Printer Friendly

Lyotard and Kripke: essentialisms in dispute.

THERE is at least one dogma shared by most analytic and continental philosophers: that they do not have anything to say to one another. Neither Jean-Francois Lyotard nor the authors of this paper[1] subscribe to this view. Lyotard's book, The Differend: Phrases in Dispute[2] employs concepts and theories derived from both traditions. In this paper we discuss the way in which Lyotard uses Saul Kripke's theory of proper names and analyze Lyotard's attempt to escape from the essentialism Kripke has derived from his own work.[3] We draw two conclusions: (i) Lyotard does not succeed in avoiding essentialism, (ii) he has not established the existence of differends.


The Differend opens by locating as its main concern the issues of validating language and understanding how it locks on to reality. In the case of historical events, for example, how is it possible to validate the fact that a (now past) event actually occurred? As a means of explicating and examining this possibility, Lyotard expounds a philosophy of language centering on the structure and function of what he calls "phrases."(4) Lyotard is unwilling to provide a definition of a phrase. This is principally because a definition is itself a phrase, and privileging the definitional phrase over other kinds of phrase would not provide an adequate account of the term (Lyotard, 106, 107). He does, however, give some examples: "French Aie, Italian Eh, American Whoops are phrases. A wink, a shrugging of the shoulder, a tapping of the foot, a fleeting blush, or an attack of tachycardia can be phrases" (Lyotard, 110).

Each phrase presents a universe which is constituted by four instances and their relations. These instances are "what the phrase is about, the case, ta pragmata, which is its referent; what is signified about the case, the sense, der Sinn; that to which or addressed to which this is signified about the case, the addressee; that `through' which or in the name of which this is signified about the case, the addressor" (Lyotard, 25).

Phrases obey rules and because of this they can be characterized as operating within particular "phrase regimens." A phrase regimen can be cognitive (and thus empirical), ostensive, etc. Each of these possible phrases must be rigorously distinguished from one another; ostensions, for example, must not be confused with descriptions. For a phrase to be under a specific regimen is for it to present a universe where each of its four instances are related in a particular way. This entails that the mode in which a phrase presents its universe is determined by the regimen to which it belongs (Lyotard, 96, 178-80). A phrase regimen, then, is correlated with a set of rules for forming phrases that pertain to a certain modality of presentation.

Phrases belonging to different regimens are heterogeneous: For each of these regimens, there corresponds a mode of presenting a universe, and one mode is not translatable into another" (Lyotard, 178). Although phrases are heterogeneous, they can be linked together to form genres (Lyotard, 179, ff.). Lyotard claims that "a genre of dis-course imprints a unique finality on to a multiplicity of heterogeneous phrases by linkings that aim to procure the success proper to that genre" (Lyotard, 180).

One can have the genre of logic, the genre of history, etc. A genre has a particular teleology, which is realized by the specific mode of linking adopted. The difference between genres and phrase regimens lies in the fact that whereas the latter provide rules for forming and using phrases, the former provide rules for securing the success of a given enterprise:

The rules of formation and linkage that determine

the regimen of a phrase have to be distinguished

... from the modes of linking that stem

from genres of discourse. As Wittgenstein observes,

the set of rules constituting the game of

tennis or chess is one thing, the set of recommendations

which form a strategy for winning

is something else (Lyotard, 185).

Hence, Lyotard is able to draw a distinction between, for example, the phrase regimen of logical phrases and the genre of logic. The logical phrase regimen provides the rules that constitute the "game" of logic, whereas the genre of logic allows for the practice of logical discourse to be undertaken with a view to a purpose, which in this particular example is the aim of arriving at a tautology able to account for the entire set of phrases that make up the genre itself (Lyotard, 175).

Another genre of discourse is that of validation, which provides those modes of linking that allow phrases to be proven. For example, "a descriptive is validated cognitively only by recourse to an ostensive (And here is the case). A prescriptive is validated juridically or politically by a normative (It is a norm that ... ), ethically by a feeling (tied to the You ought to), etc." (Lyotard, 41). Validation is carried out with respect to the observance of a set of rules proper to a particular genre.

The manner of linking phrases together within a genre, it follows, may be contingent(5) insofar as it will be subsumed within the demands set by a given goal. But it is never arbitrary, since sets of rules exist which govern how to link phrases together within any particular genre of discourse, be it that of logic, science, philosophy, etc.

The manner of linking is thus contingent because there are no imperative rules to tell us that we must obey the rules of any particular genre. That the rules are not arbitrary springs from the fact that genres stipulate finalities. In this sense also, Lyotard is making a claim which follows Wittgenstein's:

Why don't I call cookery rules arbitrary, and

why am I tempted to call the rules of grammar

arbitrary? Because "cookery" is defined by its

end, whereas "speaking" is not. That is why

the use of language is in a certain sense

autonomous, as cooking and washing are not.

You cook badly if you are guided in your cooking

by rules other than the right ones; but if

you follow other rules than those of chess you

are playing another game; and if you follow

grammatical rules other than such-and-such

ones, that does not mean you say something

wrong, no, you are speaking of something


Understanding the above in Lyotard's terms, one might say that "cookery" is analogous to the genres of discourse in so far as they are both determined by their finality, whilst language considered as a whole has no finality. Within a genre one can play "well" or "badly" according to the rules of that particular genre. However, playing "badly" might also be taken in another way, one's stakes might turn out to be different from those of the genre.

In contrast to genres, there are no stakes for language as a whole. Hence: (i) there are no "right" or "wrong" rules for language considered as a totality; (ii) since ultimately rules are simply embedded in current practices, a practice devoid of finality will be one whose rules are arbitrary.7 Thus, in relation to i) since for Lyotard "language does not have a single finality" (Lyotard, 231), "there is no 'language' in general, except as the object of an idea" (Lyotard, Thesis," p.xii) - an "object of an idea"8 is not subject to the rules applicable to establishing the reality of a referent (Lyotard, 5). With regard to (ii), Lyotard claims that there is no genre-independent way of distinguishing between playing a game "badly" and playing another game; hence, he is committed to the view that there is no objective way of individuating which rules pertain to a game.

Furthermore, putting the case in a manner somewhat alien to Lyotard, we may say that this impossibility is not merely epistemic. There is no fact of the matter as to which rules are the rules of the game. In this sense, rules considered at the level of regimens are, strictly speaking, arbitrary: there is no independent means of establishing why the cognitive regimen, for example, is cognitive (Lyotard, 188).

As we have said above, generic rules, such as those of validation, are not arbitrary because they are determined by the stakes of genres. One example of validation concerns instances in which a claim is made that a damage has occurred. A claimant asserts that he or she has been injured; this claim is made within a certain genre of discourse and the rules of the genre allow him or her to validate that claim. As opposed to a damage, however, a wrong is an instance of "a damage accompanied by the loss of means to prove the damage" (Lyotard, 7). For example, one way of losing such means is when a French citizen who is a Martinican cannot complain about the possible wrongs which she may suffer from as a result of being designated a "French citizen" because her testifying phrase is deprived of its authority Lyotard, 36). This is because the genre of French law itself, which is the only genre within which such a complaint could be lodged, prohibits the possibility of making it.

The latter is only one among many similar examples of differends that Lyotard gives in the text. The apparent clarity of such examples, however, in fact conceals what could be an ambiguity with regard to the differend. On the one hand, there is a clearly epistemic characterization of what a differend is: "I would like to call a differend the case where the plaintiff is divested of the means to argue and becomes for that reason a victim" (Lyotard, 12). On the other, however, differends also seem to be understood in terms of their semantic status: "The differend is the unstable state and instant of language wherein something which must be able to be put into phrases cannot yet be" (Lyotard, 22). The epistemic characterization implies that a differend is a phrase that is prevented from succeeding because of the rules of validation operating within the genre in which it occurs. The semantic one suggests that a differend is that which cannot even be phrased. Hence, the first characterization indicates that a differend is a phrase which is not "validatable" by means of existing rules of validation, whereas the second appears to make the stronger claim that the existing rules of formation do not even allow for the formulation of the necessary phrases. Since rules of formation determine phrase regimens and rules of validation determine genres, it seems that Lyotard is conflating his own distinction. This conflation is accompanied by a similar apparent confusion between the epistemic and the semantic.

We would hold, however, that Lyotard can be read as postulating a relational priority of the epistemic over the semantic. Because the rules of validation do not apply to it, according to Lyotard what might prima facie appear to be a well-formed phrase(9) is excluded from every regimen relevant to the genre of discourse in which it attempts to situate itself. This is due to the fact that the sense of a phrase is suspended until the phrase is linked to another phrase, which situates it within a genre and thereby subjects it to epistemic rules (Lyotard, 101).

Since Lyotard claims that sense is dependent on the relations between phrases, he endorses the view that sense is a holistic matter.(10) This is the view that nothing has sense unless other things have it, too. It is for this reason that Lyotard holds linking to be necessary (Lyotard, 136). Furthermore, since there is no last phrase (Lyotard, 17) sense is never determined, hence Lyotard's theory of language is a theory of sense in which there are no determined senses. This position is akin to Quine's theory of meaning without meanings.



It is by the linking of one phrase to other phrases that the first is validated. Thus, for example, cognitive phrases (or definitions) are validated by ostensive phrases. The cognitive is instanced, "Here is a case of it," the "of it" referring to the cognitive phrase. Thus, the phrase "Here is a red flower" can be turned into two phrases: (i) a cognitive phrase/definition in which the term "red" corresponds to a specified wavelength in the color spectrum; (ii) an ostensive phrase: "the colour of the flower, here is a case of it." The deictics then need to be removed and replaced with proper names. These proper names now allow for the possibility of any given red flower "being situated in the position of the referent as supplied in the ostensive phrase" (Lyotard, 61, 62).

The theory of the proper name that Lyotard endorses is largely derived from Saul Kripke's discussion in Naming and Necessity. Following Kripke, Lyotard holds proper names to be "rigid designators" (Lyotard: "Antisthenes Notice," p. 37). According to Lyotard, a rigid designator "does not ... have a signification, it is not ... the abridged equivalent of a definite description or of a bundle of descriptions ... it remains invariable from one phrase to the next ... Its rigidity is this invariability. The name designates the same thing because it remains the same" Lyotard, 57).

Up to this point Lyotard follows Kripke very closely. Prior to Kripke, the dominant theory of proper names in analytic philosophy was the description theory, according to which a proper name is synonymous with either one or a cluster of definite descriptions. For Kripke, in contrast, a proper name is rigid: its referent always remains the same although initially it might have been fixed by means of a description.

Kripke's discussion of the linguistic behavior of proper names is focused on their role in counterfactual sentences, rather than on their functions in phrases belonging to different genres. Hence, Kripke explicates the rigidity of proper names in terms of the permanence of their referents in all possible counterfactual worlds in which the name refers.(11) Since proper names are not synonymous with definite descriptions, it is possible to succeed in referring to a thing without knowing anything about it, that is, without knowing of any descriptive sentence that it is true of that thing (Kripke, 80).

This view has, Kripke claims, important epistemological consequences. Since the referent of a proper name is not uniquely determined by the set of properties attributed to it, two speakers may agree on which purely qualitative descriptive sentences are true of the referent and nevertheless refer to different things when using that name. In other words, what is referred to by means of a proper name is not completely determined by the sense of descriptions provided by a language.

Hence, no amount of purely qualitative evidence for a sentence containing a proper name or a deictic (which for Kripke is also a rigid designator) can succeed in establishing its truth.(12)

Likewise, Lyotard accepts that a proper name is not determined by a set of descriptions (Lyotard, 54). In other words, for him, a cognitive phrase cannot supply the content of a name. In addition, Lyotard also follows Kripke in holding that the identity of the referent is supplied by the name, not the sense. In turn, it is because a name has no sense (Lyotard, 54) that Lyotard is able to claim that it is "empty" (Lyotard, 66). The very emptiness of the name gives it its rigidity; and this rigidity entails its independence from any "current phrase" in which it may be situated. Within procedures of validation, the name thus becomes the "linchpin" between phrases belonging to heterogeneous regimens (Lyotard, 66, 67).

That proper names perform the above role, however, entails that the reality of the referent of any particular name is not guaranteed by the existence of that name.(13)


is FALSE" (Lyotard, 47, 67)

"The name `rigidly' designates across phrase universes, it is inscribed in networks of names which allow for the location of realities, but it does not endow its referent with a reality" (Lyotard, 63). Naming alone, for Lyotard, does not establish the reality of a referent. What is needed in order for this to be achieved is that the following procedures be observed: the referent must be "declared the same in these three situations: signified, named, shown" (Lyotard, 65). The reality of a referent, hence, has to be demonstrated by way of recourse to procedures of validation. Reality, then, is not given;" it has to be established by way of the above procedures. The signifying is done by means of cognitive phrases, the showing by ostensives, and the naming by nominatives, which constitute a world that is "a network of proper names" (Lyotard, 133).

Although the referent of a name is not a given, it is presented as if it were a "given" (Lyotard, 58). This presentation can occur because of the rigidity of the name, hence "the rigidity of the named is the shadow projected by the rigidity of the designator, the name" (Lyotard, 73). The apparent givenness of the referent (the named) is therefore a result of the stability of names.

Although the name, according to Lyotard, gives permanence to the referent, it is not "by itself a designator of reality" (Lyotard, 75). In order for this to be the case, senses provided by cognitive phrases must be associated with the name.(14) However, "the inflation of senses that can be attached to it is not bounded by the `real' properties of its referent" (Lyotard, 75). Such "`real' properties" do not exist independently of the phrase universes in which they are presented. They are, rather, products of particular phrases and will therefore change according to the regimen in which the name is situated. Hence, it cannot be determined a priori which senses will come to be associated with a name. To assume that the latter is possible would, for Lyotard, constitute an essentialism.

According to Lyotard, an essentialist is someone who believes that the senses expressed by cognitive phrases must already be "written into"[15] objects. Thus, an essentialist is committed to the view that objects have completely describable essences: "Essentialism conceives the referent of the name as if it were the referent of a definition" (Lyotard, 88). In Lyotard's terms, we would add, an essentialist is someone who takes the apparent rigidity of the named as a given. Facts, according to this view, already exist and are completely describable. Names, it follows, are ultimately dispensable, since the substitution of a definite description for a name would, in principle, always be possible (Lyotard, 106).

The essentialist conceives of reality as preexisting the act of phrasing and assumes the primacy of the definitional regimen. On this view a definition, if true, is a priori true. Thus, an essentialist would be someone who presupposes that there is an ontological reality (a "what") independent of and prior to claims concerning it (the "how") (Lyotard, 99). For Lyotard, in contrast, a phrase is a "what:" "The necessity of there being And a phrase is not logical (the question "How?") but ontological (the question What?") "Lyotard, 103). This should not, however, be taken to mean that language constitutes the entirety of reality; merely that no reality can be given independently of language.

Kripke's Ontological Argument

As we have shown, Lyotard's conception of the function of the name owes a significant debt to Kripke's account of it. Nevertheless, in direct contrast to Lyotard's critique of essentialism, Kripke is a figure whose influence in the sphere of analytic philosophy has in fact been to revive essentialism. Furthermore, Kripke appears to derive his essentialism from the very theory of names that Lyotard uses to advance anti-essentialist claims.

Although in Naming and Necessity Kripke does not explicitly consider the issue, it appears he believes that essentialism must be true in order for language in general and names in particular to function in the way in which they do. It is possible to read Kripke as presupposing that essentialism is the answer to the question: "How can we show any language at all ... to be possible?"[16] Kripke does not explicitly address the problem, but the answer to this question, it appears, takes the following form: (i) language succeeds in referring; (ii) it does so by means of names, which are rigid designators; (iii) a rigid designator, in order to refer, must always refer to a stable referent; hence (iv) there must be things that exist independently of language and that, in turn, validate its use; and (v) these things must have properties that make them what they are.

Essential properties, according to this view, may be attributed to things a posteriori. Although one cannot deduce a priori what properties a referent may have, the fact that language has succeeded in naming a thing implies that there must exist essential characteristics that determine what that thing is. If there were not, according to this position, language simply could not work.

Kripke's essentialism is in many ways different from previous essentialist positions insofar as he holds that there are essentialist claims that can only be known to be true aposteriori (Kripke, 104, 125). Prima facie, then, Kripke would seem to be advocating an essentialism that does not succumb to Lyotard's critique. We may, however, wish to claim that this is not the case.

Kripke's argument presents the possibility of what one might term an "a posteriori deduction" of essential properties in which the universal realm of preexisting, stable entities is generated from particular acts of nomination. This is the case with Kripke's account of "initial baptism"' (Kripke, 96-97). Here, an object is named or the reference of a name is fixed by means of a description. In order to use the proper name correctly, however, Kripke asserts that it is necessary that the addressee who in turn hears it uses it to mean the same thing as was intended by the addressor from whom it was heard; in this way it is linked onto in a correct manner (and maintains its sense). In order to do this one "takes the notion of intending to use the same reference as a given" (Kripke, 97). Indeed, it is principally through a chain of linking that the referent is determined and maintains its stability (Kripke, 135). Fixing the reference, however, is also a matter of providing a definition. Imagine an initial baptism of a "substance" which we come to call gold:" here, says Kripke, "we must imagine it picked out as by some such definition' as, Gold is the substance instantiated by the items over there, or at any rate, almost all of them."' Such a definition does not express a (completely) necessary truth" in terms of the identity it entails; the existence of gold is not dependent on the existence of any of the items, whose selection is a matter of contingency. What the definition does do, though, is "express an a priori truth ... it fixes a reference" (ibid.). Now, although what is selected as a means of fixing is contingent (since one selects an "accidental property" (Kripke, 55) as a means of doing it) the actual act of fixing a reference is not. Underlying the a posteriori essentialism, then, is a further assumption with regard to the way in which definitions must work as part of the necessary condition of language use. The a priori assumption that the act of fixing the reference entails giving a definition lurks at the heart of the assertion of a posteriori essentialism. The definitional regimen, in Lyotard's terms, is always already assumed to have pre-eminence.

It is by means of a stipulative definition that Kripke identifies substance or chemical composition as the essence of gold. Hence, the claim that its chemical structure is essential to gold follows simply from the conventions regulating the use of the term "gold." The empirical a-posteriori element is deployed only at the point at which a particular chemical structure is picked out.[17]

Lyotard's Essentialism

In contrast to Kripke, Lyotard, we may conclude, thinks that he has successfully avoided locating the referent in a manner that presupposes that it may be described according to its qualities. In other words, he thinks that he has successfully avoided the essentialist trap. Properties are divorced from the name which, as a now "empty designator," serves as little more than a formal structure, a "linchpin," into which properties may be poured, but from which they may not be so easily derived.

Significantly, however, essentialism in fact raises its head in the shape of the differend itself. Whereas proper names, as rigid but nevertheless empty designators, are basically stable, the differend, in contrast "is the unstable state and instant of language wherein something which must be able to be put into phrases cannot yet be" (Lyotard, 22). A differend, in short, "refers" to what cannot as yet be brought under reference because of the lack of adequate phrases. In it, we may remember, something "asks" to be put into phrases and is rendered a victim because of the inability of a current phrase to present it.

A differend cannot be known although, following Kripke, it can be named. This is because, as we mentioned above, Kripke's theory allows successful reference to be made without actually knowing anything about what is referred to. Similarly, for Lyotard, it is possible to name the differend, even though it cannot be put into phrases. The "differend" is in fact a proper name (both the name for a linguistic relation, and the name of a book), and is therefore a rigid designator. More accurately, we might term it a metarigid-designator.

Lyotard avoids essentialism in the case of names, which are not differends, because he does not presuppose that naming establishes the reality of the referent. As we have said before, a cognitive and an ostensive phrase are also needed to achieve this aim. However, in the case of the differend we have a proper name that must by itself establish the reality of its referent, since no cognitive phrase can be used for this purpose. Furthermore, Lyotard never doubts the reality of differends: the demand for "new competences," whereby differends may be validated, attests to this.

Lyotard is not claiming that differends cannot be phrased in principle. Rather, their status as differends is relative to existing genres of discourse. However, since Lyotard holds that there are differends, he is commited to the view that he has already established their reality. Because this cannot be done by means of the cognitive genre, it must be brought about by the act of naming. But this is only possible if the referent of the name already has essential properties that make it a differend.[18]

We must conclude, therefore, that in this case naming alone secures the reality of the referent. Nevertheless, as we have shown in our criticism of Kripke, this cannot be done. It is an illusion bred by essentialism that this is possible; an illusion to which Lyotard himself falls prey.

In other words, in order to be recognized at all, the things that "ask" to be phrased must already correspond to and be subject to the "differend." Moreover, these things must have properties independent of the phrasing whereby they can be transformed from the negative deictics, indicated by their silence, to a positive modality, in which their being instanced and validated will finally result in a new rigid designator being supplied that is capable of fulfilling a function adequate to their presentation.

More significantly, however, the referent of the "differend" itself must be rendered stable by Lyotard in order to be named at all. It follows that the "unstable state and instant of language" that epitomizes the differend is in fact something else. It thus presupposes that the properties it both contains and to which it attests are necessary attributes or properties of its being what it is.

Hence, Lyotard has not escaped essentialism and, most significantly, he has not established the case for asserting the existence of differends.

[1.] Peter Sedgwick was a student of Christopher Norris, and Alessandra Tanesini was a student of Jonathan Bennett. [2.] Jean-Francois Lyotard, The Differend: Phrases in Dispute, trans. Georges van den Abeele (Manchester: Manchester University Press, 1988). All further references are given in the text with section number. [3.] See Saul Kripke, "Identity and Necessity," in Naming, Necessity, and Natural Kinds, Stephen P. Schwartz, ed. (Cornell: Cornell University Press, 1977), p. 88. [4.] Geoffrey Bennington uses the term "sentence" rather than "phrase." See, Lyotard.- Writing the Event (Manchester: Manchester University Press, 1988). [5.] What is not contingent but necessary, however, is that linkages must occur Lyotard, 40). [6.] Wittgenstein, Zettel, G.E.M. Anscombe and G.H. von Wright, eds., G.E.M. Anscombe, trans. (Oxford: Blackwell 1991), section 320. [7.] In this connection see Wittgenstein, Philosophical Investigations, G.M.E. Anscombe, trans. (Oxford: Blackwell 1988), section 497. [8.] An object of an idea is a totality about which nothing can be justifiably predicated. [9.] But note that the notion is subject to only the formal conditions of cognitive phrases. [10.] "Holistic properties are properties such that, if anything has them, then lots of other things have them too," J. Fodor and E. Lepore, Holism.a Shoppers Guide (Oxford: Blackwell 1992), p. 2. [11.] An expression is a rigid designator if it designates the same thing with respect to every possible world in which that thing exists. Kripke, Naming and Necessity (Harvard: Harvard University Press, 1980), p. 48. All further references are given in the text with page number. [12.] Kripke thus says: "being put in a situation where we have exactly the same evidence, qualitatively speaking, it could have turned out that Hesperus was not Phosphorus; that is, in a counterfactual world in which 'Hesperus' and 'Phosphorus' were not used in the way we use them, as names of this planet, but as names of some other objects, one could have had qualitatively identical evidence and concluded that 'Hesperus' and 'Phosphorus' named two different objects" (Kripke, 104). [13.] Davidson also notes this with reference to Kripke's position. See Davidson, A Coherence lbeory of Truth and Knowledge," in Truth and Interpretation, Ernest LePore, ed. (Oxford: Blackwell, 1986), p. 318. [14.] In addition, it needs to be associated with an ostensive phrase. [15.] Wittgenstein, Tractatus Logico-Philosophicus, D. R Pears and B. F. McGuinness, trans. London-Routledge, 1992),2.012. Lyotard paraphrases this in section 88 of The Differend. [16.] Kripke, Wittgenstein on Rules and Private Language (Oxford: Blackwell, 1982), p. 62. As Kripke himself notices this question has a Kantian flavor (fn. 48). [17.] Alan Sidelle reaches similar conclusions. See, Necessity, Essence, and Individuation (Ithaca and London: Cornell University Press, 1989). [18.] We thus disagree with an anonymous referee for this journal, who has argued that essentialism does not follow if one thinks of differends as phrasable in principle. We are grateful to him or her for raising this point.
COPYRIGHT 1995 North American Philosophical Association
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1995 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Jean-Francois Lyotard, Saul Kripke
Author:Sedgwick, Peter; Tanesini, Alessandra
Publication:American Philosophical Quarterly
Date:Jul 1, 1995
Previous Article:Mutual recognition and ethics: a Hegelian reformulation of the Kantian argument for the rationality of morality.
Next Article:A dilemma for any theory of knowledge.

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters |