Literature calls into question the value of philology but is at the same time its twin.
In philology interest in theory is not a constant factor. Periods of intensive concentration on theory alternate with those in which it is neglected and its value questioned or even denied. Although theory played a major role in the founding of the discipline of philology -- in the works of F. A. Wolf, Boeckh and Schlegel -- a counter-movement set in shortly thereafter. Otto Jahn, professor of classical philology in Greifswald and Leipzig, asserted, obviously with pride, that he had never read a philosophical book.
And the renewed interest in theory in recent years has not caused it to be accepted as a matter of course in our field. The continuing uncertainty in this regard is especially evident in the current question of the reception of American deconstruction as it has become known primarily in connection with the name of Paul de Man. Here we shall take this development as a starting point for a case history of the difficulties of theory in the field of philology.
Deconstruction is usually mentioned as a literary theory. But it is precisely this assumption that we want to discuss.
Paul de Man himself defined his position clearly only near the end of his life. His answer is surprising in its reference point. He focuses his attention not on the experience of the productive power of theoretical reflection, but, quite the contrary, on strictly philological teaching of literature deliberately kept free of all theoretical efforts. According to de Man, reading texts closely as texts is far more productive and a more serious approach than any philosophical speculation or the chit-chat of evaluation, at least when it is done with the necessary analytical rigorousness. This is reminiscent of well-known criticism of theory. Does de Man consider works of art inaccessible or even superior to theory?
Indeed, this conflict between theory and practice, between conceptual thinking in the abstract and faithful reading of the original text does have a long tradition. As we know, even New Criticism in the beginning was criticized with this approach, for example by Douglas Bush in 1948. The topoi:
too heavy on theory, too much faith in science, and an avoidance of moral values -- a really timeless arsenal.
But de Man's position is not in any of the well-known camps. His assertion that so-called poststructuralistic theory is not intended to differ greatly from the close reading of the philologist makes his views hard to classify. Theory and close reading without theoretical guidance share namely a central interest in a common subject, i.e. the structure of the language. Thus de Man's surprising conclusion is that the turn to theory occurred as a return to philology. What does this mean? Did not the study of German in West Germany become so involved in theoretical questions in order to escape from a politically encumbered and epistemologically speculative or naive tradition of philology and create a "new," "future" way to study German language and literature? What is this unity of philology and theory to be like?
As de Man says, he is mainly interested in the problem of literary understanding, in theoretical questions about the possibility of literary interpretation. But he takes neither philosophical interpretation nor a science of literature which considers itself to be independent of literature as his point of orientation. Only an examination of the cognitive structure of the interpretative process can confirm the interpretation's claim to truth and legitimacy. For de Man this is also the way to examine the nature of a genuine critical discourse, and, according to my thesis, that means nothing less than the logic of the philological commentary, the scholarly language about a first text, also in the figurative sense, i.e. the "great" literary work.
First, therefore, I should like to offer a few observations about the relative historical and epistemological importance of the commentary. Then I shall outline de Man's position against this traditional background.
The commentary has long been at the center of our discipline. As late as the eighteenth century it still had a universal validity for a scholarly science responsible for all forms of knowledge transmitted in writing and which gained new knowledge by interpreting classical authors. With the rise of new methods and forms of communication among scholars, it has lost this nearly unique position. As the 18th century advanced, the unity of scholarly language and expert commentary, of eloquentia and sapientia, which long distinguished the commentary came to be considered increasingly as only a method which created 'historical' knowledge and was responsible solely for insights about individual things (notitia rerum singularum). It was not permitted to provide the more highly valued knowledge based on reason.
The commentary acquired new importance only when exegetical techniques again came into use. According to M. Foucault, it is used to analyze meaning and significance, to search for the expressive value of the primary text, and here that means the literary text. Starting with a theory of meaning, the secondary language of the commentary has the task of uncovering a deeper layer of real meaning. The commentary thus becomes an interpretation, the form in which it has, after all, shaped philology down to the present.
On the one hand de Man is completely within the tradition of this commentary philology when he considers knowledge of literature only a result of an encounter between text and commentary. On the other hand he also asks the basic question about the conditions for the possibility of the interpretational commentary as a scientific way of dealing with texts, i.e. a method based on certain criteria of truth. It is indeed by no means certain that science and the commentary are at all compatible. In the eyes of a discipline which understands itself primarily as a science, the term "commentary" seems to be only an imprecise name for the philological practice of interpreting literary texts, which is quite untheoretical and about which it may not be possible to formulate any theories. On the other hand, however, also seen as part of a strategy for producing knowledge, the existence of the commentary in its special nearness to individual works of literature is proof that a "scientific" approach is especially unsuited for dealing with literature, or rather that literature can be treated in this way only at the price of a reductionism.
De Man gets around this polemical opposition. His search for a theoretically enlightened philology is not restrained by dogmatic claims of authority. For example he takes the legitimation topos of philology based on its special "closeness" to the text and subjects this claim to an examination by asking about the possibility of immanent knowledge of literature, i.e. knowledge bound to the literary work itself. But, according to de Man, neither the logic of scientific theory nor a philosophical interpretation can answer this question. For that purpose both of these disciplines are to far from the literary text. For him the only certainty is that knowledge is always acquired by means of reading. In this view reading itself is the epistemological basis of the commentary and therefore the constitutive problem of the discipline because it is in a double sense the primary problem of the study of literature:
criticism is a metaphor for the act of reading. De Man concludes that a theory of philology is possible, if at all, only if one accepts the complexities of reading.
One could perhaps easily agree with this view if it did not at the same time present the most difficult aspect of the problem: reading is namely an act of understanding that can never be observed nor in any way prescribed or verified. To put it even more pointedly, the possibility of reading can never be taken for granted. If, however, reading cannot be subjected to a check through observation, how can the commentary avoid arbitrariness? In the logic of the commentary the answer can only be the first text, the literary work itself. De Man also accepts it as the most important criterion for measuring "divergences" between the first and the secondary text:
The work can be used repeatedly to show where and how the critic diverged from it. Immanence, he continues, is necessarily a part of all critical discourse.
This direct connection of the interpreting commentary with the literary work is, however, formulated with conspicuous caution. Clearly the act of reading is neither the mere actualization of text structures nor a result of clear reading instructions:The encounter with the language of literature involves a mental activity which, however problematical, is at least to a point governed by this language only. The validity of this principle is doubtful and, at the same time, true only in a certain respect. But then what guides the interpretation with the necessity required by its claim to truth? An answer is possible only if statements about literariness, about literature as literature, can be made. That, however, is by no means certain, and not only for de Man.
As is well known, in philology there is no general agreement as to what constitutes literaricity. A logic of the commentary cannot, however, avoid this problem of an explicit definition of literariness, as the commentary claims to be shaped in its organization by an immanent connection with the literary work. Only in the definition of literariness can, therefore, such a philological theory test the validity of its rationality criterion, i.e. answer the question as to whether it does justice to literature "as such."
De Man's attempt to devise a definition of literariness takes as its starting point the unity of difference of a first and a secondary text. Knowledge of the object can only be achieved in a commentating or interpretative language in which text and reader meet. As the view is maintained that interpretation is above all a cognitive act, an epistemological event, which owes its knowledge neither to the intuition of essayist methods of gaining access nor to interpretive skills, the determination of its cognitive status is of decisive importance.
For de Man the key to this clarification is to be found in the complex relationship between insight and blindness, the central nexus he places before the title of his omnibus volume. Every reader -- or observer -- has a "blind spot" in his perception of a literary text. A reader cannot read what he cannot read, and he cannot see that he cannot see what he cannot see. This is a limitation of perception as it is also formulated in cybernetics or system theory. Only a second reading, a commentary of the commentary, can overcome this blindness, without, however, being able to avoid a blindness itself:
The insight exists for a reader in the privileged position of being able to observe the blindness as a phenomenon in its own right. The further explanation of such a necessary connection between insight and blindness in the perception of literature links de Man's later work with this earlier basic idea. Here I shall concentrate primarily on one aspect:
on the question about literariness as the possible "cause" of this perception problem. To quote de Man himself:
What characteristic aspect of literary language causes blindness in those who come into close contact with it?
As I shall show, de Man's deconstructive philology looks for the answer, and also its own theory structure, on the one hand in the history of the formulation of the problem, that is in the commentary, and on the other hand in a reading of Rousseau as a paradigm of the definition of literariness it seeks.
But the historical point of orientation is determined less by traditional theories or terminology than by the practical reading of literature. The ambivalence of literary language noted by all attempts to define literature with immanent methods is reminiscent of a basic problem that has long shaped interpretations, i.e. the experience of the ambiguity of a literary work. As Peter Szondi observes, it has always been possible to read any part of a literary text allegorically or literally, figuratively or for its literal meaning. The problem for philology consists in deciding whether a given word, sentence, paragraph or larger part is intended to be understood metaphorically or not.
There have been innumerable ideas for interpreting, and this usually means examining, this experience of the ambiguity of a literary text. De Man himself does not take interpretive or esthetic formulations as his point of orientation. He considers the ambiguity, as does Szondi, to be a result of language and not an expression of the metaphysics of art. This makes the rhetorical organization or the rhetoricity of the language the central factor in the efforts to provide an explanation of the lack of a firm connection between the literary language and objectively accessible reality. For de Man rhetoric, in so far as it deals with language as language, becomes an epistemology of language. The central reference point from which such rhetoric is developed and presented as the key to understanding literariness is a reading of Rousseau.
Rousseau's texts not only deal with language as an object of reflection. In what they say about language they also provide information about their own nature or rhetoricity, including their own perceptibility, with paradigmatic validity for literature as such. According to de Man the self-reference of the literary language is evident in a self-commentary on the question: how is literature (as language) to be read?
De Man's views in this regard are well-known and can be summarized as follows:
language is necessarily figurative, since it is, in its very nature, in principle nonmimetic. It can have no substantial meaning; language is, to use Rousseau's analogy, comparable to music, which can only be understood as movement, as a sequence of tones, and therefore receives a (meaning) value only in the sequence. Even the text of this statement -- this is the first conclusion -- is according to this view figurative. But if this is so, this claim itself must also be figurative and cannot therefore be understood literally. In principle the text can be misunderstood, in any case if the distinction between figurative and literal, which guides the interpretation, is maintained. Thus the first sentence -- that language is figurative -- cannot develop its full validity if it is not taken seriously and literally. De Man sums up the result in a well-known sentence: Accounting for the "rhetoricity" of its own mode, the text also postulates its own misreading.
With this assertion, however, de Man attacks the usual view of the epistemological function of the commentary. For the commentary is considered especially as a critical activity which, according to Szondi, owes its knowledge to separation and decision on the basis of a long tradition. But de Man's concept of unreadability draws the commentary into a formulation of the problem (and a determination of the object) which does not follow the usual form of scientific knowledge. Here too, reading is supposed to produce information about the object through categorial distinction -- between literal and figurative. But it is the obligatory use of this relationship which is fundamentally questioned here, although, according to de Man, it is not possible to separate the understanding of a text from this unstable difference. In his view such a text containing equal but contrary arguments has to be read literally and figuratively at the same time. The decision to accept only one kind of reading fails to consider the equally important, contrary form. Literariness thus defines itself as whatever can always only be read wrongly: literature is so any text . . . that prefigures its own misunderstanding as the correlative of its rhetorical nature.
True, this view greatly restricts the commentary's claim to knowledge, measured by possibilities traditionally attributed to it as interpretation, but that claim is legitimate and necessary, as it is a consequence of the self-reference language of the primary text. De Man's "immanent" commentary, concentrating completely on the problem of the differentiation between true and false, shows that no reading can ever be considered the truth of literature in the sense of a determinable meaning of the text: as guided misguided reading it relates the literaricity of a first tale indefinable in a strictly logical sense.
From this perspective of the tradition of the interpretational commentary de Man continues to use the old topos of literature as a quite unique relationship to truth or reality. But he reformulates it in the aporetic of a self-reference literary language and makes it the central criterion of literaricity. This definition of literariness is marked, as we have seen, by a unique, special cognitive quality: the old topos of a special truth of art is alive and well in the figure of negative knowledge. This means that no statements that can be substantiated with certainty are possible. It is a knowledge of language itself and is considered as such in relation to every concrete use of the language in conceptual discourse. No science, ethic or social and political reality can be safe from this truth claim of the paradoxical figurativeness of literary language.
Viewed from this point, de Man is working on a paradox theory or, more precisely, he is reconstructing the theoretical basis of the philological commentary in the form of an aporetic reading text. This theory, however, has certain consequences, possibly intended by de Man. As is well-known, as contradictions between two equally strong arguments, paradoxes cannot be resolved. They can only be treated pragmatically. Put differently, paradox theory cannot function without pragmatism, and its possible strength can perhaps be justified only in this way. De Man's deconstruction performs this test of its own strength in the area of the didactic of literature.
The relationship between the knowledge of literature and its didactic (de Man:teaching) is, according to de Man, anything but clear. A satisfactory answer has not yet been given, but knowledge and didactic have long been considered connected with each other. Philology, according to the philologist Friedrich Nietzsche, was in its origins and has always been at the same time pedagogy. De Man maintains this pedagogical basic duty, as will be shown, in spite of the accompanying implied restrictions, with conspicuous determination:
Scholarship has, in principle, to be eminently teachable.
It is therefore hardly surprising that for de Man a program which does justice to this educational tradition of philology can in the final analysis only be an immanent theory of literature:
A theory that is true to its object lends itself better, in the long run, to being taught than one that is not. The old familiar reserve of the emphatic educational program toward all questions of usability in special social functional areas has also been taken over from tradition. But, after what has already been said, the reference point for this distance can no longer be an emphatic, idealistic-speculative or, to put it briefly, a neohumanistic concept of individuality. In terms of its justification, education is not here a function of human perfectibility asserted in the concepts of morality and nature. This does not mean that these things do not exist for de Man, but only that what is at stake is not the existence of an ethical, psychological, or theological discourse but their authority in terms of truth and falsehood. De Man therefore conducts his actual debate with traditional educational semantics not in the area of educational philosophy, but, reflecting his claim of the central role of reading, about the question of technical methods; more precisely, about the compatibility of technical methods and claimed educational content. The technical side of literary education cannot be limited to simple acquisition of apodictically asserted educational values of (Greek) antiquity, the Weimar classic period or, finally, a canon of world literature. For the assertion of the immediate testimonial force (Sagekraft) (Gadamer) of the classical text ignores the knowledge provided by poetics of the epistemology of literary language. Here that means that it suppresses (ignores?) the independence (?) of reading, which has always been entangled in the irreconcilable arguments about the textual production of meaning. Reading is too often viewed from the perspective of a pedagogically valuable object and therefore appears only as the transmitting of educational content whose validity depends on the liberal arts developed with reference to an idealized view of man.
Such a transfer of knowledge is incompatible with the deconstructive definition of literature, for if, as is claimed, the status of literature always remains uncertain, there can be no certain transition from a scientifically legitimated truth of literature to pedagogy:
no constant educational value can be ascribed to literature as a condition for its didactification as a cultural asset.
De Man seeks another didactic in literature. In doing so he argues on the one hand in the direction of the well-known topos of literature as a subversive "counter-text" to the socially standardized language. On the other hand, a second, by no means congruent line of argument aims at the general status of thematic or moral statements in the act of reading.
De Man's starting point is the contradiction, here already reconstructed, between the rigorous claim to knowledge and the inevitable disappointment. He attributes a pedagogical effect to this characteristic formulation:
for such a knowledge claim is not realized in some kind of positive knowledge or a moral value, but rather leads "only" to a treatment of the text bound to the aporetic logic of literariness. According to de Man this connection produces a subversive energy, since in the logic of literariness only negative reading can be appropriate, i.e. a method in which the actively negative relationship peculiar to literature to the disciplines of logic and grammar, which produce meaning, is brought into play. In de Man's view, literature as a function of the aporetic self-reference of language undermines the differentiating power of categorial models and thus at the same time, since no positive meaning can be asserted without it, all authoritative linguistic validity claims. This theoretical inadequacy, as de Man states with unusual emphasis, contains a necessarily pragmatic moment that certainly weakens it as theory but that adds a subversive element of unpredictability.
In the argumentation all of this seems consistent. Nevertheless the question presents itself as to how far a primarily technical method intended to refute all authoritative claims to meaning can live up to the didactical claim made for it, the claim to an everyday, serious (?) meaning. First of all, this means only that such a reading in its unlimited negative impact is contrary to the entire humanistic, rationalistic tradition of a value-oriented literature didactic. But can reading which is limited to the independent existence of the language, which retains the personal reader only as a kind of transitional factor for anonymous text processes, still show any bias? As we know, in de Man's works there are hardly any explicit statements on political and social reality or morals, but this can hardly be the answer. According to de Man, speaking as an expert on linguistics or rhetoric and not as a general communicator, the philologist is concerned with political and social reality only as a reflection of the textual complexities that lead up to it. Nevertheless, this reading described primarily from a linguistic perspective seems to be able to show a meaning relevant to the everyday world:
in its orientation with regard to meaning it is in turn intended to generate meaning (Cf. B. Johnson). It is not difficult to imagine (?) that this meaning tends most strongly to a radical scepticism toward all authoritative validity claims, as the moral world -- at least so far as it is based on the explication of texts -- according to de Man can defend its reality only in a preordained agreement about the respective referential authority. (AR, Rousseau)
To summarize briefly the points I have covered so far, a deconstructive pedagogy of literature is, seen in this way, a literature didactic essentially critical of meaning which itself does not have any positive position or, to put it cautiously, does not want to have such a position. If there is any equivalent of the old idea of educational content, then it is the proof, characteristic of literature as such, of the nonprovability of linguistic validity claims. Political-moral discourse is replaced by the epistemology of the language. It defines ideology as the confusion of linguistic with natural reality . . . and thus asserts at the same time the epistemological priority of this deconstructive linguistics of literariness as an indispensable tool in the unmasking of ideological aberrations.
All of this, however, provokes an objection familiar in the debate about criticism of ideology:
how can it be excluded that the ideology-critical force of literary language can only be achieved in the final analysis at the price of an idealistic exaggeration of literature. In this model of ideology criticism, literature seems to be a factor directly subjected to language. This does free it of everything ideological, i.e. referential meaning claims. But at the same time this poetic liberation from discourse threatens to cause a return to the abundance of meaning just criticized. Literature itself thus becomes the measure of all criticism. The hoped for defense of literature fails because of the excessive claim.
The great number of deconstructive interpretations produced in American literature departments and their literature courses and considered proof of de Man's unusual success could possibly be explained as a result of such an overestimation of literature. Many of these interpretations, however, use the theory of deconstruction only as a subject. The theory-laden basic concepts of a deconstructive method of reading are reinterpreted as content statements about the world or the work which is intended to be interpreted. The knowledge of literature is anthropologized. The result is indeed knowledge important in an everyday sense, but at the price of an "alternative" decision, contrary to de Man's intentions, for interpretation and its question about the meaning of the work.
Now in philological deconstruction thematic readings are neither illegitimate nor unwanted. However, these statements about the moral world are not an expression of a content supposedly unique to the literary work, nor is an anthropological need for meaning peculiar to the reader realized in them. They are necessary parts of every reading solely because, according to de Man, they are founded in the aporetic law of literary language:
the ethical category is imperative . . . to the extent that it is linguistic and not subjective. The ethical dimension is a consequence of the actually impossible textual reading instruction, which demands unconditionally a figurative and a literal reading, i.e. one referring to the nonlinquistic world. In this view, scientific criteria, criteria within the framework of the difference between true and false, assert, in spite of the cognitive claim of deconstructive reading, neither epistemological superiority nor a claim to exclusivity realizable with regard to thematic or in this sense moral reading. Moral statements are inevitable when the general knowledge of the literaricity of the language provides the proof, always possible, of their lack of epistemological reliability.
For de Man, too, the categories or, in the language of system theory, the binary control codes of science and morality cannot be copied on to or even reconciled with each other. But he does not conclude that therefore the knowledge of literature can only do justice to this historical differentiation process with an either-or. Philology as a scientific discipline is neither swallowed up by a scientific theory of the text, nor can it be limited to an interpretation of the literary work oriented towards moral questions. In deconstructive reading both forms of perception have equally strong validity claims. As Hillis Miller observes:
For de Man the categories of truth and falsehood can never be reconciled with the categories of right and wrong, and yet they are values, in the sense of making an unconditional demand for their preservation. The aporie of literary language is logically unsatisfactory but at the same time a binding factor for separate partial systems, each of which is subject to its own function maxims. This paradoxical unity is realized less in a programmatic theory than in the deconstructive reading didactic:
a knowledge of literature with such a double perspective must, if it is to develop the ethical implications de Man intends it to have, practice reading which generates meaning as well as destroys it in carrying out its procedure (?).
As we know, the infinite change between construction and destruction makes final meaning in such a reading process impossible. But this kind of reading does affect convictions and orientations important for the reader. In the constant change of referential and figurative text perception the experience of the contingency of the moral world becomes irrefutable (?), without it being possible to relativize such an experience by referring to a supposedly superior "poetic reality." Reading causes the reader to develop an attitude of downright ontological irony towards all "metaphysical" validity claims of the moral world. In this view the distinction between figurative and literal kinds of reading does not stand for two kinds of meaning or interpretation. It has its didactic value in the distinction between recognized and new interpretations. It breaks with familiar interpretations of the world and, because of the ubiquitous difference of figurative and literal language, potentially demands new descriptions of the moral world at every point in a reading. Searching for a deconstructive diagnosis of the present, Richard Rorty celebrates the deconstructive treatment of texts as the essence of a post metaphysical culture. Much more sober, even skeptical, de Man emphasizes, in contrast, the limits of ironic knowledge. Irony does criticize the false appearance of an analogy or even representation in which language and reality are supposed to merge. But according to de Man, as an ironist the deconstructive reader is split into an empirical self that exists in a state of inauthenticity and a self that exists only in the form of a language that asserts the knowledge of this inauthenticity. A way out of this quandary in the direction of authentic knowledge does not exist. Ironic knowledge, in spite of its insights into the lack of reliability of language, remains, as soon as it faces the moral world, totally vulnerable to renewed blindness.
Especially a theory of literature that accepts the problem of a didactic of literature tends to develop a hypertrophied view of its own importance. This is the warning of those who describe de Man's work as "negative theology." De Man seems to want to refute this criticism with two precautionary measures, and, in my opinion, he largely succeeds. On the one hand, as shown, he shifts the reference point of ideology critique to the negativeness of language theory. His second defense, more of a morality based on knowledge than a logically compelling solution, is the deliberate appeal to the traditional ethic of the philologist. For neither philosophy nor a scientific theory, but rather the literary text itself is to be considered the decisive authority:
My starting point . . . is not philosophical but basically philological and for that reason didactical, textoriented. Therefore I have a tendency to put upon texts an inherent authority. At this point the ethic of the practical philologist, who completely commits himself to the great work, and the rigorous thinking of the theoretician, who pushes the demands of an academic discipline, overlap. Practice and theory are united in the self-ironic gesture (?) of a theoretician of philology:
the literary text, not theory, is the final authority, or perhaps not? As de Man says:
I assume as a working hypothesis (as a working hypothesis because I know better than that), that the text knows in an absolute way what it's doing.
Here de Man lets the question of the validity of his work as a philological theory, i.e. a theory developed from the 'immanent' knowledge of literature, disappear in irony and leaves the answer to the reader, or more precisely the community of philologists. A theory would have to be measured by its relevance to the specific questions essential to the identity of the philological discipline. From this systematic perspective based on the history of the discipline it is clear that de Man's negative epistemology of literariness has its strong point in an understanding of literature that is astonishingly traditional, not least because it is emphatic. Seen from a historical and systematic point of view, such a theory represents a contribution to a necessary desubstantalizing of philology without on the other hand denouncing the tradition of the discipline in the name of something which should replace it. The uncertain disciplinarity of the systematic study of literature is intended to be secured, according to the reverse strategy, by an epistemological recourse to the tradition of the philological commentary. There is, however, no question that this form of philological self-criticism does not approach the radicality of a consistent systematizing or historicizing of the discipline or the knowledge of literature. To put it differently and more succintly, not the least important criterion for measuring the strength of a theory is the question of whether it is able to free literary communication from the odour of being an anachronism.
After this critical examination, the commentary retains its function. Even in a deconstructive reformulation it is the necessary counterweight to an emphatic concept of literature as is found in the topoi of the authority of the text and the idea of a unique literary education (including a socially critical concept of the 'truth' of literature).
But de Man's definition of literariness by no means guarantees the commentary the insight into the knowledge object literature. Logic can present the problem, but not solve it, as N. L., another great lover of paradoxes has remarked. And that is also true of de Man. Radical in its precision, his thinking drives the interpretational commentary into paradoxes. The cognizability of literature encounters a decisive limit here. On the other hand, however, this does not mean that, when confronted with two equally valid kinds of reading, the commentary must simply dismiss(?) the work as unreachable or succumb immediately to irrationalism. Conversely, it could be argued that the aporetic form of the commentary is productive in the communication about literature institutionalized in the systematic study of it and in pedagogy. The difference between figurative and literal reading functions as a criterion for the acceptance or rejection of a commentary:
literary communication remains lively and productive, because the sterile dogmatism of repetitive reading, as well as reading which in its openness for other acts of communication that follow is arbitrary, i.e. which is not connected with any preceding or following commentary, is excluded. To put it concretely:
a deconstructive reader can entertain the traditional idea of the special importance of a literary text. He can also accept it as a work, because he knows at the same time that he can defend his rejection of this idea of the understandability of a (closed, unique, meaningful etc.) text.
In this view the commentary as a traditional form of knowledge of literature survives even without the idea of a 'deeper' meaning of literature. But this is only possible because the traditional definition of the commentary as interpretation is reformulated. The deconstructive commentary stands at the intersection of two basic ways of dealing with a text. On the one hand, deconstructive reading aims at a critical-linquistic analysis of the general laws of literary language. It does not attempt exegesis, but strives for a regularity (?) of the language even without any concrete reference to the nonlanguage world. Deconstructive commentary is thus bound to and at the same time separate from interpretation. If deconstruction is what happens in the literary text itself, then deconstructive reading is interpretation, because it repeats what is in the text. But what we call its result refers only in a negative sense to possible meanings and is thus far removed from asserting a meaning as the content of the work. Its result is thus 'only' the formalized description of the mechanics of the text, instability due to its rhetorical organization. The aporetic rhetoricity of the literary text, which shows itself only when one asks the text about its meaning, makes reading a formal analysis of the language itself, although such a formalization is not identical with the meaning sought. Formalization and interpretation are thus, as Michel Foucault explains in an astonishingly precise description of a deconstructive commentary perhaps previously found only in the works of Nietzsche, two correlative techniques between which there can be no simple choice, a choice namely in which philology itself can decide between the past, which believes in meaning, and the present (the future) . . . , which has discovered the significances (?). A resolution of this conflicting nature in one direction or the other inevitably involves, as a result of having to abandon the understanding of the text, the loss of the awareness of the problem already achieved.
A theory which is at one and the same time a science and a defense of literature, which in its procedures moves between the interpretation of the literary work and the linguistic analysis of language, can, however, never be certain of its success among the scholars of its particular field. Its double form is irritating, as it is incompatible with the accepted divisions of work and responsibility within the discipline. Deconstructive philology mixes things that normally do not belong together:
it insists on the necessity of scientific abstraction as well as on the emphasis on literature, and thus avoids the separation of text and work as well as the polemic between theory and practice. It measures itself by the tradition of an immanent, generally value-conservative philology and nevertheless at the same time tries to develop a linguistic-theoretical ideology critique as the basis of literary education. Thus the difficulty with deconstruction is also the difficulty philology has with its own traditions.