Deleuze and Analytic Philosophy

Jeffrey A. Bell

Presented as the SEP-FEP Joint Conference

September 9, 2006

Fifty years ago H.P. Grice and P.F. Strawson published their now famous essay, “In Defense of a Dogma,” as in large part a reply to W.V. Quine’s equally famous essay, “Two Dogmas of Empiricism.” Both essays have set the agenda for a number of the debates that have arisen in analytic philosophy in the past half century. When Quine’s essay was first published in 1951, in fact, analytic philosophy was not nearly as dominant in the academy as it has since become. The pragmatist tradition in American philosophy was still, by 1950, quite strong in the United States; and in Britain the influence of Wittgenstein, who is seen by many American pragmatists such as Rorty as a kindred spirit, was also quite strong. In most graduate programs in philosophy today, by contrast, in the United States and Britain, one finds the clear, undisputed dominance of analytic philosophy. This dominance is not recent, moreover, for from the 1950s on analytic philosophy quickly acquired the dominance it has maintained to this day.

Rather than add to the litany of work that has either sought to explain the rise of analytic philosophy and the subsequent rift between analytic and continental philosophy (in which I include pragmatism), or the work that has attempted to overcome the difference between them altogether, we will focus, instead, upon concepts of philosophers in both the continental and analytic traditions. These concepts, it will be argued, need to be understood relative to the problems these philosophers are responding to, and these responses, as we shall see, are of use to philosophers in both traditions. It is in this spirit, then, that we return to the debate between Grice/Strawson and Quine, showing how the problems within this debate, problems which quickly became the dominant themes of the analytic tradition, were problems equally addressed by later philosophers both in the analytic tradition – especially the work of Saul Kripke and Hillary Putnam – and by those writing in the continental tradition – we will bring in the work of Gilles Deleuze and Bruno Latour (and in a long version of this paper, Alain Badiou). In setting these comparisons forth we will conclude, finally, that it would be to the detriment of both analytic and continental philosophers to remain indifferent to or even dismissive of the work going on in other traditions.

At the heart of the debate between Grice and Strawson and Quine is the analytic-synthetic distinction. Yet as Grice and Strawson readily admit, Quine’s objections to this distinction can be broadened to include much more: ‘Quine’s objection is not simply to the words “analytic” and “synthetic,” but to a distinction which at times philosophers have supposed themselves to be expressing by means of such pairs of words or phrases as “necessary” and “contingent,” “a priori” and “empirical,” “truth of reason” and “truth of fact”…’[1] If Quine’s objections were to hold up, therefore, then much more would be at stake, according to Grice and Strawson, then the analytic/synthetic distinction. For Grice and Strawson, though, Quine’s objections do not hold up, and they base their response to Quine upon a specific notion of meaning, what they refer to as “cognitive synonymy,” a notion they take to be part of the “analyticity-group.”[2]

Two expressions are cognitively synonymous, Grice and Strawson argue, is “roughly” equivalent ‘to what we ordinarily express by saying that x and y have the same meaning or that x means the same as y …’[3] For Quine, however, as Grice and Strawson read him, to mean the same as, when applied to predicate-expressions, differs from and goes ‘beyond the notion of being true of just the same object.’ The predicate-expressions, to use Quine’s example, ‘the creature with a heart’ and ‘the creature with a kidney’ may indeed be true by virtue of referring to one and the same object, and yet they are not ‘cognitively synonymous,’ much as Frege’s famous example of the morning star and evening star were not cognitively synonymous. In other words, it is not what is thought about that determines cognitive synonymity, but rather it is the way it is thought about that is the same, which is clearly not the case in the examples just given.

The problem, for Quine, is that there would be no example that would or could satisfy the conditions of cognitive synonymity, or analyticity as Quine extends his critique. The very notion of analyticity itself, Quine argues, derives from Kant, who ‘conceived of an analytic statement as one that attributes to its subject no more than is already conceptually contained in the subject.’[4] Among the problems with this understanding of analyticity is the manner in which an analytic statement can ‘conceptually contain’ all that is necessary without reference to anything outside, to something that is contained, and thus containment must be taken ‘metaphorically.’[5] In his use of the term, Kant appears, according to Quine, to take a statement to be analytic ‘when it is true by virtue of meanings and independently of fact.’ If taken in this way, which is precisely how Grice and Strawson understand Quine, then analytic statements can in no way be subject to revision for they already ‘contain’ the meaning that makes them analytic, and they do not refer to an outside or other which would prompt a possible revision. Quine is clear on this point: ‘Furthermore, it becomes folly to seek a boundary between synthetic statements, which hold contingently on experience, and analytic statements which hold come what may. Any statement can be held true come what may, if we make drastic enough adjustments elsewhere in the system.’ ‘Even a statement very close to the periphery,’ Quine adds, ‘can be held true in the face of recalcitrant experience by pleading hallucination or by amending certain statements of the kind called logical laws’; or, by contrast, even ‘the logical law of the excluded middle,’ a worthy candidate for an analytic, unrevisable statement, has been proposed for revision ‘as a means of simplifying quantum mechanics.’[6] In other words, the analytic distinction is to be understood as a difference of degree on a continuum of variation, whereby analytic are those statements most resistant to variation and synthetic statements most susceptible to variation. To maintain the distinction between an analytic, necessary, and unrevisable statement and a synthetic, contingent, and revisable statement is thus, for Quine, untenable. The analytic/synthetic distinction, in short, is made possible by what Quine will later refer to as the web of belief, and a proper empiricism, an empiricism without the dogmas associated with analytic and synthetic statements, will better appreciate the complexities associated with our knowledge claims.

Returning now to Grice and Strawson’s defense of the analytic-synthetic dogma, they argue that despite Quine’s arguments that efforts to define analyticity in terms of cognitive synonymy, or despite the fact that a formal, conceptually contained definition is not possible, it does not warrant the assertion that there is no meaningful distinction to be made here. ‘The fact,’ they say, ‘if it is a fact, that expressions cannot be explained in precisely the way which Quine seems to require, does not mean that they cannot be explained at all.’[7] In support of this view they offer the following example:

1. My neighbor’s three-year-old child understands Russell’s Theory of Types

2. My neighbor’s three-year-old child is an adult

Statement (1) is, however unlikely, verifiable. There is a way in which this statement can be taken to be true. Statement (2), by contrast, is something we cannot understand unless we undergo a wholesale revision of our concepts ‘child’ and ‘adult.’[8] The second statement would be an example, for Grice and Strawson, of an analytic statement. Barring a conceptual revision, no empirical evidence will lead us to accept that a three-year-old child is an adult; whereas the first statement could indeed be found to be true if such a child prodigy were found, in experience, to exist, and thus this would be an example of a synthetic statement. Quine’s criticism here, as Grice and Strawson read him, is that ‘those who believe in the distinction [between analytic and synthetic statements] are inclined at least sometimes to mistake the characteristic of strongly resisting revision … for the mythical characteristic of total immunity from revision.’[9] Stated otherwise, Quine’s assumption seems to be that ‘As soon as we give up the idea of a set of experiential truth-conditions for each statement taken separately, we must give up the idea of explaining synonymy in terms of identity of such sets.’[10] And yet, Grice and Strawson argue, we need not take a statement separately or even reject the notion of conceptual revision. In fact, they argue that ‘All we have to say now is that two statements are synonymous if and only if any experiences which, on certain assumptions about the truth-value of other statements, confirm or disconfirm one of the pair, also, on the same assumptions, confirm or disconfirm the other to the same degree.’[11] Furthermore, if we can make sense of the notion that the same form of words, given one set of assumptions, may express something true, and, given another set of assumptions, express something false, ‘then,’ they conclude, ‘we can make sense of the idea of conceptual revision. And if we can make sense of this idea, then we can perfectly well preserve the distinction between analytic and synthetic.’[12]

Before moving on it will be helpful to take stock of some of the key elements in the debate between Quine and Grice and Strawson. First, and most especially for our purposes, is the notion of meaning as a set which contains certain elements. As Quine understood an analytic statement, its meaning wholly contains the elements within itself and, because it is closed off to contingencies of empirical experience, it necessarily involves ‘the mythical characteristic of total immunity from revision.’ Grice and Strawson, as we just saw, argue that we can continue to adhere to the distinction between analytic and synthetic if we can make sense of the notion that conceptual set associated with a meaningful statement can be revised even though they may be strongly resistant to revision – e.g., the concepts child and adult. What remains consistent throughout this debate, however, is a continued adherence to the notion that there is a relationship between the meaning of a statement as an abstract set or property and the elements that embody the use of this statement, or the elements of experience that verify this meaning and/or give it content.

The continued adherence to this distinction becomes apparent in later work in the analytic tradition. In fact, it is perhaps not inaccurate to argue that this relationship between sets and contents of sets is the problematic that has generated much of the work in analytic philosophy – and hence the importance of Russell’s paradox (i.e., is the set of sets that are not members of themselves a member of itself?) and Tarski’s theory of truth that attempts to circumvent the paradox. Alain Badiou’s work in the continental tradition has also maintained an adherence to this distinction between sets and elements, understood in this context from a Cantorian perspective of infinite sets. Badiou does argue that the relationship between sets and elements is not a simple matter of sets including their elements; to the contrary, the Cantorian sets that most interest Badiou entail a fundamental impasse, an abyss without mediation, between the elements that belong to and are presented in the set and the unpresentable set of subsets that is included and represented within the set. Badiou’s work, however, as he himself admits, is nevertheless work that continues to develop within the analytic tradition the ‘mathematico-logical revolution of Frege-Cantor.’[13]

At this point we can turn to the work of Gilles Deleuze, for in his writings we find an effort to understand the relationship between an abstract set of rules on the one hand and the actual, material content on the other. Deleuze, however, offers an account that may well elucidate and extend the work of analytic philosophy in that he attempts, with his notion of the ‘abstract machine,’ to offer an understanding of abstract rules that are both inseparable from and dependent upon the actual, material systems of which they are the rules, and rules that are nonetheless distinct from and not to be confused with the material systems that actualize them. In a telling and appropriate passage for our purposes, Deleuze argues that ‘All methods for the transcendentalization of language, all methods for endowing language with universals, from Russell’s logic to Chomsky’s grammar, have fallen into the worst kind of abstraction, in the sense that they validate a level that is both too abstract and not abstract enough.’[14] Their logic and grammar is too abstract in that it separates—abstracts—the form from the content, the abstract set of formal, logical rules from the content of this set; and it is not abstract enough, Deleuze argues, ‘because it is limited to the form of expression and to alleged universals that presuppose language.’[15] In other words, the abstract logic and grammar, as confined to presupposing language, is unable, as an abstraction, to be applied to understanding other material processes. It is only good for understanding language, and hence it is not abstract enough.

What Deleuze offers as an alternative to the approach of Russell and Chomsky is the notion of an ‘abstract machine.’ The abstract machine does not presuppose the distinction between an abstract set and the content of these sets, or between an abstract set of formal rules and the concrete behaviours that are the embodiment of these rules. As Deleuze puts it,