A Redescription of "Romantic Art"

Niklas Luhmann
MLN, Vol. 111, No. 3, German Issue. (Apr., 1996), pp. 506-522.
Stable URL:
http://links.jstor.org/sici?sici=0026-7910%28199604%29111%3A3%3C506%3AARO%22A%3E2.0.CO%3B2-T
MLN is currently published by The Johns Hopkins University Press.

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained
prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in
the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/journals/jhup.html.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

The JSTOR Archive is a trusted digital repository providing for long-term preservation and access to leading academic
journals and scholarly literature from around the world. The Archive is supported by libraries, scholarly societies, publishers,
and foundations. It is an initiative of JSTOR, a not-for-profit organization with a mission to help the scholarly community take
advantage of advances in technology. For more information regarding JSTOR, please contact support@jstor.org.

http://www.jstor.org
Mon Dec 3 12:56:04 2007

A Redescription of "Romantic Art"

w

Niklas Luhmann

A sociologist taking up a theme like "romantic art" should endeavor

to add nothing new to the subject matter itself. Faithfulness to the object is called for-even if only in the ordinary sense of "empirical." In
what follows it is therefore not a matter of competing with literary or
aesthetic inquiry or of offering new interpretations of Romantic texts
or other contemporary works of art. Nor shall I intervene in the broad
discussion bearing on the relationship of Romanticism, and above all
early German Romanticism (Friihromantik),to modern society and its
self-description as "modern";' this discussion is too dependent on
crude evaluations (for example, "irrationalism") and will necessarily
remain controversial as long as the concept of modernity itself remains
controversial. It is, then, not a question of hermeneutics, not a matter
of 'knowing better' in the domain of the critical analysis of art; in fact
it is not even, at least not directly, a question here of a more adequate
understanding of key Romantic concepts such as poetry (Poesie), irony,
arabesque, fragment, criticism. Such may emerge as a byproduct of
our investigation. But disciplinary discourses operate in their own specific recursive networks, with their own intertextualities, their own selffabricated pasts, which, for instance, determine what one has to do in
order to assume the standpoint of second-order observation and to remain intelligible-regardless of whether one continues the discursive
tradition or suggests particular changes. And, as is well-known, it is dif-

See Karl Heinz Bohrer, Die Kritik der Romantik, Frankfurt 1989
MLN, 111 (1996):506522 O 1996 by The Johns Hopkins University Press

ficult (if not impossible, at least dependent on coincidences hard to

foresee) to intervene in disciplines from the outside in the name of interdisciplinarity2
This should be emphasized in advance when, as here, it is a matter
of redescribing with systems-theoretical instruments what happened
when Romanticism discovered its own autonomy and realized and
worked through what had already taken place historically, namely the
social differentiation of a functional system specifically related to
art."here is a considerable literature bearing on this development, a
literature that takes as its point of departure the notion that the specific character of Romanticism as well as subsequent reflections of art
is conditioned by the reorganization of society along the lines of funcIf Romanticism was modern and still is, then
tional differentiati~n.~
not because it preferred the "hovering" (das "Schwebende") or the "irrational" or the "fantastic," but because it attempts to endure system
autonomy. Up till now, however, there have not been any investigations
that seek to make clear, at the level of abstraction of general systems
theory, what is to be expected when functional systems are differentiated as self-referential, operationally closed systems. This process cannot be grasped according to the schema-still predominant at the
time of Romanticism--of part and whole. The same goes for general
concepts of the advantageous division of labor or, negatively formulated, of the eternal conflict of apriori binding values; phrased in
terms of proper names: the point holds for Emile Durkheim and Max
Weber. For neither can one assume that an "organic" solidarity corresponding to the division of labor emerges on its own, nor is it justified
to conceive of values as fixed points on the horizon of action orientation. Today entirely different theoretical instruments are available for
a discussion of these foundational issues.

On these difficulties, but also on possible parallelisms among developments in the

natural sciences, cybernetics, and literary studies, see the book by the English scholar
trained in chemistry: N. Katherine Hayles, Chaos Bound: Orderly Disorder i n Contemporary
Literature and Science, Ithaca 1990, esp. p. 37.
The concept of "redescription" is here employed in the sense of Mary Hesse, Models and Analogzes i n Science, Notre Dame 1966, p. 157ff. One should, however, speak of
"metaphorical redescription" only if one accepts that n o theory can d o without
metaphors and furthermore that the concept of metaphor is itself a metaphor that uses
"metapherein" in a figural, extended, or translated sense.
See, for example, Siegfried J. Schmidt, Die Selbstorganisation des Sozialsystems Literatur
im 18. Jahrhundert, Frankfurt 1989; Niels Werber, Literatur als System: Zur Ausdifferenzierung literarischer Kommunikation, Opladen 1992. Cf. also Gerhard Plumpe, Asthetische
Kommunikation der Moderne, Vol. I : Von Kant bis Hegel, Opladen 1993.

NIKLAS LUHMANN

Important changes in the conceptual repertoire of systems theory result when one substitutes "essential definitions" ( WesensdeJinitionen),but
also so-called analytic system concepts, with the theoretical notion of
the operative closure of systems. Essential definitions rested on a heteroreferential (jirerndrefkrentiell)orientation, analytic definitions on a selfreferential orientation of the observer. The notion of operative closure
and, related to it, the theory of autopoietic systems presuppose that selfreferential systems must be observed. They are just that which they
make out of themselves. An observation is therefore only then appropriate if it takes the self-reference of the system and, in the case of systems operating with meaning (sinnhaft operimend), the self-observation
of the system into account. The "paradigm shift" that is thereby accomplished displaces systems theory from the level of first-order observation (systems as objects) to the level of second-order observation (sys
tems as subobjects or obsubjects, to employ formulations ofJean Paul) .5
With this turn, the distinction between self-reference and heteroreference is relocated within the observed observing system. Not only
the scientific observer must be able to distinguish between him/herself
and others (that is, between concepts and objects); this verba/res distinction is valid for all observing systems, even when they are occupied
with sense perceptions and have to rely on the external world without
. ~ generalizabeing able to distinguish between reality and i l l ~ s i o nThe
tion of the concept and the structural problems of observing systems
has far-reaching consequences, which only became apparent through
mathematical analyses. This detour via mathematics frees us at the
same time from the mystifications previously attached to concepts such
as "meaning" (Sinn) or "mind" (Geist). They enable us to see today
more clearly why and how something like "imagination" is required and
in what sense construction/deconstruction/ reconstruction as an ongoing process, an ongoing displacement of distinctions (Derrida's dzffhance), is necessary in order to dissolve paradoxes in and as time.'
See Clauis Fzchteana seu Labgeberiana, in Jean Paul, Werke, vol. 3, Munich 1961,
pp. 1011-56, or Flegeeljahre, eine Biographie, in Werke, vol. 2, Munich 1959, pp. 567-1065,
esp. 641.
"his special condition of an unavoidable trust in the world that can only be disrupted through critical reflection holds, however, only for psychic systems. For this reason we can leave it out of consideration in what follows.
The parallels between deconstruction and second-order cybernetics are treated
more thoroughly in: Niklas Luhmann, "Deconstruction as Second Order Observing,"
Nrw Literary History 24 (1993), pp. 76382.

In what follows we rely on the calculus of form developed by George

considerations are to be found in the secondSpencer B r o ~ nSimilar
.~
order cybernetics which Heinz von Foerster has elaborated."Here the
consideration as to what happens when the output of a system is immediately reintroduced into the system (that is, when the system forms
reflective loops within itself) leads to the concept of the non-trivial,
and therefore unpredictable, machine. And here too the problem
consists in the fact that the space of possibilities of the system is so
greatly expanded through self-reference that neither internal nor external observations can predict the operations of the system. A further
inference that can be drawn from these mathematical analyses: the system requires meaning in order to deal successfully (zurechtzukommen)
with both itself and its world.1
With reference to this problematic locus Spencer Brown employs
the concept of the "re-entry" of a distinction into itself.llHere too it is
a question of deploying possibilities of ordering that cannot be
achieved through the normal operations of the arithmetic and algebra and can only be demonstrated as paradoxes. Spencer Brown's
mode of presentation has the advantage of being directly applicable
to a very formal concept of observation. Observation is, in this context, nothing other than the use of a distinction for the indication of
one and not the other side of the distinction, however the system that
performs this might be constituted. For this reason the analysis concludes by referring back to its beginning in the equation of observing
and drawing a distinction: "We see now that the first distinction, the
mark, and the observer are not only interchangeable, but, in the form,
identical."12
For the analysis of the Romantic world, the consequences of such a
re-entry are of central importance. If it can be accomplished (whether
it is accomplished is then an empirical question), the system reaches
a state of "unresolvable indeterrninacy."l"he
decisive aspect of this
See George Spencer Brown, Laws ofFurm, New York 1979. Cf. also Dirk Baecker, ed.,
Kalkul derForm, Frankfurt 1993.
Heinz von Foerster, Observing Systems, Seaside, Cal. 1981. See also the German edition expanded with several additional contributions: Heinz von Foerster, Wzssen und
Gmissen. Versuch einer Briicke, Frankfurt 1993.
l o This presupposes, of course, a de-subjectification of the concept of meaning. For a
thorough elaboration of this point see Niklas Luhmann, SoziaZe Systeme: CrundnjU einer
allgemeinen Themie, Frankfurt 1984, p. 92ff.
l1 On this point and on what follows, see Spencer Brown, Laws oJFwm, 54ff., 69ff.
Ibid., p. 76.
l 3 Ibid., p. 57.

"

510

NIKLAS LUHMANN

concept is that the indeterminacy is not explained with reference to

dependence on an overpowering, itself indeterminable environment,
but rather is caused by the re-entry within the system itself: It is thus a
matter of seIf-generated uncertainty with which the system in one way or
another, but in any case selectively, must deal.
In order to do this the system requires:
1. a memory function. Memory must be understood here as the presentation
of the present as the result of the past; or alternatively as the result of an
ongoing discrimination between forgetting and remembering.14 The
memory function is thus a necessary accompaniment to all operations of
observing systems. It is by no means a matter of the occasional calling up
of memories on the time-scale of the past (after the pattern: where did I
put my glasses?).

2. an oscillator function. This can be interpreted-going

beyond Spencer
Brown- as the correlate of the use of distinctions. With every deployment of distinctions in observation the system will also observe (mitbeobachten) the possibility of crossing the border of the distinction with a
further operation and thus moving from one side to the other-for example: from the positive to the negative, from the good to the bad, from
the allowed to the prohibited, from the useful to the non-useful, from the
profane to the sacred, etc., from the realistic to the fantastic and back
again.

With the memory function the system binds itself to its own, now unalterable past. In this way it produces a present with a past horizon and
motivates itself to proceed from the present state of the world rather
than presupposing everything as new and unknown at every moment
and thus always starting from the beginning.15 For this reason there is
no "originary" present, no present that would be its own origin. With
the oscillator function the system holds its future open-and not
merely as the freedom of performing this or that action, but with regard to the fact that everything can arrive different; and this not arbitrarily, but depending on the distinction being used, which, because it
l4 Hence of the freeing-up and the reimpregnation of the observational capacities of
the system. This according to Heinz FBrster, Das Gedachtnis. Eine quantenmechanische Untersuchung, Vienna 1948. This formulation, by the way, shows how identities emerge,
namely through confirmation (Bmahrung) in reimpregnation or, in the terms of
Spencer Brown (Laws ofForm, p. l o ) , through condensation and confirmation; in any
case, however, through the ongoing equation (Abgleichung) with new irritations but not
with fixed contents of the environment.
l 5 In doctrines of wisdom the opposite requirement is occasionally stated: "The wise
perceive every thing as new, in attentive observation if not at first glance." Baltasar
Gracian, Criticon oder Ueber die allgemeinen Laster des Menschen, Hamburg 1957, p. 15.

includes what it excludes, indicates what in any given case can be otherwise. This too does not require, but rather makes possible a chronometric ordering of future temporal positions.
The difference between the simultaneously required memory and oscillator functions makes the construction of time necessary, the distinction of past and future and the insertion of a present between them in
which alone the system can operate. Via temporal difference modaltheoretical paradoxes can be dissolved, for example the supramodal
necessity of contingency that was once so important to theology. The
necessary can now be seen as a consequence of its being past, the contingent as a feature of the future. With the distinction of past and future the system can, additionally, deal with the requirement that it simultaneously (!) generate and hold in store both redundancy and
variety; the requisite redundancy will then be attributed to the past, the
requisite variety to the future. And that still leaves the question open
whether one conceives of the present as constant, as enduring, and
time as flowing through it, or construes the present of the system as
process, as a movement out of the past in the direction of the future.
The system can think of itself as static and as the correlate to the eternity of God, for example as a soul which must endure its temporal existence; or as dynamic and with the necessity/impossibility of using the
present in order to shape the future. This distinction can then be used
to adapt the temporal structures to socio-cultural configurations. In any
case, however, the constructivist analysis compels one to conclude that
every present is furnished with past and future horizons and for that
reason that the future can never become present.16 The temporal horizons only shift with, indeed by virtue of, the operations of the system so
that from moment to moment new pasts and futures are being selected.
Reformulated in terms of the theory of games, what follows from
this analysis is that the game can only be played within the game and
only with distinctions that identify the individual operations a n d simultaneously the play itseV1' That's why Adam (in Paradise Lost) had to
have the world explained to him by the archangel Raphael; and that's
why Henry Adams can only describe his education as the play of an indeterminate I against an indeterminate world.18
"'n
this point see Niklas Luhmann, "The Future Cannot Begin: Temporal Structures in Modern Society," Social Research 43 (1976),pp. 130-52.
l 7 For several mathematical variants of this theme, cf. Louis H. Kauffman, "Ways of
the Game-Play and Position Play," Cybmnetics and Human Knowing2/3 (1994),pp. 1734.
la Henry Adams, The Education ofHenry Adams, New York 1918.

NIKLAS LUHMANN

In certain respects, mathematical theories have today overtaken what

in the so-called human sciences (Geisteswissenschaften), but in sociology
as well, had always been intuited and expressed through a rather ambiguous use of language. This is true above all for chaos theory.lg It is
also true of the catastrophe theory of Reni. Thom and of the postGodelian calculus of forms of Spencer Brown discussed above. Of
course, it is not to be expected that Romanticism anticipated and more
or less intuitively took such a development of formal theory programs
into consideration. However, a close examination of the texts of
Romanticism can disclose so many correspondences that it becomes
unavoidable to ask how they can be explained.
An overhasty conclusion would be to say that Romanticism is nothing
other than the poetic paraphrase of a mathematical problem, a poetic
version of mathematics. We shall leave that view aside and instead make
our way via a sociological theory that can sustain empirical verification.
This intention was already alluded to above. Its point of departure is the
notion that the functional differentiation of modern society can be conceived in terms of autonomous, operatively closed, autopoietic functional systems. That leads to the hypothesis that all functional systems
draw limits or borders and therefore must reproduce the difference
between inside and outside internally as the difference self-reference,'
hetero-reference. The transition from hierarchically fixed positional
orders describable as nature to the primacy of the distinction between
self- and hetero-reference is considered a characteristic, if not the decisive feature of Romantic literature20 (and, one can add, Romantic art in
general). That encourages us to be on the lookout for the above described consequences of re-entry. For in the final analysis the distinction
between self- and hetero-reference is nothing other than the re-entry of
the distinction system/environment into the system itself.

With the differentiation of the art system and its disconnection from
external compulsions, an excess of communicative possibilities

'"

On this point see Hayles, Chaos Bound (note 2 ) . On the discussions set into motion
by the theory o f thermodynamics, see Kenneth D. Bailey, Socioloa and the New Systems
Themy, New York 1994.
' O See esp. Earl R. Wasserman,The Subtler Language: Cntzcal Readings ofVeoclassica1and
Romantic Poems, Baltimore 1959.

emerges internally and must be internally controlled and brought into

form.21The relationship of redundancy and variety, which for a long
time had accompanied the description of art,22shifts in the direction
of a flood of variational possibilities that can hardly any longer be mastered.23The "marvelous" is not an invention of Romanticism, but of
the cinque cent^;^^ but when its differentiation is fully accomplished, art
can more forcefully distance itself from a pre-given reality.25More and
more, art must generate the requisite redundancies itself, and this
through the restriction of variety. Today one would speak of "selforganization." For this reason, Romanticism discovers itself as if new
born in an empty space and called on to give itself its own meaning.
How that's supposed to happen becomes a question in terms of which
one can gather together diverse themes of Romantic literature.
For example, the call for a new mythology.26With a formulation
coined to describe postmodern architecture but entirely applicable to
Romanticism, one could say: "Whereas mythology was given to the
artist by tradition and by patron, in the postmodern world it is chosen
'
That
can happen in an entirely "sentimentalist" fashand i n ~ e n t e d . " ~
ion by drawing on antiquity and Christianity, through borrowings that
reflect on the fact that they form their observations from a different

"

O n this see Peter Fuchs, Moderne Kommunikation: Zur The& des aperativen DisplaceFrankfurt 1993, p. 79ff.
ments,
'
For example, for the Renaissance in the twin concepts unita/moltitudine or, distinguished from these, va'simile/meraviglioso. For a representative example, see Torquato
Tasso, Discorsi dell'artepoetica e in partzculare sopra ilpoema eroica (1587), in: Prose, Milano
1969, where (p. 366) it is stated that the poet should rely more on the one than the other
("0 piu del verisimile o piu del mirabile") in order to produce "magior diletto." The
sphere of the "marvelous," however, is limited by the fact that means have to be found
"per accoppiare il meraviglioso co'l verisimile." (p. 367) Beyond this example, one could
of course recall such ancient cosmological distinctions as ordo/varietas or unitas/diversitas.
'3 At the same time, biology reorients its inquiries from pre-given essential characteristics to "irritability" as that characteristic which enables the evolution of living beings. SeeJean-Baptiste Pierre Antoine de Monet de Lamarck, Philosophie zoologzque, Paris
1809, reprint Weinheim 1960, esp. vol. I, p. 82ff.
'4 Cf. Baxter Hathaway, Marvels and Commonplares: Renaissance Literary Criticism, New
York 1968.
Of course, that doesn't mean that art can indicate the one-way traffic on Fifth
Avenue incorrectly or claim that Carthage defeated Rome. In this, Tasso is still right
(Discorsi,p. 367), but today that's no longer the problem.
P6 For example, in the sense of the "5lteste Systemprogramm des deutschen Idealismus," here cited from G. W. F. Hegel, Werke,vol. I, Frankfurt 1971, pp. 23436, or in the
sense of Friedrich Schlegel.
P7 CharlesJencks, "Postmodern vs. Late-Modern," in Ingeborg Hoesterey, ed., Zeitgeist
in Babel: T h Postmodernist Controversy, Bloomington 1991, pp. 9-21; here, p. 9.
~

"

514

NIKLAS LUHMANN

temporal position. In contradistinction to the Renaissance, the great

discovery of which was that there had once been perfection in this
world, the directive difference no longer lies in the distinction between
secular and theological descriptions, but in the temporal difference
between present and past.
Or-second
example-the accentuation of writing as a form in
which absences (author or content) can appear as present.28 "Die
Schrift hat fiir mich," Friedrich Schlegel confesses, "ich weiR nicht
welchen geheimen Zauber, vielleicht durch die Dammerung von
In Ludwig Tieck's William Lovell,
Ewigkeit, welche sie ~mschwebt."~"
the characters reveal themselves and their opinions only through writing. What's Romantic in this is not the presentational form of the epistolary novel, but rather the fact that an image of "voriiberfliegenden
Gefiihlen, die mit unserer Vernunft (nicht) in eins zu schmelzen
(sind),"" is fixed in writing. And when that which has been supposedly written down is published, the reader can dissolve the narrative
and accept as his/her own one of the possible points of view. Writing
evidently compensates for the displacement of an enduring present
with process, since it can be reused in the present, but also read differently. It fixes itself, as it were, but not the reader.
And above all-third example-criticism (Kritik), conceived as the
ongoing labor in reflection on the never-complete artwork. Romanticism, then, seeks forms with which it can respond to the necessity/
impossibility of transcending the limits of the imagination. The expressive devices on the literary plane that correspond to this are irony
and the fragment, in music the preference for the piano with its context- and continuation-dependent tonal qualities. The unambiguous
distinctions are no longer sufficient, every frame of observation refers
to a further frame of observation, which it confirms by realizing itself
in it.31
Systemic autonomy, to which Romanticism in this way endeavors to
respond, isjust what happened to the art system as a result of the functional differentiation of society. One can no longer expect instruction

''Here too the parallel to postmodernism, in this case to Derrida, is astonishing.

"

"Uber Philosophie," in Friedrich Schlegel, Werke, Berlin 1980, vol. 11, pp. 101-29;
here, p. 104.
30 Ludwig Tieck, Friihe Erzahlungen und Romane, Munich 1963, p. 378.
31 In this regard also the correspondences to postmodernism are not accidental. See
David Roberts, Art and Enlightenment: Aesthetic Theory aaferAdorno, I,incoln, Neb. 1991;
"Die Paradoxie der Form in der Literatur," in Dirk Baecker, ed., Problemr &Form, Frankfurt 1993, pp. 22-44.

from the religious system, the political or economic systems, nor from
the households of the most important families as to how artworks are
to be made. For this reason one could almost say: autonomy becomes
the destiny which is interpreted as a defence against external intervention; o r the invisible cage in which the artist is forced to select,
to be original and free. Romanticism thus views and deals with the
problem of autonomy on the level of the artwork and the creative freedom of the artist derived from it, but not on the social level of the functional system of art; for only in this way can Romanticism define its position. The social system of art lets itself be represented through the
idea of art.
All that reads like a commentary on the self-generated "unresolvable indeterminacy" that is unavoidable as soon as one reintroduces
the difference between system and environment within the system itself. And just as in mathematics imaginary numbers or imaginary
spaces are required in order to absorb paradoxes,32Romanticism condenses the imaginary to the fantastic, and thereby to forms that precisely do not mean what they show, but are nothing other than materialized irony.33 But that by no means implies that all forms dissolve,
that no distinction any longer holds, that everything becomes arbitrary. On the other hand, it does not suffice to postulate with Kant that
freedom is given for its rational use or that the genius must make a disciplined and cautious use of his geniality.34 Rather, the artwork receives the task of demonstrating its own contingency and being its own
program; and that makes very severe demands on both productive and
receptive observation, which therefore cannot happen "just any way."
Self-generated indeterminacy does not by any means imply that n o
meaningful operations, no determinations are possible; merely that
determinations must be recognizable as self-determinations and as
such observable. In other words, communication must be transferred
to the level of second-order observation.
Against this background the reason that the Romantics begin to play

3
' See Spencer Brown, Laws ofFwm, p. 58ff, where a tunnel is introduced beneath the
surface on which the system performs its acceptable calculations. Cf. Dirk Baecker, "Im
Tunnel," in Dirk Baecker, ed., Probleme drrForm, pp. 1437.
33 On the further development of this tendenc-th
ever new outraged opponents-up to surrealism, see Bohrer, Die Kritik dm Romantik, p. 39ff.
34 This is, by the way, a longstanding, pre-romantic idea. One encounters it in the distinction libertas/licentia of natural law theory or in the disegno doctrine of the n'nquecento
with its distinction between creative imagination and the skilled and practiced execution of a drawing.

516

NIKLAS LUHMANN

with "reality," doubling identities in the form of Doppelganger, twins, exchanged names, and mirror images, becomes intelligible: in order to
show that the same can also be otherwise and must be set into relation
with itself. Instead of the ontological guiding difference (Leitunterscheidung) between being and non-being-which on the side of being
congeals to substance so that in the reapplication of the distinction to
itself the side of being is confirmed-other
guiding distinctions
appear, for example, the distinction finite/infinite (determinate/
indeterminate) or, alternatively, inside and outside.35 Ontological
metaphysics, which took only one possible primary distinction as its
point of departure, now had to be outtrumped by a meta-metaphysics,
which could take shape with the typically Kantian question regarding
conditions of possibility. The localization of reality with respect to the
distinction inside/outside was then as now a hardly solvable p r ~ b l e m : ~ "
"1st das Reale auBer uns: so sind wir ewig geschieden davon; ist es in
uns: so sind wirs ~ e l b e r . " ~However,
'
because no adequate, sufficiently
rich, many-valued logic is available, the problem is displaced onto aesthetics. Translated into constructivist terminology, that means that the
decision as to what can be treated as reality and what not is made internal to the system. The reality test of "resistance" doesn't have to be
given up as a result, but it is no longer a matter of a resistance of the
environment to the system, rather of system operations to system operations within the same system-above all the resistance of the selfproduced memory against new impulses or occurent ideas, or the resistance of the already begun artwork or narrative against something
which can no longer be added to it. Viewed in this way, reality is nothing more than the correlate of consistency tests within the system, and
this can occur in such a way that magic, ghosts, the supernatural, etc.
are introduced into a tale so as to acquire narrative plausibility, which
can then be revoked within the tale itself when, at the end, a perfectly
natural explanation for all the strangeness is provided.38 The figure of
the Doppelganger thus means nothing more than that in reality there is
35 On the plurality of such "primary distinctions," see Philip G. Herbst, Alternatives to
Hierarchies, Leiden 1976, p. 88. Herbst's work is, by the way, quite probably the earliest
sociological response to Spencer Brown.
36 On the contemporary version of the problem, see N. Katherine Hayles, "Constrained Constructivism: 1,ocating Scientific Inquiry in the Theater of Representation,"
in George Levine, ed., Realism and Representation:Essays on the Problem ofRealism in Relation to Science, Literature, and Culture, Madison, Wisc. 1993, pp. 27-43.
37 Jean Paul, Vorschule derAsthetik, in Werke, vol. 5, Munich 1963, p. 7-514 (445).
38 This is a well-known narrative technique of Ludwig Tieck's, from William Louell to
Llas ZauberschloJ.

no assymmetry of original and copy; rather, that this is a distinction art

alone requires for itself, an entry on the cost side in the balance of its
autonomy.
All this can be handled with the de-reification (Entdinglichung) of the concept of world introduced already by Kant. World is
no longer a totality of things, an agpgatio corpororum, a uniuersitas rerum, but rather the final, and therewith unobservable, condition of
possibility of observations, that is of every sort of use of distinctions.
To formulate this another way, the world must be invisiblized so that
observations become possible. For every observation requires a "blind
spot,"ggor more precisely: it can only indicate one side of the distinction being used, employing it as a starting point for subsequent observations, but not the distinction itself as a unity and above all not the
"unmarked space," precisely the world from which every distinction,
as soon as it is marked as a distinction, must be delimited.
This invisibilization of the nevertheless doubtlessly given and presupposed world had dramatic consequences for Kant, Fichte, and
above all for the Romantics. It leads to an overburdening of the individual with expectations regarding the production of meaning and
therewith to the collapse of the communication weighed down with
such expectations. The individual endowed with reflection now receives the title of "subject." But the higher and more complex the expectations that subjects direct toward themselves and their others, the
greater is the probablity of a failure of their communications. Texts exemplary of this are Jean Paul's Siebenkas (the marriage scenes) and his
Flegeljahre.40 The forcing of subjectivity as the single answer to the
problem of world makes intersubjectivity difficult, indeed, if one is
conceptually rigorous, actually impossible. Today this necessarily leads
to the question whether the "human being," the "subject," or similar
collective singulars are a possible starting point for social theory at all.
The Romantics used them and couldn't give the matter a second
thought, for they had in any case no chance to develop an adequate
theory of society. For them this position was occupied by the concept
of "spirit" (Gezst) and by the French Revolution.

"

On this point, see Heinz von Foerster, "Das Gleichnis vom blinden Fleck," in Gerhard Johann Lischka, ed., D m entfrsselte Blick: Symposium, Workshops, Ausstellung, Bern
1992, pp. 15-47.
40 See also Ludwig Tieck, William Lowell, p. 603: "Es ist ein Fluch, der auf der Sprache
des Menschen liegt, daB keiner den anderen verstehen kann." Cf. also p. 383 (Balder's
letter to William 1,ovell).

NIKLAS LUHMANN

Nearly contemporaneous with Romanticism a new sort of concept of

"culture" (Kultur) arises, offering itself as a serviceable "memory function" for modern society. One can see this with respect to the Romantics, but also other "humanistic" (geisteswissenschaftlich) endeavors, including religion (Schleiermacher) and philosophy (the late work of
Husserl). From the middle of the eighteenth century, the term "culture" is employed as an independent expression, that is: it is no longer
related to the care of something else as in "agriculture" or "cultura
animi" (Cicero). Formally, culture is distinguished from nature, but
that is merely an external delimitation and says nothing about the contents that are seen as cultural and, as such, approved or disapproved.
Here too one must distinguish between themes and functions: the
themes of culture and its function with regard to the autopoiesis of a
highly complex societal system. The themes of culture are formulated
with reference to possible comparisons, in particular regional (at first
national) and historical comparisons. Historically, such comparisons
can in principle reach back indefinitely, as far as the "sources" that are
always being discovered allow. With respect to content, cultures are related to ideas (Zdeen) or values, for which an "apriori" validity, or at least
a fixed orientation, is presupposed well into the twentieth century. Following the schema laid down in the Kantian critiques or by some other
method, a plurality of validity types can then be posited, the unity of
which either remains unreflected or is described as a tragic conflict
(Weber) or as endless discourse ( H a b e r m a ~ ) . ~ ~
Ideas, values, validity claims of all sorts emerge as correlates or, as it
were, as secretions of the comparative construction of culture. In this
way one endeavors to retransform contingency into necessity, with the
result, however, that contingency reappears in daily practice-be it as
the merely approximate realization of ideas, be it as the ever renewed
necessity of deciding in cases of value conflict. This problematic occupies the thematic horizon of modern society, but still doesn't show
wherein the persuasive force of the comparative method consists. It
seems to be rooted in the fact that extremely diverse states of affairs
can neuertheless be compared, in the conspicuousness value of the
equality of the diverse, which is to say: in the successful solution of a
paradox. What is similar fascinates and, so to speak, proves itself by

4 1 One can speculate that Kant's Kntik der Urteilskraft aimed at such an integration,
but failed to provide it.

virtue of the fact that it is found unexpectedly. This is called "wit" ( Witz)
and is found "intere~ting."~'
One can show that the same is different
and that diverse things allow identities to be known so long as one directs the comparison in terms of this cognitive interest. But why should one
do that? For the reason that it is a cognitive strategy that makes it possible to deal with extraordinarily complex, in the final analysis worldsocietal states of affairs. The semantics of the society is keyed to its
structural complexity and one component of this is that talk of ideas
and values provides a surface description that prevents inquiry from
reaching the paradox of the equivalence of the different and thus
from developing modes of description sufficiently complex to grasp
the complexity of the society.
One could speak in this connection of a cultural ~ ~ r n ~ t o m o l o g y . ~ "
The themes of culture have a symptomatic function. They do not
merely mean themselves, but also something else; and that becomes
especially noticeable when they are formulated as unconditional, transcendental, or absolute, and are introduced into the communicative
process with precisely this import. Thus there arises in the course of
the nineteenth century a second culture, a culture of suspicion that
raises the question of what is being disguised by the themes of culture.
I am referring, of course, to Marx, Nietzsche, Freud, and the sociology
of knowledge that follows in their path.
Poking around in allegedly latent structures is a way of searching out
hidden interests. The appropriate response to such searching is a tu
quoqueargument, namely the question as to the interest behind this interest in latency. The suspicion of veiled motives becomes universal
and therefore trivial; it is then a matter of nothing other than a double description of reality with first- and second-order observation.
The considerations set forth in the previous sections allow for a reformulation of the question as to the function of cultural themes. Society requires a memory function that allows it to accept the present
as the result of the past and as the starting point for subsequent operations. A memory, however, does not merely hold past events in reserve; it accomplishes above all a continuous discrimination of forgetting and remembering. Most everything sinks away and very little
is so condensed and reconfirmed that it can be reused. This sortal

4P For the subsequent development of this configuration, see Karl Heinz Bohrer,
Pldtzlichkeit: Zum Augenblick des asthetzschen Scheins, Frankfurt 1981.
43 This is the formulation of Matei Calinescu, "From the One to the Many: Pluralism
in Today's Thought," in Hoesterey, ed., Postmodmist Controversy,pp. 15674; here, 157.

520

NIKLAS LUHMANN

function serves the ongoing adaptation of the system to that which it

can construct as repetition. However, as a sortal function it must remain latent because otherwise it would also remember what is forgotten. The memory must, to put the matter differently, accomplish a reentry of the difference between forgetting and remembering within
forgetting, and the form in which this occurs seems to be the construction of themes-of identities and generalizations that can be
in other
.~~
fixed in communicatively available d e ~ i g n a t i o n sThemes,
words, make possible a forgetting of forgetting, and at the same time
the way in which themes are constructed serves the ongoing adaptation of the system to itself, the continuing inscription of a consistent
"reality."
To return to Romanticism after this long digression: one can assume
that this systems-theoretical concept will contribute to a socio-historical understanding of Romanticism. With a peculiar preference for
transitional tones, for paradoxes, for the narratively produced believability of the unbelievable, for the cognition of what cannot be communicated, the Romantics cultivate a symptomology that avoids congealing to theses, which could then be accepted or rejected. The
previously binding, early European tradition has to be forgotten in
order to free u p new capacities, and then restaged in a timely form
(zeitgemaJ) with a nostalgia that reflects on itself. In Romantic poetry
and criticism ideas are evoked and simultaneously marked as unreachable.
The temporal conceptions of the Romantics also fit with this analysis. Time is still presupposed as a movement in the old sense and therewith related implicitly to the cognitive possibilities of conscious perception. But the present is experienced as precarious, as a caesura, as
the "Differential der Funktion der Zukunft und V e r g a ~ ~ g e n h e i t . " ~ ~
The ambivalence in the evaluation of the French Revolution provides

44 "Themesn-the reference, of course, is to communicating and therefore social systems. For perceptual (psychic) systems one would have to speak of "objects."
45 Novalis, Werke,ed. Ewald Wasmuth, Heidelberg 1957, vol. I, p. 129 (fragment 417).
Cf. fragment 2225 (vol. 11, p. 125): "Ale Erinnerung ist Gegenwart. Im reinen Element
wird alle Erinnerung uns wie notwendige Verdichtung erscheinen." Or Bliilhenslaub 109:
"Die gewohnliche Gegenwart verknupft Vergangenheit und Zukunft durch
Beschrankung. Es entsteht Kontiguitit, durch Erstarrung, Krystallisation. Es gibt aber
eine geistige Gegenwart, die beyde durch Auflosung identifiziert." Werke, Tagehiicherund
BriejFriedrich uon Hardenbergs, ed. Hans Joachim Mahl and Richard Samuel, Darmstadt
1978, vol. 2, p. 283. Cf. also Jean Paul, fitan, in Werke,ed. Norbert Miller, Munich 1969,
vol. 11, p. 478: "Nein, wir haben keine Gegenwart, die Vergangenheit mu13 ohne sie die
Zukunft gebiren."

a political illustration of the same tendency. And that seems to suffice

as a symptom of the insecurity of the Zeitgeist. One does not find the
way to an adequate theory of time although the idea of a three-phase
passage from the past through the present to the future has already
been refuted by the experience of the precarious character of the present, by its d e - o n t o l o g i ~ a t i o n . ~ ~
he
present
is valued precisely because
of its undecidablity (but wouldn't one then have to say: because of the
necessity of deciding?) and is projected onto the historical moment of
European society. The past loses itself in history. One can forget or remember it4'; one has to prophesize it, as Friedrich Schlegel c l a i m ~ . ~ ~
And the future becomes the best guarantee for the fact that the world
is indescribable, and will remain so.
Despite this historicization and, if one can put it this way, rendering
precarious of temporal conceptuality, however, the Romantics d o not
entirely succeed in detaching the concept of time from the premises
of ontological metaphysics. Their concept of the world is too strongly
oriented in terms of the human being for that. In contradistinction to
many animals,49for humans a thing remains identical to itself when it
shifts from rest to movement. And that suggests an ontologically
nested concept of time, oriented in terms of the phenomenon of
movement, a concept that presupposes identities that bridge the distinction movement/non-movement and can sustain not merely movement but also the change from non-movement to movement and vice
versa, that is, the "crossing" of this distinguishing limit. Even Heidegger will still have difficulty with this. From the perspective of a radical
constructivist theory of observation, however, identity is not a timeindependent given, but merely an instrument for binding time when
it is a question of mediating past and future in the present.
Science, including systems theory, cannot afford such cultivated undecidabilities in the temporal, material, and social dimensions. It must
aim for refutable theses. That does not, however, exclude attempts to
46 On this point, see Ingrid Oesterle, "Der 'Fithrungswechsel der Zeithorizonte' in
der deutschen Literatur," in Dirk Grathoff, ed., Studien zurAsthetik und Literaturgeschichte
der Kunstperiodz Frankfurt 1985, pp. 11-75.
47 A concept of memory based in quantum physics that fits this state of affairs can be
found in Heinz von Foerster, "Was ist Gedachtnis, daR es Ritckschau und Vorschau ermoglicht?" in Wissen und Gewissen,pp. 299-336. See also by the same author, Das Gedachtnis (note 14).
4X Werke (n. 29), vol. I, p. 199.
4Y For example, frogs. See J. Y. Letbin, H. R. Maturana, W. S. McCulloch, and W. H.
Pitts, "What the Frog's Eye Tells the Frog's Brain," Proceedings ofthe Institute of Radio Engineers 47 (1959), pp. 1940-59.

522

NIKLAS LUHMANN

d o justice to Romanticism in a theoretical redescription. The systemstheoretical instruments of description break with the semantic repertoire in terms of which Romanticism sought to understand itself. For
the actual aim of this redescription is a theory of modern society for
which Romanticism can only have-but this in a most revealing waysymptomatic value.
University ofBielefild