Wittgenstein as Conservative Deconstructor Author(s): Samuel C. Wheeler III Source: New Literary History, Vol. 19, No. 2, Wittgenstein and Literary Theory (Winter, 1988), pp. 239-258 Published by: The Johns Hopkins University Press Stable URL: http://www.jstor.org/stable/469335 Accessed: 07/02/2010 18:41 Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. 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/action/showPublisher?publisherCode=jhup. 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. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected]. The Johns Hopkins University Press is collaborating with JSTOR to digitize, preserve and extend access to New Literary History. http://www.jstor.org as Conservative Deconstructor Wittgenstein Samuel C. Wheeler III I. Introduction _DtE CONSTRUCTIVE LITERARY THEORY has been an important part of literary thinking for some time. Belatedly, analytic philosophers are coming to realize that deconstruction has common ground with "analytic" philosophy both in presuppositions and in strategies.' This realization typically starts with reading Der- rida but leads to an interest in literary theory generally. Literary thinkers have led some philosophers to read philosophy in the shade of different questions. Some analytic philosophers have thus come to appreciate the possibility of rhetorical readings of philo- sophical texts and the relevance of such readings for philosophy. More kinds of reading than "getting the argument" are beginning to seem interesting. More directly, there are some parts of the philoso- phy of language on which literary thinking is vastly more subtle and sophisticated than that of analytic philosophy. On a topic such as met- aphor, learning from literary theorists and critics can remove the under-practiced, self-confident naivete with which philosophers have dealt with figuration.2 Among the connections between literary theory and philosophy are very strong affinities between deconstructive thought and the thought of the later Wittgenstein. The single most striking difference is that Wittgenstein does not take his deconstructions to have dramat- ic consequences outside philosophy. Wittgenstein's diagnoses of inco- herence in philosophical theories seem to leave the application of the deconstructed concepts untouched. I call this Wittgensteinian reac- tion to the dissolution of a dualism "conservative deconstruction." This paper will discuss deconstructive arguments and focus on the question of what follows from a demonstration that a dualism is inco- herent. In particular, I show that Wittgenstein's thinking has impor- tant consequences for those who practice various kinds of "reading for contradiction." In particular, I want to show that some political critiques which use deconstructive arguments can, by adapting Witt- genstein's conservative deconstructive thinking, avoid the apparently crushing rebuke that their political position uses dichotomies which themselves are subject to deconstruction. 240 NEW LITERARY HISTORY II. in Dichotomy Philosophy The history of philosophy has numerous examples of the following kind of move: a pair of contrasted terms in ordinary language is ex- tended and adapted to provide an account of a certain kind of phe- nomenon. The dichotomy represents the actual phenomena as a mixture of both elements in the dichotomy. The sides of the dichoto- my are thus, by a systematizing ontological move, projected into a pure case and a generally amorphous "other." Since the concrete phenomenon is supposed to be really a mixture, each side of the di- chotomy must be thinkable as conceptually, if not actually, uncontam- inated by the other. Examples of such dualisms are the literal versus the metaphorical, rational consequence versus rhetorical connection, cognitive versus "other" meaning, analytic versus synthetic truth, and conceptual scheme versus the material organized by a conceptual scheme. These distinctions can be understood as descendents of the basic Platonic dichotomy of the middle dialogues. Plato distinguished be- tween the character itself, which is in no way other than its own na- ture, and an amorphous Other, variously described. Physical things which have a given character have other features and randomnesses besides, due to mixture with the amorphous Otherness. Such physical cases of the character are thus imperfectly and indeterminately of that character. The Forms in the middle dialogues are the pure, un- contaminated cases of which the material instances are more or less corrupt copies. The negative, completely unstructured side of the di- chotomy is some kind of Otherness, whose characterization turns out to be a problem. The dichotomy of Forms and stuff supported and was supported by an epistemological dichotomy. For Plato, knowledge of something is a relation which can only be total.3 If knowledge is to be total, every feature of an object must be known if the object is known. Thus the only objects that can be known are either very simple ones or objects whose features can be grasped in a mathematical way. Since Forms, if they are numbers or ratios, have an infinity of distinct properties, they can be known in the way that the number series is known, by a kind of rational grasp of an infinite totality. But clearly nothing with contingent properties or indeterminacies can be known. The concep- tion of knowledge as a total grasp of the entire object requires that what is known have a deductive, single-focus structure.4 Knowledge is thus domination of utterly perfect entities, while belief and lower forms of awareness are uncompleted graspings toward incomplete failures to be. Total mastery requires that both the knowing and what WITTGENSTEIN AS CONSERVATIVE DECONSTRUCTOR 241 is known be organized deductively around one principle. Such a de- ductive structure both masters and allows mastery. Both knowledge and being are conceived in terms of a mathemati- cal structure which can be grasped as a whole even when it is infinite- ly complex. Greek mathematics was geometrical, and the model is that of a visualizable totality. Such a foundational model does not allow the kind of folding-in on itself that is characteristic of the re- sults of deconstruction discussed below. The self-referential, involut- ed, invaginated structures deconstruction points out cannot be visual- ized, since their unlimitedness is different from the masterable infinity of a mathematical progression. A language which grasped forms would denote logoi. The relations traced among logoi would be logical and would follow the true ac- count of the natures of their signifieds, rather than mere rhetorical and poetical connections of word to word and sentence to sentence. Someone persuaded by logic would be persuaded by the real connec- tions revealed to reason, not by the accidental and irrational connec- tions that obtain at the level of mere physical words. The distinction between the logical and the rhetorical appears in a pair of distinctions in analytic philosophy which continue the Platonic pattern of deductive mastery and purity versus mixture. These dis- tinctions are those between cognitive meaning and other aspects of language we are tempted to call "meaning," and that distinction's supplement, the analytic/synthetic distinction. The analytic/synthetic distinction is driven by the demand that, since truth is determined by both the world and convention, it ought to be possible to separate out the contribution of the world so as to be left with the pure contribu- tion of the language. The logical relationships would then be the re- lationships among the meanings and could be given a total and com- plete account. The distinction between the cognitive meaning of an expression and the emotive, connotative, or "other" meaning of a term is a pre- condition of the analytic/synthetic distinction. That is, both sides of the analytic/synthetic distinction are themselves components of cog- nitive, truth-value-determining meaning. That the cognitive can be purified out of concrete language use is yet another instance of the dichotomous, two-factor, theory-building scheme. The dualisms above have three major features I want to focus on: First, they are foundational. "Foundational" must be explained in terms of the intentions of theory constructors. A foundational dual- ism is a pair of contraries designed to be the unproblematic compo- nents out of which an explanatory scheme is built. These contrary component terms are derived from a descriptive contrast that already 242 NEW LITERARY HISTORY exists in the conceptual equipment of the system builder. A descrip- tive term and its contrary are extended and generalized. By the char- acteristic dichotomous move, these expanded terms are taken to di- vide the phenomenon absolutely and exhaustively into components. Second, there is a value hierarchy implicit in the dichotomies. The left side of each dichotomy is the positive side and the right side the corrupting or negative side. The pure case of the left side of the di- chotomy is typically the state from which the other is taken to have fallen, or relative to which the other is defective. Whether this is of any significance for their real character depends on whether the cog- nitive meaning/other meaning dichotomy resists deconstruction. If it does not, then the hierarchical aspects of a pattern of thought are as relevant to its evaluation as thought as are its "cognitive," factual parts. Third, since some of the dichotomies are concerned with the ques- tion of what constitutes good grounds for conclusions, the very dis- cussions are reflexive. This format of theory building-which sys- tems built on these dichotomies seek essentially and exclusively- endorses the "Platonic-Formal" kind of connection among beliefs and a corresponding conception of the systematic. Thus the very demand for a certain kind of completeness is itself one side of the dichotomy- format which is at issue. A serious attack on the distinction between logical and rhetorical connection, for instance, will (according to its lights) change criteria for when an attack has succeeded. Both sides of the discussion have this self-referential, question-begging aspect, since the very rules of discourse are among the points at issue. The central dichotomy for purposes of understanding deconstruc- tion, and this reflexive structure of the debate between theory and deconstructivist argument, is the "cognitive meaning" versus "other" meaning distinction. "Other" here includes virtually all aspects of words, including emotive, rhetorical, metaphorical, historical, pho- nic, etymological, and so forth. The related distinction between cog- nitive connection and other connections among words and sentences is correspondingly diverse on the "other" side. Whether this distinc- tion remains erect after deconstructive attempts to undermine it basi- cally determines what argument and reason-giving are. III. of Deconstruction Strategies Deconstruction is a currently appropriate term for the practice in the history of philosophy of arguing that neither pure case of a di- WITTGENSTEIN AS CONSERVATIVE DECONSTRUCTOR 243 chotomy makes sense. The condition for a dichotomy to function foundationally is that the pure cases of the sides be coherently de- scribable. A deconstructive argument shows that each of the pure cases in fact presupposes the other half of the dichotomy. Ontologi- cally speaking, the allegedly pure cases of one side must in fact con- tain the other. So both sides of the dichotomy are in fact impure, and the foundational project cannot succeed. We can provisionally distinguish two phases of deconstruction as it applies to contemporary philosophy: The first is the rhetorical strate- gy that applies when one is attacking a standard while using that same standard. In this case, the notion of argument that applies is the then current cognitive meaning standard. Derrida's early writing and some of Quine's and Davidson's work correspond to this model.5 The second phase of deconstruction applies when the cognitive/other dis- tinction has been abandoned because it has been shown to be logically incoherent. Then the deconstruction need no longer be carried out according to the predeconstructive standards of argument and what counts as incoherence. Once the notion of the "incoherent" is ex- panded to cover more than "logically incoherent," "deconstruction" will be likewise expanded to cover the exposure of such broader inco- herence. Given an initial success, deconstruction becomes as much a rhetorical practice as a logical one, because it now no longer recog- nizes that distinction.6 There are basically four ways such deconstructive arguments proceed: First, the philosopher can attempt to show systematically that no dichotomy drawn in the way that a given dichotomy is will work. Such an argument shows that a certain kind of division is in principle de- fective because a theory employing the dichotomy must presuppose the other side of the dichotomy. Any account of one ineliminably mentions the other, so that the phenomenon cannot coherently be imagined to consist of a mixture of the two components. The diffi- culty with this strategy, historically, is that the point of view from which one establishes such a conclusion is itself a theory, a founda- tional, totalizing, conceptual edifice which is itself subject to the same kind of criticism. That is, to show that a given kind of result is in principle impossible, there have to be some strong principles which determine what kinds of conceptual configurations there can be.7 Second, a philosopher can try to undercut the dichotomy by show- ing that what the theory takes to be a natural division is in fact a product of other factors entirely. This shows that a position is ideo- logical, not scientific-that, for instance, economic factors, gender practices, diseased spirits, Freudian phallism, or other extralogical 244 NEW LITERARY HISTORY factors actually motivate a dichotomy. When such an argument shows (in some post-deconstructive sense) a "contradiction" in this ideology, it counts as deconstruction.8 Thinkers arguing in this way often suc- cumb to a temptation to be essentialists and theory builders about their own preferred dichotomy. For instance, to claim that something is ideology and not science supposes that there is something that would be pure of the contaminations by the interests and other fac- tors that constitute an ideology. It is difficult to use a dichotomy to attack another dichotomy without supposing that one's own set of contrast terms can provide a foundational theory. Third, the most characteristic strategy of deconstruction as prac- ticed by Derrida is to discuss a particular text and show how that text undermines itself, in that the division it is arguing for is implicitly denied in the very text itself. (The use of "implicitly denied" is rhe- torically complex here, since it must employ criteria for being implicit which are themselves under discussion.) Such deconstruction does not suppose that there is a well-defined single "kind of view" under- lying a whole culture or literature. Such an underlying view would presuppose the division between real meaning (or the essence of a view) and the mere words in which it is put. By attacking only texts, one avoids proposing a theory, dividing rhetoric from logic, and positing the thesis behind the text. The most accessible example of this strategy in Derrida is his treatment of Husserl's notion of the expressive sign.9 There he shows that the account of pure presence in fact presupposes what amount to positings into the past and future- that is, nonpresence. What makes this accessible is that he uses no modes of persuasion which are not already part of standard philo- sophical practice. What makes later deconstructions so bizarre is that they operate via new notions of consequence. Fourth, it is possible to attempt to end dichotomous foundational theorizing by showing that the dichotomy has no basis in what we say. This strategy does not show that the dichotomy misrepresents the "facts" about the language, but that there is no logical compulsion to restructure what we say on the basis of the theory. In the purest case, Wittgensteinian deconstruction, the argument form is basically one of observation of what is actually said when. This strategy transfers the burden of proof to the dichotomizer by denying that the conceptual scheme we have is dichotomous at all. This strategy would deny that "pure case" extrapolation is required by anything other than the ille- gitimate impulse to theorize. Thus, a given "incoherent" dichotomy is argued to be incoherent only because of an illegitimate oversimplifi- cation of an unproblematic practice. The practice itself only appears to have the problems the theory was designed to solve, because the WITTGENSTEIN AS CONSERVATIVE DECONSTRUCTOR 245 theory insists on extending concepts past the points where they con- tinue to make sense. The dichotomy itself, as theoretically under- stood, is an illegitimate projection. Thus, for such a deconstructor, what is deconstructed is not the equipment of thought, but a certain philosophical disease of theorizing. The dichotomy is shown to be incoherent only as extended, not in the form in which it exists in its original habitat. This is the basic idea of Wittgensteinian deconstruction, the para- digm of the "conservative deconstruction" discussed in the next sec- tion. Wittgenstein is not committed to any theory at all, or to any views about the perfection of ordinary language. An "ordinary lan- guage" attack on dichotomies which did not itself presuppose foun- dational, dichotomous accounts of language would concentrate on dichotomies in the philosophy of language, and with understanding the workings of language in a way which avoided theorizing.10 IV. of Deconstruction: Consequences and Conservative Revolutionary So what if a dichotomy has no pure cases, so that no theory built on the dichotomy can have the kind of simplicity, completeness, and vi- sualizable totality that is hoped for? What follows from a "proof" that the extended and completed dichotomy is incoherent? There are two characteristic positions taken by deconstructors, the revolutionary and the conservative positions. For the revolutionary, a deconstructed dichotomy should be tran- scended, discarded completely and replaced by a way of speaking which does not depend on this dichotomy. The incoherence disquali- fies the distinction as an acceptable piece of linguistic equipment. (A conscientious deconstructor would ironize the "tool" metaphor, since it both implies choice and implies the independence of tool and user.) The difficulties of "reform" and communication during the process of reform, and of coherence while a reform is being argued, are some of the most interesting topics in deconstructive thought. The revolutionary position is not that dichotomies can just be abandoned. The apparently radical view that the concepts surround- ing a defective dichotomy can just be jettisoned presupposes the di- chotomies that deconstruction overturns, and is inconsistent with the notions of concepts and persons and culture that "deconstructive" nondichotomous thought implies. The notion of jettisoning a con- ceptual scheme supposes that the linguistic is separable from the fac- tual, that there is a way of changing the words and keeping informa- 246 NEW LITERARY HISTORY tion constant. As Davidson has shown, the very notion of a conceptual scheme into which we can put our factual beliefs about social justice or the correct relations among the sexes is incoherent."1 Barring a Platonic conception of meaning as something which stands outside and links languages across times and cultures, the division into scheme and material organized must be rejected. But rejecting this, and recognizing that we are constituted by and not users of a language, makes rejection of an important part of language paradoxical. The revolutionary thus has a characteristic difficulty: To dismantle one dichotomy, one must use a language which employs other di- chotomies which are similarly defective. To attack a dichotomy in "Platonic" philosophical terms and by Platonic standards is to pre- suppose other dichotomies. Furthermore, these dichotomies are in- terconnected, and more or less invade our entire apparatus for think- ing about theoretical issues. Thus, given a general project of excising incoherent dichotomies, the target dichotomies cannot be attacked one by one, because they presuppose one another. Really getting beyond a dichotomy and thinking apart from it is to think apart from all dichotomies which cannot be sustained without it, and this would seem to mean thinking in entirely new ways. But to abstain from using all such dichotomies at once is to drop out of the conversation by abandoning the links of understanding which make communica- tion possible. To eschew "dichotomous" concepts and argument forms is to abandon the formats of argument that are recognized as serious dis- course. Strategies such as using terms "under erasure" acknowledge this problem without solving it. The actual result of an acute con- sciousness of this kind of problem is a kind of ironic philosophy, where the writer is distanced from the conclusions and expressions of the text, hoping to keep at bay the implications of all the terms being used. Such consequences seem to some thinkers to disqualify decon- struction as a tool for those who would change our ways of thinking. The accusation that "You're presupposing the same kind of totality" seems an effective way to force deconstructive arguers into quietistic irony. Conservative deconstruction shows a way out of such binds. The conservative tactic avoids self-referential incoherence by rejecting the theorizing extension of a dichotomy. Unless a dichotomy is forced to pure cases, uncontaminated by their opposites, incoherences do not show up. But unless we adopt Platonic premises, there are no grounds for supposing that those incoherences are already implicit in ordinary language, waiting, as it were. Unless the logical demand that WITTGENSTEIN AS CONSERVATIVE DECONSTRUCTOR 247 a dichotomy be purified, extended, and turned into theory is already there in the true nature of our concepts, the fact that deconstruction shows the incoherence of theorized dichotomies says nothing about the nontheorized concepts. But, of course, there is no "true nature" of concepts beyond what our discussions in their terms construct. So the theorizing impulse is the villain, not the dichotomy. This is the idea behind Wittgenstein's deconstructions and the conclusions he reaches therefrom. I call this "conservative" deconstruction because it attempts to pre- serve the patterns of what is said when-that is, the language without which we do not exist. Since dichotomies are a very important orga- nizing part of this language, the dichotomies (or something like them) must be preserved as well. Conservative deconstruction claims that the theory can be removed while the dichotomy is still in place. That is, the supposition is that the theory is inessential to the dichoto- my, which can be used in its normal ways without the kind of hyper- bole which constitutes philosophical theorizing. According to conser- vative deconstruction, there is nothing wrong with the dichotomy itself. Rather, the difficulty is with the theorist who insists on pushing the dichotomy past the point where it makes sense. Its failure to make sense is basically the same kind of "presupposition of the oppo- site" that other deconstructors point out. For a conservative deconstructor, for instance, the failure to make coherent a principled distinction between the metaphorical and the literal just means that principled distinctions are not required. One need not find a principled distinction between the metaphorical and the literal in order for there to be a difference between more and less literal speech. A principled distinction would derive from a statement of necessary and sufficient conditions embedded in an appropriate complete theory, defining the pure cases from which real world mix- tures are derived. Conservative deconstruction tries to preserve the useful dichotomies ("ordinary language is in order") by eschewing the drive to theorize. The theorized and overextended dichotomy is the target of deconstruction, not the dichotomy itself. The overween- ing Platonistic demand for total vision is something to be cured. V. as Conservative Deconstructor Wittgenstein In this section I sketch some of the ways in which the later Witt- genstein shares important views and attitudes about language with Derrida, Quine, and Davidson. I then argue that Wittgenstein's de- scriptions of language, remarks on what must be and what need not