau scurta cunotin despre om i despre
nsuirile sale [Antropologa o corto conocimiento sobre el ser humano y sobre sus
caractersticas] (1830), pp. 103-115, cuya significativa actividad mdica incluye tambin la
preocupacin por crear un lxico mdico rumano especializado, aunque su obra tambin contiene
trminos populares.
En cuanto a las ciencias de la naturaleza (botnica, zoologa, geologa), pp. 116-257, se
analizan textos que pertenecen a los siguientes autores:
Amfilohie Hotiniul, con Gramatica de la nvtura fizicii [Gramtica sobre la
enseanza de la fsica] (1796), pp.116-146 ya que, segn la autora, se trata de la primera
escritura que representa un principio de terminologa cientfica botnica, zoolgica y geolgica en
rumano, caracterizndose por la presencia de los prstamos y de los trminos populares;
Iacob S. Cihac, con Historia natural [Historia natural] (1837), pp. 147-182, miembro de
honor de la Sociedad Acadmica Rumana, autor del primer tratado de ciencias naturales impreso
en rumano, que recurre a prstamos lxicos para crear un lenguaje especializado utilizable en los
mbitos cientficos rumanos;
Iulius Barasch, con Minunele naturei [Las maravillas de la naturaleza] (1852), pp. 183209, que se ha hecho importante por haber introducido la terminologa especializada que, segn la
autora, se sigue utilizando incluso en los manuales actuales de la enseanza media y por haber
adoptado trminos cientficos del francs;
Dimitrie Iarcu, con Elemente de istorie natural [Elementos de historia natural] (1860),
pp. 210-243, autor y traductor de obras cientficas francesas, cuya obra contiene trminos
cientficos que siguen siendo utilizados aun en el presente;
D. Brndz, con su Curs elementar de istoria natural. Zoologia [Curso elemental de
historia natural], pp. 244-257, el padre de la ciencia botnica rumana, fundador del Jardn
Botnico de Bucarest y miembro de la Academia Rumana, representa un momento de madurez
del lenguaje cientfico rumano, ya que utiliza en el manual analizado en Din nceputurile
terminologiei tiinifice romneti una terminologa que, en su gran mayora, sigue siendo
vigente, segn la autora.
Las obras de estos autores reflejan, en buena medida, a nivel lingstico, no slo las
edades de la terminologa en los Pases Rumanos, sino tambin el desarrollo de la ciencia en este
espacio rumano.
A este vasto captulo dedicado a los textos del mbito mdico y de las ciencias naturales
le sigue un Glosario (pp. 259-262) que rene trminos arcaicos de todos los mbitos
mencionados con sus correspondientes de la terminologa rumana actual. Una recomendacin en
cuanto a la construccin del glosario sera que ste fuera ms extenso y contuviera, en caso de una
posible reimpresin, unas cuantas menciones en lo que concierne la obra de procedencia (el autor,
el ttulo, el ao, la pgina), para facilitar las investigaciones de los que quisieran continuar este
estudio. Aunque los anlisis de las obras de los autores mencionados contienen tambin
enumeraciones de los trminos populares o neolgicos, retomar estos vocablos en un glosario
final hara ms visible la continuidad o la detencin/suspensin de la evolucin terminolgica de
143
una obra a otra situadas en intervalos temporales diferentes. Sin lugar a duda, es posible que
algunos trminos sean presentes en los textos de varios autores o, todo lo contrario, sean
especficos de un solo autor, pero justamente estas coincidencias o diferencias pueden representar
puntos de inters en un futuro estudio terminolgico.
Los Anexos incluyen fotos de las hojas de titulo/portadas y algunas pginas de las
ediciones consultadas (pp. 263-272), mientras que Izvoarele [Las fuentes] (pp. 273) contienen una
enumeracin de las obras cientficas en orden alfabtico segn los autores.
El estudio de Liliana Soare puede representar un importante punto de partida tanto para
sus propias investigaciones como para los dems investigadores que se propongan adentrarse en
un cierto mbito y seguir la evolucin del lenguaje cientfico rumano, su aproximacin paulatina
de los requisitos terminolgicos internacionales.
MIOARA ADELINA ANGHELU*
David Deterding and Salbrina Sharbawi, Brunei English: A New Variety

in a Multilingual Society, Springer, Dordrecht, 2013, 170 pp.
David Deterding and Salbrina Sharbawis book consists of the Conventions in the
transriptions, the list of Abbreviations, eight chapters, five appendices, the References and
an Index.
Chapter 1, Introduction (pp. 1-11), is an overview of the history, demographic profile
and sociolinguistic situation of Brunei. Particular attention is paid to the status of this variety, i.e.
English in Brunei or Brunei English. The authors argue (p. 7) in favour of using the latter
label, while acknowledging the fact that this is a controversial option. Also discussed is the
considerable variation typical of Brunei English. The chapter ends with a detailed presentation of
the data analyzed in the book: recordings from 53 formal interviews (38 female and 15 male
subjects); an extensive informal interview with Umi, a 33-year-old female fluent speaker of
English; data from the University of Brunei Darussalam Corpus of Spoken Brunei English
(UBDSCSBE); written data from issues of the two local English-language newspapers (The
Brunei Times and the Borneo Bulletin); extracts from texts printed at the Kampong Ayer
Cultural and Tourism Centre; data from on-line postings on the internet discussion forum
BruDirect.
In chapter 2, Education in Brunei (pp. 13-21), the authors discuss the history of the
educational system in Brunei, highlighting the bilingual education policy implemented via the socalled dwibahasa dual languages bilingual system of education. In this context, they also
address the issue of the educational divide in Brunei (pp. 19-21), and account (p. 21) for their
decision to use data illustrative of the proficient and fluent variety of English that exists among
well-educated young Bruneians to the exclusion of the learner English of the less welleducated.
Chapter 3, Phonetics and phonology (pp. 23-47), the longest in the book, is a detailed
description of the pronunciation of Brunei English. The data analyzed consist of the 53
UBDCSBE recordings of the Wolf passage, the 53 five-minute interviews of the same
UBDSCSBE speakers, the extensive interview with Umi, and attempted phonetic transcriptions
*
Universidad de Bucarest, Facultad de Lenguas y Literaturas Extranjeras, Departamento
de Lingstica Romnica, Lenguas y Literaturas Ibero-romnicas e Italiano,
mioara.angheluta@gmail.com.
144
by students. The sections on segmental phenomena (pp. 24-41) are concerned with the following
topics: TH-stopping, i.e. the tendency towards the substitution of the stops [t] and [d] for the
fricatives [] and [] respectively; the reduction of word-final consonant clusters, via consonant
deletion; intrusive [t] in word-final position; the realization as a glottal stop [] of word-final /t/
and /k/; obstruent devoicing in word-final position; /l/-vocalization; /l/-deletion; rhoticity; the
tendency towards the levelling of the length distinction in the lexical sets2 KIT and FLEECE, FOOT
and GOOSE; the confusion of the vowels in FACE and TRAP; the tendency towards the
monophthongization of the FACE and GOAT vowels; the absence of reduced vowels; instances of
spelling pronunciations and of idiosyncratic pronunciations. The suprasegmental phenomena
discussed (pp. 41-47) include word stress, compound stress, rhythm, sentence stress, deaccenting, and the utterance-final rising pitch.
Chapter 4, Morphology and syntax (pp. 49-70), is based on the interviews in the
UBDCSBE and it focuses on a selected number of salient morphological and syntactic features of
Brunei English. The issues covered are: the occurrence vs. non-occurrence of the plural suffix -s;
the treatment as count-nouns of nouns which are uncountable in standard varieties of natively
spoken English; the omission of the plural suffix -s after one of; the occurrence of plural forms
such as brother-in-laws or runner-ups; the classifier-like use of piece; subject-verb agreement,
with -s either unexpectedly occurring on or missing from present tense verbs; the use of theres
with plural noun phrases; the occurrence of -s with modal verbs; the effect on agreement patterns
of an intervening noun between the head of the subject and the main verb, which triggers the use
of the suffix -s, e.g. most of the words of English comes, or of were, e.g. the performance of the
participants were satisfactory; the use of the present tense for past time reference; frequentative
will; the widespread use of would to express tentativeness and also as a stylistic variant of will;
the use of the auxiliary verb do in affirmative declarative sentences; the occasional occurrence of
ever in affirmative sentences; the occurrence of null subjects; the occurrence of subject-auxiliary
inversion in embedded questions; the variable usage of determiners; the absence of the definite
article before names of countries, e.g. shes in UK right now or US would be good; the use of
affirmative answers to negative questions; the complementation patterns Adj to V and Adj Ving; the occurrence of redundant prepositions, e.g. discuss about/on, emphasize on, grasp at.
Chapter 5, Discourse (pp. 71-87), is an overview of the most typical discourse patterns
of Brunei English. The analysis is based on data of both spoken and written discourse; the former
are from the UDSCSBE interviews while the latter come from the two national English-language
newspapers The Brunei Times and the Borneo Bulletin and from texts displayed at the Kampong
Ayer Cultural and Tourism Centre. The authors deal in turn with: the discourse particles bah
perhaps the favourite particle for Bruneians (p. 72), lah and kan; the use of yeah to signal the
end of the turn or a change of mind; the use as discourse markers of the phrases sort of/kind of;
the particle tsk (where <tsk> stands for an alveolar click); topic fronting, e.g. my grandparents
they speak Hakka; the use of -wise to mark a noun as the topic, e.g. job-wise, I wouldnt mind any
job that lets me travel; the occurrence of compared to instead of than; reduplication; the repetition
of lexical items; the widespread occurrence of words with similar meanings connected by and,
e.g. love and affection, enhance and upgrade; instances of tautology, e.g. total overall; the cooccurrence of such as and and so forth/and so on in the same sentence; the tendency towards
what the authors call overdoing explicitness (p. 84), e.g. the diabetes disease; the tendency
towards using whereby as an all-purpose connector; the tendency towards lengthy sentences; the
occurrence in written discourse of ill-formed run-on sentences.
Chapter 6, Lexis (pp. 89-106), is concerned with the characteristics of the vocabulary of
Brunei English. First, the influence of other languages is illustrated: the borrowings from Arabic
via Malay, consisting mostly of religious terms; the borrowings from Malay of terms such as for
2
The lexical sets (see p. 23) are taken from J. C. Wells (1982), Accents of English,
Cambridge University Press, Cambridge.
145
addressing or referring to the royal family, to designate items of clothing, for specific cultural
items, e.g. traditional dances and festivals; the (probably Chinese-influenced) use of three or/to
five years; calques from Malay, e.g. four-eye meeting. Next, the authors discuss phenomena such
as: the widespread occurrence of initialisms, e.g. BSB Bandar Seri Begawan, the capital city of
Brunei; the use of specific clippings, e.g. ex-coms ex-committee members; blends, e.g. promex
promotion exam; the occurrence of semantic shifts, e.g. bring take, stay live, sober
ashamed; the existence of a special terminology for participants in various sports, e.g. shuttler
badminton player.
Chapter 7, Mixing (pp. 107-116), focuses on code mixing and code switching involving
Malay and English. Also addressed, though only tentatively, are the factors that account for the
widespread occurrence of both code mixing and of code switching: what the authors call the
inability to think of a word (p. 111); the need to explain something; the lack of English
equivalents for terms relating to the Islamic religion, to food and to other elements of local
culture; direct quotations; stylistic reasons. The authors conclude that both code mixing and code
switching appear to be the norm (p. 116), i.e. the unmarked option3 of speakers.
In chapter 8, Brunei English in the world (pp. 117-125), the authors first assess the
status of Brunei English in terms of the called Dynamic Model4 and conclude that Brunei
English is still in the third stage, i.e. nativization. The other issues covered in this chapter are the
place of Brunei English within World Englishes, the intelligibility of Brunei English in
international settings, and the pedagogical implications of the findings in light of the paradigm
shift away from strict adherence to the norms of native-speakers.
There are five Appendices: Appendix A, The female UBDCSBE speakers (pp. 127128), lists the age, ethnicity, first language and second language of the female subjects in the
UBDCSBE corpus; Appendix B, The male UBDCSBE speakers (pp. 128), lists those of the
male subjects; Appendix C, The Wolf passage (pp. 128-129), consists of the passage used for
the readings; Appendix D, Transcripts of the interview with Umi (pp. 129-151), is the
orthographic transcript of the .wav files of the extensive interview with Umi; Appendix E, The
BruDirect data (pp. 151-153), provides the identification code, title, date of the initial post, and
URL of the 15 discussion threads from the BruDirect internet discussion forum, analyzed in
chapter 7.
Deterding and Sharbawis monograph is the first book-length description of Brunei
English. The analysis of Brunei English data includes references to other New Englishes, in
particular to those spoken in Malaysia, Singapore5, the Philippines and Hong Kong, as well as to
natively spoken varieties. Due consideration is given to the complex interplay of factors in the
emergence and development of Brunei English, including the influence of Malay and Chinese.
Controversial issues and less well documented facts are discussed in a remarkably objective way.
The frequently tentative nature of the findings, given the insufficiency of the data in many areas
of interest, is repeatedly underscored. The book is written in a lively and engaging style, which
makes it accessible to a wide readership. The volume has been very carefully edited and indexed.
There are very few inaccuracies: with is about (p. 78) instead of which is about; hybrid
mature (p. 110) instead of hybrid nature; Schneiders modal (p. 120) instead of Schneiders
3
In the sense of Carol Myers-Scotton (1993), Social Motivations for Codeswitching,

Clarendon Press, Oxford.
4
First presented in Edgar W. Schneider (2003), The dynamics of New Englishes: From
identity construction to dialect birth, Language, 79, pp. 233-281. See also Edgar W. Schneider
(2007), Postcolonial English. Varieties around the World, Cambridge, Cambridge University
Press, chapter 3, and Edgar W. Schneider (2011), English around the World: An Introduction,
Cambridge, chapter 2.
5
The first author is a well-known specialist in Singapore English; see e.g. David
Deterding (2007), Singapore English, Edinburgh University Press, Edinburgh.
146
model; Ishaminas first name is Athirah, not Athura (pp. 157 and 166); the reference (p. 159) to
Noor Azam Haji-Othman (2012b) does not include the page numbers.
In conclusion, Brunei English: A New Variety in a Multilingual Society is a welcome
description of a hitherto under researched variety of New English, advancing our understanding
not only of south-east Asian Englishes but also of World Englishes, for which the authors are to
be commended.
ANDREI A. AVRAM*
Mihaela Tnase-Dogaru, The Methodology of Teaching English as a Foreign

Language, Editura Universitii din Bucureti, Bucharest, 2013, 293 pp.
The book under review here is a course in foreign language teaching. It is addressed to
second and third year BA students, as well as second year MA students. Its main part is divided
into five chapters, each devoted to one aspect of language teaching. The chapters are freestanding and can be used independently of one another.
The book starts with a Foreword, followed by an Introduction, and closes with a short
Bibliography.
Chapter 1, Methods and approaches to the teaching of English as a foreign language
(pp. 15-117), presents the main methods and approaches currently used in EFL classrooms. This
is a good starting point for such a book as I do not believe that the study of language teaching
methods should be excluded from language teacher education. The chapter starts with the most
traditional and widely criticized method, the Grammar-Translation Method, and ends with the
most flexible and widely praised approach, the Communicative Approach. Between these two, we
find other methods and approaches, such as the Direct Method, the Audio-Lingual Method, the
Silent Way, the Total Physical Response Method, Community Language Learning, and
Suggestopedia. Each of the eight sections presents the principles of the respective method, its
distinguishing features and techniques, as well as the criticism(s) it has received during the years.
Thus, the author goes beyond the simple presentation of the different methods of teaching English
as a foreign language. For instance, she discusses the strong and weak points of each of them, and
gives her own opinion about which method would be appropriate in Romanian schools. However,
we should not be blinded by the criticisms of any of these methods, and thus fail to see their
invaluable contribution to teacher education. Throughout the chapter, we are reminded of the fact
that despite their drawback, individual techniques, even if combined, may prove to be successful
in certain cases and may lead to good results. The last part of the chapter is a brief review of these
methods and approaches, and it contains the key to exercises.
Chapter 2, Teaching grammar (pp. 118-157), is dedicated to the hotly-debated topic
whether grammar should be taught at all or not, and if yes, then how. The first part confronts
teachers with the two ways grammar can be taught: teachers may choose the inductive way of
teaching grammar, by means of which they let students discover the rules themselves, or they can
choose the deductive way of teaching grammar and give the students all the rules they need to
know. The second part of the chapter provides communicative activities for the three stages of
teaching grammar. The tight teacher-controlled activities of the first stage (the presentation and
explanation of new material in a clear and comprehensible way) are followed by students
performances (practice to consolidate knowledge), and these in general culminate with the final
*
University of Bucharest, Department of English, andrei2.avram@gmail.com.
147
stage, which offers great learner freedom (free production). These stages may not occur only in
this order, i.e. they do not represent a rigid linear classroom routine. Moreover, they may
sometimes be combined within one and the same activity, and it is possible to both start and finish
a lesson with the same stage. In order to make sure the readers understand the main ideas of the
chapter, the author discusses some problematic lessons and explains what seems to be wrong with
them.
Chapter 3, Teaching integrated skills (pp. 158-219), focuses on teaching the four basic
skills: reading, listening, writing, and speaking. Although most learners can cope with a higher
level in receptive skills than they can in language production, an equal importance is given to
both the receptive and the productive skills. First, it attempts to explain a number of specialist
skills which learners have in their own language but which need to be re-activated when it comes
to reading in a foreign language, i.e. English. Then, it discusses some ways of, and reasons for,
reading a text, and clarifies some principles and aspects of the nature of reading. As reading, just
like listening, involves active participation on the part of the reader, special emphasis is laid on
pre-reading, while-reading, and post-reading activities. The teaching of listening skills follows the
methodological model of the teaching of reading skills, but training students in the former skills
presents problems not found in reading material. The reasons for teaching listening may be
different as well, and the listening material itself can come in at least two different forms: audio
tapes and video. As spoken language differs markedly from written text, teachers should prepare
differently and should pay attention to other aspects of pre-listening, while-listening, or postlistening activities. As far as writing is concerned, unfortunately it is often considered the
forgotten skill and does not attract great attention. Nonetheless, writing activities can be very
helpful in reinforcing the new language, in helping students learn the grammatical structure of the
foreign language, in getting them to acquire the abilities and skills they need in order to produce a
range of different kinds of written texts, as well as some of the basic conventions of writing. The
largest part of the section proves that there are numerous activities and tasks that can stimulate
writing. Also, what the author emphasizes is that teachers should not underestimate the role of
correction of written work and should know how to provide feedback on writing. Finally,
speaking seems to be the most important skill, and it would not be an exaggeration to say that
most foreign language learners are primarily interested in learning to speak. Evidently, speaking
shows differences from writing, and this implies different reasons for teaching this skill and
different types of exercises, which, in their turn, focus on other aspects of language and demand
different levels of correctness.
Chapter 4, Teaching vocabulary (pp. 220-258), reinforces the importance of teaching
vocabulary and reconsiders its once marginal status relative to the main purpose of language
learning: the acquisition of grammatical knowledge about the language. And this should come as
no surpise, as vocabulary expansion is one of the central aspects learners express interest in. The
chapter talks about ways and techniques of memorizing and remembering vocabulary, the criteria
that should be taken into account when selecting vocabulary, and what students need to know
when it comes to vocabulary teaching. The core of the chapter is based on the pattern we are
already familiar with: the three important stages of teaching vocabulary are presentation,
discovery techniques, and practice.
Finally, the response to the well-known question whether literature should be included in
the course or not is affirmative, as Chapter 5, Teaching literature (pp. 259-287), emphasizes the
fact that teaching and learning literature is very important. Some of the reasons why we should be
in favour of including literature in courses are the following: (i) it arouses and maintains students
interest in literature and contributes to enjoyment; (ii) it helps them explore their own responses
to literature; (iii) it has intrinsic educational and aesthetic value; (iv) last but certainly not least, it
is a rich source of language. To give only one example, in order to show that language can be
taught through literature, the author gives the example of the novel Mrs. Dalloway, in which
Virginia Woolf makes use of free indirect discourse.
148
Generally speaking, this is a well structured book. It provides an excellent overview for
teachers who need some background about terminology and methods that is not too technical.
Attached to each chapter is a special section entitled Checking understanding, which
presents a wide variety of communicative activities and seminar worksheets. I consider that they
may be used by anyone looking for new ideas or teaching resources.
We witness a student-friendly approach and a student-centered way of explaining the
methodology and it is precisely this approach that I liked best. Let me mention only one example
related to teaching listening. The author first mentions (pp. 174-176) two (negative) versions of
the same lesson plan, from which she summarizes the most important things to remember when
devising a listening activity. Then, she invites her readers to reconsider the lesson procedures and
redesign them so as to include the newly-discovered features of teaching listening. She concludes
by extracting some general principles of teaching this integrated skill.
The authors style of presenting the material is easy to read and she speaks directly to her
readers. Take a look at the following examples: What I want to make you understand is that
(p. 120), What would you do to change that? (p. 122), In order for you to see the importance
of language planning, I (p. 123), Before establishing together some of the most important
principles (p. 173), etc. Readers are invited to answer questions, think about possible
approaches and solutions to various problems or difficult situations, they are encouraged to
elaborate upon their reasons, motivate their standpoint, select from the multitude of activities, or
design their own activities. The author provides very many examples and there are a lot of
practice exercises.
In the Introduction, the author mentions that this book can be used as a textbook meant
to teach future teachers about English taught as a foreign language (p. 13). I consider that this is
an underestimation: the present book, precisely because it is both a scientific and a
methodological approach to teaching English as a foreign language, is a useful tool not only for
future teachers, but also for trainee or novice teachers, and some of its material may also be found
interesting and helpful by more experienced practitioners.
IMOLA-GNES FARKAS6
Isidor de Sevilla, Etimologii XI-XII, edicin bilinge, traduccin del latn

al rumano, estudio introductorio, cronologa y notas por Anca Criv, Polirom,
Iai, 2014, 280 p.
El presente volumen es una edicin bilinge de las Etymologiae (Originum sive
etymologiarum libri viginti) escritas por San Isidoro, obispo de Sevilla en 627630 A.D. El
original latino viene acompaado por su traduccin al rumano de dos de los libros de las
Etimologas: Liber XI De homine et portentis (Libro XI El hombre y los monstruos) y Liber
XII De animalibus (Libro XII Las bestias). La estructura del volumen se integra del corpus de
textos mencionado anteriormente, contextualizado por un estudio y una cronologa emplazados a
principios del libro y seguido por unas notas en el apartado posterior de ste.
La edicin en su conjunto resulta impresionantemente generosa en ofrecer informaciones
minuciosas tanto desde el punto de vista cuantitativo como cualitativo, manteniendo una
6
Babe-Bolyai University, Faculty of Letters, Department of English Language and

Literature, farkas.imola.agnes@gmail.com.
149
estructura detallada pero sin embargo muy clara, casi matemtica. El estudio introductorio ofrece
antes que nada detalles sobre el contexto externo a las Etimologas isidorianas que influy en su
creacin, para que posteriormente sumergiera en las entraas de la estructura interior de la obra
explicando el hilo de su filosofa. Sus apartados siguientes tratan primero de sus fuentes y luego
de las influencias que tuvo posteriormente en otros escritos, para que el final se dedicara a unos
ejemplos prcticos de etimologas isidorianas, enfocando en los ms impactantes e ingeniosas.
Con vistas a demostrar desde el principio el hecho de que a San Isidoro no se le poda
pasar por alto en cualquier enumeracin de los espritus cientficos de la Edad Media, Anca Criv
parte en su estudio de una hermosa imagen del Paraso de Dante, la cual trata del momento en
que el poeta florentino viene acompaado por Beatrice hacia el cuarto cielo, el del Sol. Ah los
espritus felices de quienes se hubieran dedicado la vida al estudio y al conocimiento se mueven
en crculos de luces brillantes, al son de una cancin de una harmona perfecta. Se encuentran all
las ms importantes personalidades: Toms de Aquino, Alberto Magno, Graciano, Pedro
Lombardo, el sabio rey Salomn, Dionisio Areopagita, Paulo Orosio, Boecio, Isidoro de Sevilla,
San Beda el Venerable, Ricardo de San Victor y Siger de Brabante.
A continuacin nos explica la autora que la inclusin de San Isidoro en esta encantadora
imagen dantesca viene justificada por el xito que debi de tener sus escritos durante los siglos
VIIXV, entre los cuales el estudio introductorio destaca un tratado de dogmtica y moral
cristiana, Sententiae/ De summo bono, considerado como una Summa theologica del siglo VII y
un escrito de carcter mstico, Synonyma/ Soliloquia con el subttulo De lamentatione animae
peccatricis. La autora tambin muestra que a San Isidoro se lo conocieron como histrico y
naturalista, y enfoca sobre todo en su importante papel como enciclopedista, debido a su obra de
proporciones impresionantes, las Etymologiae sive Origines. Aade que sta fue una de las ms
conocidas, copiadas y citadas obras medievales, que, hasta que se public la edicin princeps en
Augsburgo en 1472 por Gnter Zainer de Reutlinger, haba circulado entre los siglos VII y XVI
en ms de mil copias manuscritas.
Con respeto a las fuentes de las Etimologas, el estudio establece las auctoritates para San
Isidoro, quien practica el humanismo cristiano enseado por las lecciones de San Agustn. As, en
primer lugar estn los escritos de San Ambrosio, San Jernimo y San Agustn de Hipona, el
ltimo siendo el ms importante por encontrarse citas agustinianas a cada paso en las
Etimologas. En segundo lugar los autores cristianos, Casiodoro y Gregorio Magno, y en el
tercero, los escritores de la Antigedad pagana.
En lo relativo a las influencias que tuvo en las producciones posteriores, la enciclopedia
de San Isidoro se configur como el fundamento de todas las enciclopedias medievales, sobre
todo durante el siglo XIII, la edad de oro del enciclopedismo medieval, de los lapidarios y los
bestiarios, como por ejemplo lo fue el Fisilogo latino B, que ms tarde se convirti en la fuente
de los bestiarios franceses.
En lo que concierne al contexto de la obra, la autora apunta datos significativos sobre los
textos que informan de los escritos de San Isidoro comenzando con el siglo VII y explica la
naturaleza que el concepto de enciclopedismo poda tener en aquella poca. Anca Criv advierte
que, aunque el trmino enciclopedia en el sentido que hoy conocemos y utilizamos se consagra
despus del ao 1559, se pueden mencionar como ejemplos de obras de tipo enciclopdico
anteriores a las Etimologas de San Isidoro, Disciplinarium libri de Marco Terencio Varrn,
Naturalis Historia de Plinio el Viejo, Prata de Suetonio, Nocte satticae de Aulo Gellio y
Saturnalia de Macrobio. Enfatiza que la enciclopedia pagana que instituye las siete artes liberales
como esqueleto fundamental del sistema educacional medieval es el tratado escrito por Marciano
Capella en el siglo V, De nuptiis Philologiae et Mercurii. Este tratado representa la boda del dios
de la Retrica, Mercurio, con la virgen Filologa, quien recibe como regalo a las siete artes
liberales: la gramtica, la dialctica y la retrica el trivium y la aritmtica, la geometra, la
msica y la astronoma el quadrivium.
150
Este conjunto de disciplinas que haban formado en la Antigedad la
(enkkliospaidea la cultura general) se conserv a lo largo de la Edad Media a travs del canon
educacional del trivium y el quadrivium que permite no solamente la conservacin de la cultura
antigua, sino ms que nada la asimilacin de sta como propedutica sine qua non para el
entendimiento de las sagradas escrituras. Esto se da sobre todo comenzando con la poca y los
escritos de San Agustn de Hipona, quien se remonta en la tradicin de Filn de Alejandra en
virtud de la cual Moiss y los profetas no fueron sino predecesores de los filsofos antiguos.
Como ya se conoce, fue San Agustn quien promulg la idea de que anteriormente a la
Encarnacin, Dios les haba hablado a los hombres por medio de los filsofos.
La autora describe el trayecto del fenmeno de la cristianizacin del saber antiguo desde
el inicial rechazo de las artes liberales por parte de los padres de la iglesia hasta el momento en
que el tratado de San Agustn, De doctrina christiana, ofrece el programa esencial de la cultura
eclesistica en una sntesis del cristianismo y la cultura griega y latina, a la vez que consagra la
idea de la philosophia ancilla theologiae, es decir el estatuto ancilar de la filosofa antigua frente
al conocimiento de la Sagrada Escritura y al acercamiento a Dios.
Todo ello explica, por ejemplo, cmo fue posible que aproximadamente en el ao 552
Casiodoro escribiera Institutiones divinarum et saecularium litterarum, un programa de estudio
para los monjes del monasterio de Vivarium, que se fundara en los siete artes liberales.
La siguiente parte del estudio ofrece datos y explicaciones amplias y detalladas sobre el
escrito en frente del cual estamos, las Etymologiae, desentraando su estructura y filosofa. Los
datos se refieren al ttulo original, a la datacin de la obra, al hecho de que inicialmente se dedic
al monarca visigodo Flavio Sisebuto, quien falleci el ao 621. Tambin nos proporciona Anca
Criv informaciones sobre la primera mencin de las Etimologas por Braulio, obispo de
Zaragoza y discpulo de San Isidoro, en su trabajo, Renotatio. Braulio explica la divisin de las
Etimologas en libros, y adems, advierte la autora, por ofrecer tanto indicios obvios sobre los
conocimientos comprendidos en ellos, como indicaciones sobre la modalidad de organizacin de
este contenido de conocimientos, da razn de la concepcin isidoriana de tales conocimientos y
de tal modalidad de la organizacin de stos. Harto fundamento para que el estudio ofrezca
explicaciones sobre el sistema organizativo de la materia en 20 libros, para el anlisis del modo
de encadenamiento de los captulos, de los puentes que dan coherencia al conjunto y de las
referencias encrucijadas. La conclusin de este apartado mantiene que las Etimologas son de
hecho un inmenso base de datos del siglo VII, cuya coherencia se debe al mtodo explicativo
empleado sistemticamente por San Isidoro: la etimologa.
Para que lector se percatara de tal mtodo, enumeramos a continuacin unos cuantos
ejemplos de etimologas de San Isidoro. El Libro XI. El hombre y los monstruos, Captulo 1. Del
hombre y sus partes empieza con una de las etimologas confirmadas por la filologa moderna:
homo ex humo. sta viene como motivo para recalcar la construccin binaria del hombre, duplex
est homo, interior et exterior, la superioridad frente a los animales por un lado y su propensin
hacia la contemplatio mundi, por el otro. Los sentidos, sensus, hacen que el alma, anima, mueva
el cuerpo entero a travs de la fuerza de las sensaciones. Las cosas presentes, praesentia, se
localizan frente a los sentidos, praesensibus, as como lo que est en frente de los ojos se halla en
la vista. Las lgrimas, lacrimas, se refieren a laceratio mentis laceracin de la mente. Los
hinojos, genua, se llaman as porque en el vientre materno se hallan cara a cara con los prpados,
genae, por lo cual tienen rasgos en comn con los ojos. Ello explica como los hombres, nada ms
se prosternen de hinojos, que inmediatamente empiezan a llorar, puesto que la naturaleza, por
haber dispuesto a los hinojos aplastantes en las mejillas, quiso recordar que antes de que salieran
a la luz, se haban encontrado juntos en la oscuridad.
Siguiendo a San Agustn, De trinitate 4.4.7, San Isidoro enumera en el segundo captulo
del mismo libro a las seis edades del hombre. Anota la autora del estudio que los est conectando
al tiempo grande de los seis das de la Creacin segn la Biblia y a los seis perodos de la
historia de la humanidad: de Adn a No, de No a Abraham, de Abraham a David, de David a la
151
esclavitud babilnica; de sta a Cristo, y el sexto perodo despus de Cristo. De esta forma el
hombre atraviesa la infancia, siendo infans mientras no sabe hablar y fari hasta los siete aos, la
niez, pueritia, es decir la edad pura hasta los 14 aos, la adolescencia hasta los 28, la juventud es
la ms poderosa de todas hasta los 50, la madurez hasta los 70, y la vejez, cuya ltima parte se
llama senium. Difunto, defunctus, de llama el que ya haya concluido su oficio de vida,
functiofficio, y sepultado, sepultus, el que ya no tiene pulso, sine pulsu.
El captulo tercero trata de los monstruos (portenta, ostenta, monstra y prodigia) que se
llaman as porque parecen prever (portendere), mostrar (ostendere), demostrar (mostrare) y
predecir (praedicere) los hechos del porvenir. De los escipodas llamados en griego skiopodas se
conoce que vivan en Etiopa y que slo tienen un pie, que, al alzrselo en tiempo de mucho calor,
le sirven a estos etiopios para hacerse sombra. Tambin se imagina que hubo tres sirenas, de
cuerpo mitad virgen, mitad ave, que tenan alas y garras. Entre ellas, una cantaba, otra tocaba la
flauta y la tercera la lira. Ellas hacan que los marineros naufragaran al ser hechizados por su
msica. San Isidoro explica que de hecho se trataba de tres prostitutas de los que, habiendo dejado
a los pasajeros pobres, se dijo que les hacan naufragar. Las alas y las garras eran smbolos del
dios Amor quien vuela y hierre a la vez y tambin se dijo de ellas que vivan en las olas, porque
Venus haba nacido de las olas.
El cuarto captulo, sobre las metamorfosis, da cuenta entre otras de las transformaciones
de varios cuerpos y de la consecuente aparicin de varios tipos de formas de vida. Por ejemplo, de
la carne ptrida de vaca nacen las abejas, de la de caballo los escarabajos, de la del asno los
saltamontes y de la del cangrejo los escorpiones.
El Segundo libro de las Etimologas incluido en el presente volumen es el XII y trata de
las bestias. En el primer captulo hay explicaciones sobre el cordero, agnus, nombre que viene de
la palabra griega hagns puro, que equivale a santo. Explica San Isidoro que los latinos
consideran que lleva este nombre porque reconoce, agnoscere, a su madre entre los otros
animales an en una manada numerosa por su manera de balar. Al otro extremo, el chivo, hircus,
es un animal lascivo, sin vergenza, que mira siempre en diagonal, y de ah su nombre, porque
hirqui se usa para designar las partes laterales de los ojos. Su natura es tan caliente, que incluso el
diamante, que ni el fuego ni el hierro lo pueden romper, se disuelve solamente en sangre de chivo.
Del segundo captulo citamos las explicaciones sobre el nombre del oso, ursus, que
proviene de su hbito de dar forma a su cachorro con su propia boca, ore suo, y de ah orsus
significa comienzo. Del sptimo, que trata de los pjaros, aprendemos que las aves se llaman as
porque no siguen vas, viae, ciertas, sino que se desplazan en direcciones no determinadas, avia.
Tambin la grajilla, monedula, tiene su nombre de monetula, porque una vez que encuentra oro,
se lo lleva y lo esconde, por lo cual se le conoce como que roba el oro. El cuervo, corvus, corax,
tiene su nombre del sonido de su voz, coracinare. Gregorio Magno cuenta en Moralia que este
pjaro alimenta suficientemente a sus hijos slo despus de que les hubiera crecido el plumaje
negro oscuro tpico, de modo similar al predicador que se niega a nutrir a sus discpulos con las
verdades sutiles de la fe hasta que stos no se hubieran convertido conscientes de la negrura de
sus pecados.
Para entender mejor y contextualizar tal inditas e interesantes explicaciones de palabras
y conceptos, volvemos a la teora sobre la ciencia de la etimologa y encontramos en el estudio
introductorio la clasificacin de Jacques Fontaine, que la autora utiliza para desarrollar las seis
tradiciones que configuran este tipo de reflexin lingstica: la etimologa popular, la tradicin
gramatical, la retrica, la dialctica, la filosfica o inicitica derivada de la conviccin neopitagrica de que ciertas prcticas etimolgicas son capaces de desvelar la verdadera naturaleza
de la realidad, y la sexta, que surge de la exegesis cristiana, y a la que se denomina compleja
por englobar la tradicin hebraica de interpretacin de los nombres bblicos.
Entre stas, la visin de San Isidoro parece emplear a la etimologa como el mtodo para
definir el valor autntico de las palabras, o bien su cara verdadera: vis verborum. sta ltima se
establece a partir de su origen, pero a la vez permite el conocimiento de la realidad por medio del
152
origen de las palabras que designan. Anca Criv advierte que el obispo sevillano apoya el viejo
debate a encontrar tambin en el dialogo platnico Cratylos, el de la coexistencia de los dos tipos
de etimologa, la secundum naturam y la por convencin arbitraria, una coexistencia que no
permite el desvelo de la realidad entera por medio de las palabras.
Adems, en base de la clasificacin de las tradiciones etimolgicas hecha por Fontaine, la
autora del estudio considera que San Isidoro emprende un mtodo gramatical etimolgico, dado
que la gramtica es la ciencia piloto que facilita el acceso a los textos fuentes del saber. Este
mtodo no solamente conserva el valor metafsico que tuvo en ciertos corrientes de pensamiento
de la Antigedad, sino que an ms, hay que darse clara cuenta de que se trata de unos
conocimientos provenidos de los autores antiguos prestigiosos, considerados unas auctoritates
durante la Edad Media.
Con referencia a San Isidoro y en apoyo de la importancia de su impactante obra que
apunta a la transmisin del saber, Anca Criv puntualiza no sin cierta emocin que decir que los
antiguos hablan por medio de tu persona significa asumir conscientemente la posicin de alguien
que se siente capaz de llevar el conocimiento adonde est en peligro de desaparecer
definitivamente y de salvar, en medio de un mundo de guerreros que amenazan a caer en la
barbarie, los valores de una cultura que muchos ya no pueden entender.
SILVIA TEFAN*
Valerie Pellatt, Eric T. Liu, and Yalta Ya-Yun Chen, Translating Chinese
Culture. The Process of Chinese-English Translation, Routledge, London and
New York, 2014, viii + 191 pp.
The book discussed here comes as a natural continuation of Thinking Chinese
Translation. A Course in Translation Method: Chinese to English7, where the authors touched
upon many of the problems young translators encounter when dealing with Chinese texts, both
linguistically and culturally. The present book goes a step further and explores the very specific
challenges young translators face when dealing with texts which are intimately related to the
Chinese culture, such as writing or calligraphy, clothing or traditional painting, nursery rhymes or
poetry.
One of the most common remarks one hears when it comes to Chinese culture is about its
uniqueness, about how Chinese culture has got an intimate structure which is very hard for an
outsider to understand and virtually impossible to translate. It is this kind of attitude that the book
fights against and the authors are quite successful in showing that translating Chinese culture for a
Western audience is not necessarily doomed to fail. For this, however, the translator must pay
attention to many extra-linguistic elements, starting with the similarities and differences between
the cultures involved, the background of the source text and the author, the target audience, etc.
The nine chapters of the book, each of them dealing with a specific area of expertise,
could be loosely grouped around visual or performing arts. Each chapter follows the same pattern,
*
Universidad de Bucarest, Departamento de Lingistica Romnica, Lenguas y Literaturas
Iberoromnicas e Italiano, silvia.stefan@lls.unibuc.ro.
7
Valerie Pellatt and Eric T. Liu (2010), Thinking Chinese Translation: A Course in
Translation Method: Chinese to English, Routledge, London and New York.
153
starting with an introduction of the area of expertise, followed by case-studies organized
hierarchically, from the less to the more complex ones, and finish with practical exercises. The
authors do not forget to provide the reader with possible translations for the examples and
exercises in the book, with linguistic analysis of the choices they make, or discussions regarding
various methods of translation. They pay very much attention to the cultural background of the
texts chosen as examples and the extra-linguistic elements that influence the process of
translation, helping the young translator get a better understanding of their work and avoid a
common mistake made by many inexperienced translators, that the translated text must be
linguistically as close as possible to the original text. The book can help students in translation
find the fine line between excessive domestication or foreignization of a text.
Most of the areas of expertise chosen for discussion are familiar to all students in
Chinese: calligraphy, painting, poetry, drama, etc. There are also elements which might not be
covered by a special course, in many universities, such as clothing, or nursery rhymes, but this
does not make them less important, especially because it is highly probable to encounter them
when translating Chinese literature. Most of the areas covered in the book have their own
specialized vocabulary and, just as the authors observed, most of the times, this vocabulary is not
translated, especially when appearing in texts destined to people familiar with the Chinese
culture, who are supposed to understand the vocabulary as such. But what happens when the text
is destined to a larger audience? How should these terms be translated when they refer to a type of
reality inexistent in the target culture? The chapters dedicated to calligraphy and clothing are very
good examples of how to deal with this type of texts.
Dedicated to calligraphy, the third chapter shows not only the difficulty of translating the
vocabulary specific to this art form, but also touches upon the different ways of organizing a text
in Chinese and English. There are many Chinese authors who feel that, when writing about
traditional Chinese culture, one needs to use long elaborate sentences with structures borrowed
from the classical language. The translator needs to show a lot of skill in order to preserve the
flavor of the text, by choosing the right register, and, at the same time, to come up with a coherent
and natural text in the target language. It is also the translator who needs to decide how to
translate the specific terms and how much information his/ her translation should contain in order
to make the text as informative as needed.
Similar problems pose the translation of clothing, as shown in the fourth chapter. The way
people dress, especially in a highly hierarchized society such as imperial China, was extremely
important since it said a lot about the status a person enjoyed in the society. How can a translator
use words to make the reader visualize the clothes characters wear, especially when transliteration
or contextualization does not work the magic? The answer given by the authors is semantic
fields hyponyms or hyperonyms. Good mastery of the semantic fields, corroborated with
knowledge about dressing culture, can help the translator make the right choice based on the
function and the appearance of the term to be translated.
The need for a translator to be bold and daring is shown in the sixth chapter dealing with
translation of poetry. Translating poetry is notoriously difficult, because, more than in any other
case, it involves both the content and the form. By looking at the process of translating poetry as
transcreation, the authors encourage the translator to use his/ her creativity and spontaneity and
became aware of the fact that classical Chinese poetry would be extremely difficult to be
translated from classical into modern Chinese, let alone a foreign language. There are many
instances where translators tried their best to recreate the poem either by employing the same
composition rules as in Chinese, or by imitating well-known classical poems or poets in the
source culture. Needless to say that, most of the times, the results are disappointing. The
translator needs to reach an audience which is absolutely different from the one for which the
poem was composed, therefore the poem needs to be translated in such a way that it reaches the
new audience and it allows for them to identify with it.
154
Translation as a collaborative project is discussed in the eighth chapter, dedicated to
drama translation. Collective translation was a very common phenomenon in China, where most
of the Buddhist sutras in the local language are the result of the corroborated efforts of a large
team of monks. Due to the special destine of the text to be staged in front of the audience, there
is a lot of discussion regarding drama translation, whether it should be more literal or more
performance-oriented, how domesticated the text should be to ensure that it reaches the audience,
what role the director, or of the actors, should play in the process of translation. The two case
studies described in this chapter, the translation of Cao Yus Peking Man and Wang Fangs
Poison, might be ideal cases where the team involved in producing and staging the play also acts
as translator. Thus, the students involved in the process became much more aware of the
difficulty of their task, when they needed to take into account not only the text, but also to
accommodate their colleagues understanding of the plot and anticipate the potential problems
related to staging the text. It is rarely the case that the team acting the play be also the one that
translates it. Nevertheless, the chapter emphasizes the importance of collaborative work when it
comes to translating drama, the fact that people who might have no special knowledge of the
source culture, such as the director or the actors, can actually contribute to a successful
translation.
Translating Chinese Culture. The Process of Chinese-English Translation. is a rich book
which gives food for thought not only to young translators from Chinese, but also to other people
interested in the Chinese culture. It might seem that the authors overemphasize the introductory
part of some of the chapters, paying less attention to the concrete text analysis than to the cultural
background, but from the point of view of someone involved in teaching students how to deal
with a Chinese text, I can understand that emphasizing the role of culture to a successful
translation never seems enough. After all, to quote Bassnett (2002: 23)8, language is the heart
within the body of culture, and it is the interaction between the two that results in the continuation
of life-energy. In the same way that the surgeon, operating on the heart, cannot neglect the body
that surrounds it, so the translator treats the text in isolation from the culture at his peril.
MUGUR ZLOTEA*
York.
Susan Bassnett (2002), Translation Studies, 3rd edition, Routledge, London and New
University of Bucharest, Faculty of Foreign Languages and Literatures, Department of

Oriental Languages and Literatures, mugur.zlotea@g.unibuc.ro.
CONTRBUTORS
ANDREI A. AVRAM is Professor of English Linguistics at the University of Bucharest.

He holds a PhD in linguistics from the Iorgu Iordan Al. Rosetti Institute of Linguistics of the
Romanian Academy, Bucharest (2000), and a PhD in linguistics from Lancaster University
(2004). His research areas are pidgins and creoles, language contacts, and phonology. His
publications include Pidginurile i creolele cu baz englez i francez ca tip particular de
contact lingvistic (2000, Bucharest), On the Syllable Structure of English Pidgins and Creoles
(2005, Bucharest), Fonologia limbii japoneze contemporane (2005, Bucharest), and articles in
English World-Wide, English Today, Journal of Language Contact, Cochlear Implants
International, Linguistics in the Netherlands, Acta Linguistica Hafniensia, Lengua y migracn,
Papia, Linguistica Atlantica, Acta Linguistica Asiatica, Romano-Arabica, Revue roumaine de
linguistique.
DANIELA BORDEA is a graduate of the Faculty of Chemistry of the University of
Bucharest (1980) and of the Faculty of Orthodox Theology, Department of Theology - French
(1996), obtaining a PhD in chemistry (2002) and another in philology (2011). The title of her PhD
dissertation in philology is LAdjectif, du figement laffranchissement. Modle mathmatique
tridimensionnel. She is currently Lecturer at the Department of Modern Languages (specializing
in French), Faculty of Foreign Languages and Literatures, University of Bucharest. Her areas of
interest are linguistics and mathematical notions applied to French linguistics.
GABRIELA ANIDORA BROZB defended her PhD dissertation in 2011. She teaches
English phonetics and phonology as well as pragmatics at the University of Bucharest, and ESP
(mainly English for business and economics) at the Romanian-American University in Bucharest.
Her research interests are phonetics and phonology, non-native varieties of English, and
sociolinguistics. Her publications include Between Reality and Myth: A Corpus-based Analysis of
the Stereotypic Image of Some Romanian Ethnic Minorities (2010, Munich), The Phonology of
New Englishes (2012, Bucharest), articles in Bucharest Working Papers in Linguistics, Analele
Universitii din Bucureti, Analele Universitii Ovidius.
ALINA BUGHEIU received her PhD from the University of the West in Timioara.
She is currently a research assistant at the North University Centre of Baia Mare (Technical
University of Cluj-Napoca). Her main research interest is in the field of onomastics, especially in
relation to commercial names and unconventional anthroponyms, such as nicknames and user
names. Her publications include Once upon a Time in Angela Carters Magic Toyshop: Fairy
Tales, Myths, and the Sixties (2012, Saarbrcken) and articles in The Journal of Linguistic and
Intercultural Education JoLIE, British and American Studies, Analele Universitii din
Bucureti, Caiet de semiotic. She also edited, together with Ovidiu Felecan, Onomastics in
Contemporary Public Space (2013, Newcastle upon Tyne).
MARIA AURELIA COTFAS received her PhD from the University of Bucharest in
2012. The title of her PhD dissertation is On the Syntax of the Romanian Subjunctive: Control
and Obviation. She is Lecturer in the Department of English of the University of Bucharest. Her
research area is generative syntax. Her publications include articles in Bucharest Working Papers
in Linguistics, Analele Universitii din Bucureti.
156
JING DENG is Lecturer of English and Linguistics at Nanjing University of Science and
Technology in Nanjing. She is currently conducting post-doctoral research at Fudan University in
Shanghai. Her research focuses on cross-cultural pragmatics, pragmatics of the Chinese language,
the grammar-pragmatics and sociolinguistics interface. Her publications include articles in
Linguistica Atlantica and Bucharest Working Papers in Linguistics.
SABINA POPRLAN is Lecturer at the University of Bucharest, Department of
Oriental Languages and Literatures, where she teaches courses on Hindi language and
culture. She received her PhD in philology in 2009 from the University of Bucharest. Her
research interests focus on Hindi linguistics, mostly from a typological perspective and
comparing Hindi to French and Romanian, but also on Indian culture. Her publications
include LAnaphore verbale. Domaine typologique: franais, hindi, roumain (2010,
Bucharest) and articles in Revue roumaine de linguistique, Analele Universitii din
Bucureti.
FRANCESCO VITUCCI is Professor of Japanese Language and Linguistics and
Japanese Philology at the School of Foreign Languages and Literature, Interpreting and
Translation and of the Alma Mater Studiorum University of Bologna and teaches Japanese
Language at the Department of Asian and Mediterranean Africa Studies of Ca Foscari
University, Venice. His research interests are the multimedia teaching of modern Japanese, with
particular attention to audiovisual translation. His publications include Nihon JP-1 (2007,
Bologna), Nihon JP-2 (2010, Bologna), Eserciziario Orale di Giapponese Moderno (2009,
Bologna), Il Giapponese per viaggiatori (2011, Bologna), the monograph La didattica del
giapponese attraverso la rete - Teoria e pratica glottodidattica degli audiovisivi (2013, Bologna)
as well as numerous articles in international journals.
ANALELE UNIVERSITII BUCURETI (AUB)

LIMBI I LITERATURI STRINE
N ATENIA COLABORATORILOR
Pentru o cooperare eficient ntre editori, autori i casa editorial, autorii de
articole i de recenzii sunt rugai s respecte urmtoarele norme:
Articolele pot fi trimise n englez, francez, romn, italian, spaniol,
german.
Articolele trebuie s fie trimise pe suport electronic (e-mail sau CD) n
format WORD (.doc or .rtf).
Articolele trimise trebuie s conin numele i afilierea instituional a
autorilor, ca i adresa de e-mail. Autorii sunt rugai s predea i o scurt
prezentare auto-bio-bibliografic (cca. 10-15 rnduri).
Articolele trebuie s fie nsoite de un rezumat (10-15 rnduri), urmat de 5-7
cuvinte-cheie, ambele n englez (font Times New Roman, corp 9, la un rnd).
Toate articolele i recenziile vor fi redactate cu diacritice; dac sunt
folosite fonturi speciale (Fonetic, ArborWin etc.), se va trimite i tipul de font
folosit.
Formatul documentului: pagin A4 (nu Letter, Executive, A5 etc.).
Marginile paginii: sus 5,75 cm; jos 5 cm ; stnga i dreapta
4,25 cm ; antet 4,75 cm; subsol 1,25 cm.
Articolele trimise trebuie tehno-redactate cu font Times New Roman, corp 11,
la un rnd.
Titlul articolului trebuie s fie centrat, cu majuscule aldine (font
Times New Roman, corp 11).
Numele (cu majuscule aldine) trebuie s fie centrat, sub titlu (font
Times New Roman, corp 11).
Rezumatul (nsoit de titlul articolului tradus, dac articolul este n alt
limb dect engleza) preced textul articolului (font Times New Roman, corp 9,
la un rnd); cuvintele-cheie (Times New Roman, corp 9, italic) urmeaz
rezumatului.
Notele trebuie s apar n josul paginii (cu font Times New Roman, 9, la
158
un rnd).
Trimiterile bibliografice, indicarea sursei pentru citate se vor indica
n text, dup urmtoarea convenie: (Autor an:(spaiu)pagin) (Pop 2001: 32);
(Pop/Ionescu 2001: 32).
Se pot utiliza n text abrevieri, sigle (SMCF, vol. II, p. 20) care vor fi
ntregite la bibliografia final, dup cum urmeaz:
SMCF Studii i cercetri privitoare la formarea cuvintelor n limba
romn, vol. II, Bucureti, Editura Academiei Romne, 1961.
LR Limba romn etc....
RITL Revista de istorie i teorie literar etc.... RRL Revue roumaine de
linguistique
Bibliografia va fi indicat dup urmtorul model:
(1) Pentru cri, volume, monografii se indic numele autorului,
prenumele prescurtat, anul apariiei, titlul cu italic, editura, oraul (eventual volumul
sau numrul de volume). n cazul n care una dintre componentele trimiterii
bibliografice lipsete, se vor folosi normele consacrate [s.l.], [s.a.]. La
volumele colective se va indica ndrumtorul/ coordonatorul/ editorul prin
(coord.) sau (ed.)/ (eds.) dup nume i prenume. n cazul n care exist mai
muli autori/ coordonatori/ editori, doar primul nume va fi inversat (Zafiu, R., C.
Stan...).
Coteanu, I., 1982, Gramatica de baz a limbii romne, Editura Albatros,
Bucureti, Riegel, M., J.-C. Pellat, R. Rioul, 1999, Grammaire mthodique du
franais, Presses
Universitaires de France, Paris.
Zafiu, R., C. Stan, Al. Nicolae (eds.), 2007, Studii lingvistice. Omagiu
profesoarei Gabriela Pan Dindelegan, la aniversare, Editura Universitii din
Bucureti, Bucureti.
(2) Pentru articole din volume colective se indic numele autorului,
prenume, an, titlu ntre ghilimele, urmat de in + prenume (prescurtat), numele
editorului/ editorilor (ed./ eds.), titlul volumului n italice, editura, oraul, pagini
Zamboni, A., 1998, Cambiamento di lingua o cambiamento di sistema?
Per un bilancio cronologico della transizione, in J. Herman (ed.), La tranzitione
dal latino alle lingue romanze. Atti della Tavola Rotonda di Linguistica
Storica, Universit CaFoscari di Venezia, 14-15 giugno 1996, Tbingen,
Niemeyer, pp. 99-127.
Portine, H., 2012, De la synonymie la reformulation , in S. Berbinski,
D. Dobre, A. Velicu (ds.), Langages(s) et traduction, Editura Universitii din
Bucureti, Bucureti, pp. 47-62
159
(3) Pentru articole din reviste se indic numele autorului, prenumele
autorului, anul, titlul articolului ntre ghilimele, urmat de in + numele revistei
cu italic (neabreviat), volumul/ tomul, numrul, pagini. n cazul n care exist mai
muli autori, doar primul nume va fi inversat.
Fischer, I., 1968, Remarques sur le traitement de la diphtongue au en latin
vulgaire , n Revue Roumaine de Linguistique, XIII, nr. 5, pp. 417-420.
Cornea, P., 1994, Noiunea de autor: statut i mod de folosin, n Limb i
literatur, vol. III-IV, pp. 27-35.
Sorea, D., A. Stoica, 2011, Linguistic Approaches to Verbal and Visual
Puns, in Analele Universitii Bucureti. Limbi i Literaturi Strine, anul LX2011, nr. 1, pp. 111-127.
Toate referinele bibliografice din text trebuie s apar n bibliografia final
; pentru mai multe detalii despre normele de editare (Guidelines for authors),
se poate consulta adresa : http://www.unibuc.ro/anale_ub/limbi/index.php
Articolele trimise vor fi discutate de o comisie de specialiti n domenii
filologice: lingvistic, literatur, studii culturale, studii de traductologie.
Articolele trebuie trimise la urmtoarele adrese de e-mail:
sabina_ioana@yahoo.com; ruxandra.visan@lls.unibuc.ro.
THE ANNALS OF THE UNIVERSITY OF BUCHAREST

FOREIGN LANGUAGES AND LITERATURES
NOTES FOR CONTRIBUTORS

The authors of the articles and book reviews are requested to observe
the following publication guidelines:
The articles can be edited in English, French, Romanian, Italian, Spanish,
German.
The articles should be submitted electronically (by e-mail or CD) in a
WORD format (formats .doc or .rtf).
The articles should contain the authors full name and affiliation,
along with the authors e-mail address. The authors are requested to supply an
auto-bio-bibliographical note (approximately 10-15 lines).
The articles should contain an abstract (10-15 lines), followed by 57 Keywords (Times New Roman, 9, single spaced), both in English.
All the articles and book reviews must be edited using diacritical
marks; if there are special Fonts, these should also be sent.
The page format: paper A4 (no Letter, Executive, A5 etc.);
The page margins: top 5,75 cm; bottom 5 cm; left and right
4,25 cm; header 4,75 cm; footer 1,25 cm.
The articles submitted for publication must be typed single spaced, in
Times New Roman, 11.
The title of the article should be centered, bold, all capitals (Times New
Roman, 11)
The authors name (bold capitals) should be centered, under the title
(Times New Roman, 11).
The abstract (with the translated title, if the article is written in other
language than English; Times New Roman 9, single spaced) precedes the text
of the article; the Keywords (Times New Roman 9, bold) follow the abstract.
The notes should be indicated by superscript numbers in the text and
typed at the bottom of the page (single spaced, Times New Roman 9).
The references or the quotations sources should be indicated in the
text, following the format: (Author year:(space)page) (Pop 2001: 32);
(Pop/Ionescu 2001: 32).
The abbreviations or abbreviated titles (SMCF, vol. II, p. 20) can be used
in the papers; they will be included completely in the listed references at the end of
162
the article, as it follows:
SMCF Studii i cercetri privitoare la formarea cuvintelor n limba
romn, vol. II, Bucureti, Editura Academiei Romne, 1961.
LR Limba romn etc....
RITL Revista de istorie i teorie literar etc....
RRL Revue roumaine de linguistique
The references should observe the following styles:
1. Books Basic Format: Author: last name, first name (only the name of the
first author is inverted), year of publication, Title of Work, publisher, location.
Coteanu, I., 1982, Gramatica de baz a limbii romne, Editura Albatros,
Bucureti. Riegel, M., J.-C. Pellat, R. Rioul, 1999, Grammaire mthodique du
franais, Presses Universitaires de France, Paris.
2. Edited Books Basic Format : last name of the editor, first name,
(ed./ eds.), year of publication, Title of Work, publisher, location (only the
name of the first editor inverted).
Zafiu, R., C. Stan, Al. Nicolae (eds.), 2007, Studii lingvistice. Omagiu
profesoarei Gabriela Pan Dindelegan, la aniversare, Bucureti, Editura
Universitii din Bucureti.
3. Articles or Chapters in Edited Book Basic Format: last name of the author,
first name, year of publication, Title of article/ chapter, in name of the editor/ editors
(ed./ eds.), in Title of Work, publisher, location, pages of chapter.
Zamboni, A., 1998, Cambiamento di lingua o cambiamento di sistema?
Per un bilancio cronologico della transizione, in J. Herman (ed.), La tranzitione
dal latino alle lingue romanze. Atti della Tavola Rotonda di
Linguistica Storica, Universit CaFoscari di Venezia, 14-15 giugno 1996,
Tbingen, Niemeyer, pp. 99-127.
Portine, H., 2012, De la synonymie la reformulation , in S.
Berbinski, D. Dobre, A. Velicu (ds.), Langages(s) et traduction, Editura
Universitii din Bucureti, Bucureti, pp. 47-62
4. Articles in Journals Basic Format: last name of the author, first name
(only the name of the first author is inverted), year of publication, Title of the
article, in Title of Periodical, volume number (issue number), pages.
Fischer, I., 1968, Remarques sur le traitement de la diphtongue au en latin
163
vulgaire , in Revue Roumaine de Linguistique, XIII, nr. 5, pp. 417-420.
Cornea, P., 1994, Noiunea de autor: statut i mod de folosin, n
Limb i literatur, vol. III-IV, pp. 27-35.
Sorea, D., A. Stoica, 2011, Linguistic Approaches to Verbal and Visual
Puns, in Analele Universitii Bucureti. Limbi i Literaturi Strine, anul LX2011, nr. 1, pp. 111-127.
All the bibliographical references should appear in the final bibliography.
For
some
more
details
(Guidelines
for
authors),
visit also
:
http://www.unibuc.ro/anale_ub/limbi/index.php
All the papers will be peer-reviewed by a committee of specialists in different
philological fields: linguistics, literature, cultural studies, translation studies.
The first version of the articles should be submitted to the e-mail
addresses: sabina_ioana@yahoo.com; ruxandra.visan@lls.unibuc.ro.
Tiparul s-a executat sub c-da nr. 823/2015 la

Tipografia Editurii Universitii din Bucureti

05 08 23 29AUB Foreign Languages and Literatures 2014 No. 2 PDF

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

05 08 23 29AUB Foreign Languages and Literatures 2014 No. 2 PDF

Uploaded by

Copyright:

Available Formats

ANALELE

LINGVISTIC / LINGUISTIQUE / LINGUISTICS

GABRIELA ANIDORA BROZB, A Few Remarks on the Vowels of

A FEW REMARKS ON THE VOWELS OF KENYAN ENGLISH

University of Bucharest, Department of English, brozba.anydora@gmail.com.

A FEW REMARKS ON THE VOWELS OF KENYAN ENGLISH

comparison is Received Pronunciation (henceforth RP). The standard of

Figure 1: The vowel systems of SBE, WAfrE and EAfrE

(adapted from Schmied 1991a: 61 and Mutonya 2008: 437)

tend to merge in East African varieties of English towards vocalic

GABRIELA ANIDORA BROZB

Figure 2: The vowel systems of SBE, EAfrE and KenE

A FEW REMARKS ON THE VOWELS OF KENYAN ENGLISH

phonological level, and to show whether there is inter-speaker variation (and

KIT and FLEECE

in KenE (Gusii1, SAA)

GABRIELA ANIDORA BROZB

be explained either as a result of phonetic environment or as an instance of

Figure 4: GOOSE vowel in KenE (Gusii1, SAA)

A FEW REMARKS ON THE VOWELS OF KENYAN ENGLISH

Gusii2 Kikuyu1 Kikuyu2 Kikuyu3 Kiswahili3 Kiswahili8 Luo1 Luo2

Table 1: The KIT and FLEECE vowels in SAA samples of KenE

Attested in Singapore English (see Deterding 2007: 38).

GABRIELA ANIDORA BROZB

Table 2: Mean F1/F2 values of Luo1

They occur rather randomly than at sense phrases.

Figure 5: Monophthongs of Luo1

A FEW REMARKS ON THE VOWELS OF KENYAN ENGLISH

Luo (Dholuo), which belongs to the Nilotic branch of the Nilo-Saharan

Table 3: Mean F1/F2 values of Gusii1

Gusii (Ekegusii), which belongs to the Bantu family of languages, has

GABRIELA ANIDORA BROZB

Figure 6: Monophthongs of Gusii1

A FEW REMARKS ON THE VOWELS OF KENYAN ENGLISH

general, as well as of others, which seem to point to the uniqueness of KenE

(Simo Bobda 2000: 263)

In addition to being cases of vowel epenthesis, examples (1a-b) also illustrate

(Owino 2003: 57)

(Owino 2003: 89-90)

Note also the occurrence of paragoge in (1b, c) as well as in the following

(Schmied 2008: 162)

As is well known, the occurrence of paragoge has been pinpointed as a hallmark

(Simo Bobda 2000: 263)

GABRIELA ANIDORA BROZB

A FEW REMARKS ON THE VOWELS OF KENYAN ENGLISH

SYLLABLE RESTRUCTURING IN EARLY SOLOMON ISLANDS

University of Bucharest, Department of English, andrei2.avram@gmail.com.

SYLLABLE RESTRUCTURING IN EARLY SOLOMON ISLANDS PIDGIN ENGLISH: AN

*ONS/sO: onsets made up of /s/ and an obstruent are disallowed.

See also Avram (2000).

*ONS/sO, MAX-IO, L-ANCHOR >> DEP-IO, CONTIG

1920s tarch (< E starch) starch

Similarly, vowel prothesis occurs once in a sample of bush pidgin (Jourdan

This constitutes a violation of what Kager (1999: 105) calls minimal

SYLLABLE RESTRUCTURING IN EARLY SOLOMON ISLANDS PIDGIN ENGLISH: AN

COPY: add a copy of the etymological vowel

The ranking of the constraints is:

*ONS/sO, MAX-IO, L-ANCHOR, COPY >> DEP-IO, CONTIG

2.2. Reflexes of the /sl/ cluster

For other relevant evidence see section 2.5.

*ONS/sL: onsets made up of /s/ and a liquid are disallowed.

*ONS/sL, MAX-IO, L-ANCHOR >> DEP-IO, CONTIG

Clearly, if ONS/CG outranks GLIDE-VOCALIZATION the glide /w/ surfaces as

ONS/CG, MAX-IO, L-ANCHOR, DEP-IO, CONTIG >> GVOCAL

ONS/CG, MAX-IO, L-ANCHOR, DEP-IO, CONTIG >> GVOCAL