Representation and Reality in Wittgenstein's “Tractatus.”

In deference to the extraordinary richness and dialectical subtlety of the Tractatus, José Zalabardo does not attempt a comprehensive treatment of the work. Instead, he focuses on two topics of unquestioned centrality: representation and reality. As it seems to me, he interprets what the Tractatus has to say—or apparently say—about representation more satisfactorily than its pronouncements about reality. It is not clear, however, to what degree this amounts to a criticism of Zalabardo's approach. For he denies that “faithful interpretation of Wittgenstein's thought” is his ultimate goal (9). In keeping with this disavowal, he claims license to develop the arguments he attributes to Wittgenstein “at a level of detail that goes beyond what the evidence could sustain” (9). In practice, however, Zalabardo does not merely develop certain nascent lines of Tractarian thought. His readings, while inventive and interesting, sometimes positively run counter to the plain meaning of the text. I will substantiate this charge shortly. But let me first say a word about what I find to be his more congenial discussion of the Tractatus's treatment of representation.The kernel of the Tractatus's account of linguistic and psychological representation is contained in what is usually called the “picture theory of the proposition.” The central idea of this theory is the thought that a proposition—itself a kind of fact—shares a “form” in common with the reality it depicts. This form is “logical form,” the least determinate species of “picturing form.” As Zalabardo notes, in treating the form of the representing fact as identical with the form of the represented fact, Wittgenstein departs dramatically from Bertrand Russell's conception in 1913 of the bearers of truth and falsity. This conception is enshrined in Russell's “multiple relation theory of judgment,” a theory according to which the form of the represented fact is a component of the representing fact. Zalabardo argues plausibly that Wittgenstein's reconceptualization of how the form of the represented fact figures in representation is one of the most distinctive contributions made by the picture theory (45–46). The picture theory provides an account of how the bearer of truth and falsity can furnish information about how the elements of the represented fact are to be combined—thus solving what Zalabardo calls “the mode of combination problem”—while at the same time allowing for the possibility of false representation. Because the Tractatus takes logical form to be a species of pictorial form, it treats a proposition as representing in precisely the manner of a picture. This suggests that the Tractatus really is presenting a picture theory of the proposition—a theory of the nature or essence of a proposition (see, by way of comparison, 4.016)—rather than, as is sometimes suggested, merely a picture “analogy.” Although Zalabardo does not emphasize this last point himself, his discussion certainly prompts this thought (see 46–56).I find myself less satisfied with Zalabardo's treatment of Tractarian views on reality. One source of this dissatisfaction is his practice of occasionally treating the remarks of other philosophers—for example, Russell (114–15) and Frege (112–14)—as if those remarks by themselves constituted evidence for the correctness of his interpretations of Tractarian remarks. Moreover, these extraneous texts are even sometimes accorded more weight than the testimony of the Tractatus itself. This happens most notably in connection with the question of the unity of the atomic fact. In 1913 Russell maintains that the work of unifying the constituents of an atomic fact is performed by one of those very constituents, namely, by the relation that unites the terms. Russell had earlier compared the items to be unified in a fact to the bricks in a wall, that which unifies them to the cement. Framed in terms of this image, Russell's view in 1913 is that every fact must feature some brick that (nonintrinsically) plays this cementing role.This proposal faces familiar difficulties. First, since Russell takes the unifying “brick” always to be a relation, the proposal provides no account of the unity of simple “subject-predicate” facts. Second, it invites the questions: What is it that cements the terms to the relating relation? Will this be a further relation? And if so, will this further relation also be a part of the fact? If it will, won't we be lumbered with an infinite regress?There are various ways of addressing these difficulties. We might claim—with Russell himself—that the copula expresses a relation. And we might permit facts to have infinite complexity, thus embracing the regress while denying its viciousness. The Tractatus, however, avails itself of neither recourse. Instead, it offers a wholly novel proposal:2.03 In the atomic fact objects hang one in another, like the links of a chain.In other words, there's no privileged constituent responsible for securing factual unity. Instead, the work of unification is distributed equally across the bricks in the structure, which stand in immediate combination. Naturally, such an account will leave some unmoved. One might worry, for example, that it leans too heavily on the uncashed metaphor of a chain. But, whatever its merits, the proposal is unmistakably Wittgenstein's stated answer to the problem of the unity of atomic facts in the Tractatus—an answer that addresses both the problem of the privileged element (there is none) and the problem of regress (since the links themselves make the connection with one another, there's no need for anything third to connect them—compare Wittgenstein's letter to Ogden, April 23, 1922).And yet, as if in simple defiance of Tractatus 2.03, Zalabardo claims that the Tractatus's response to the problem of factual unity takes an altogether different form. Wittgenstein, he maintains, is dissolving rather than solving the problem; for he is denying that facts are composite or compounded at all (108). They are rather “ultimate indivisible units” (108). And since objects are not components of facts (115), the question of how they are cemented together simply fails to arise (112).Undeniably, this suggestion has its attractions. It's apt to sound impeccably Wittgensteinian, after all, to urge that a problem is to be dissolved rather than solved. The stubborn truth, however, is that the reading's textual fit is poor. For one thing, one wants to ask: what then is the purpose of Tractatus 2.03 if not to state the Tractatus's answer to the question of factual unity? For another, one searches the Tractatus in vain for compelling evidence for this reading. Nor, indeed, does Zalabardo seem to want to claim that there is any compelling evidence. Instead, he rests the weight of his case on an invocation of Frege, whom he supposes to have anticipated the view he attributes to Wittgenstein in remarks such as the following: “I do not begin with concepts and put them together to form a thought or judgment; I come by the parts of a thought by analyzing the thought” (“Notes for Ludwig Darmstaedter,” quoted by Zalabardo, 113–14). Here Frege is saying that the whole thought is, in some sense, prior to its parts: we don't assemble a thought from its elements; instead, the thought (or judgment) is given to the mind as a whole and we somehow carve out its constituents subsequently. “A concept,” Frege says, “can be distinguished within, but not separated from, the complex in which it occurs (“On the Foundations of Geometry,” quoted on 113).If these remarks of Frege's about thought and judgment are to support Zalabardo's account of the unity of (an atomic) Tractarian fact, the parallel idea would have to be that in such a fact the whole is somehow prior to its parts. Such a thought would indeed serve to dissolve the problem of unity so long as we could arrive at a clear, nonmetaphorical understanding of the notion of “priority” involved. But one has to wonder whether this idea is really the one Zalabardo means to be attributing to the Tractatus. For the Fregean point involves viewing a proposition as actually having parts, albeit ones that are subsequent to the whole. Zalabardo, by contrast, seems to be suggesting that for Wittgenstein neither facts nor propositions have parts at all. For he insists that Wittgenstein declines to regard them as composites (108). This further thought, however, simply fails to follow from the Fregean idea to which Zalabardo appeals. Worse, if the Fregean idea is correct, then the problem of factual unity is (dis)soluble without making the stronger claim that facts lack mereological structure. That claim, accordingly, looks otiose. Worse still, if we give the texts their due, we must conclude that, contra Zalabardo, the Tractatus does not in fact conceive of facts as lacking parts. At 2.011, for example, Wittgenstein says, “It is essential to a thing that it can be a constituent part [Bestandteil] of an atomic fact.” And similar remarks occur elsewhere (Tractatus 3.31, 3.35, 4.024).In view of this apparent lack of textual fit, charity might recommend drawing a distinction between the “parts” of a fact and its “constituents.” The idea would be to understand “parts” as components existing prior to the wholes in which they figure, “constituents” as existing posterior to those wholes. Then Zalabardo might perhaps be read as suggesting that, although Wittgenstein denies that facts contain “parts,” he does take them to contain “constituents.”As far as I can tell, however, Zalabardo would decline this more moderate line of interpretation. For he sees the Tractatus as implicitly recommending that we regard such entities that the early Russell would view as parts of propositions as what the Tractatus calls “features” or “expressions” (112). And he seems to suggest that these “features” or “expressions” are not to be thought of even as constituents “distinguished within” facts (in Frege's phrase). Rather, they are to be understood as properties of them. This view, however, also seems to run contrary to the text.3.31 Every part [Teil] of a proposition which characterizes its sense I call an expression (a symbol) (emphasis added).It's not at all clear then that “expressions” are properties of facts as opposed to parts or constituents of them.There is much more that could, and ideally should, be said about this rich and interesting book. And the hesitations I've expressed here must be placed in the context of my generally favorable view of the work. It contains numerous inventive and patiently developed lines of thought and treats the Tractatus as the work of a philosopher rather than a spiritual master. It deserves to be read with care by every serious student of Wittgenstein

Contemporary philosophy · Epistemology · General interest · Political science · Representation (politics) · Law · Philosophy · Wittgensteinian philosophy and applications

• Tractarian Functions are not Propositional Functions

  Open Access•Clare Hay•History and Philosophy of Logic•2025

  Add to list