170. Oskari Kuusela, Wittgenstein on Logic as the Method of Philosophy: Re-examining the Roots and Development of Analytic Philosophy

COMO CITAR:

Silva, João Esteves da. «Oskari Kuusela, Wittgenstein on Logic as the Method of Philosophy: Re-examining the Roots and Development of Analytic Philosophy». Forma de Vida, 2021. https://doi.org/10.51427/ptl.fdv.2021.0053 .

DOI:

https://doi.org/10.51427/ptl.fdv.2021.0053

João Esteves da Silva*

As Aristotle makes clear, theoretical science concerns itself with anything we seek to know for its own sake. But what logic deals with is not sought for its own sake, but as a support to the other sciences. So logic is not classified as a major branch of theoretical science, but as something that serves philosophy, providing it with tools

Aquinas, Expositio super Librum Boethii de Trinitate

1. James Conant has described Oskari Kuusela’s first book, The Struggle against Dogmatism (2008), in the following way:

Rather than being framed around the assumption that the crucial difference between an early and a later Wittgenstein lies in their respective philosophical doctrines, it takes its point of departure from the assumption that early and later Wittgenstein equally aspired to practice philosophy in a manner which eschewed all doctrine. The book then seeks to articulate the crucial differences between early and later Wittgenstein in terms of the details of the respective ways in which they sought to realize such an aspiration. (Conant 2011, 623)

There Kuusela sought to expound Wittgenstein’s general conception of philosophy as “an activity . . . [that] consists essentially of elucidations” (TLP 4.112), that aims to clarify our thought and talk, not to put forward doctrines or theses. (“Philosophy is not a theory . . . [it] does not result in “philosophical sentences”, but in making sentences clear” (ibid.; amended translation).) Now, with Wittgenstein on Logic as the Method of Philosophy (2019), his second book, he further develops the picture laid out in the former and examines its place within the context of twentieth-century analytic philosophy. As the title indicates, the primary emphasis is on Wittgenstein’s philosophy of logic and metaphilosophy, which, from the point of view recommended by Kuusela, emerge as virtually one and the same. Arguably, his central claim is that the conception of philosophy as an activity of logical clarification, a Fregean and Russellian inheritance, should be understood as applying to all of Wittgenstein’s thought, unifying it to a considerable extent. Far from denying its radical transformations, what Kuusela proposes is an alternative assessment of Wittgenstein’s development, notably as involving a largely unnoticed “paradigm shift” (Kuusela 2019, 143) in philosophical logic. By drawing attention to such a shift and offering an account of the novel paradigm he identifies in Wittgenstein’s later works, Kuusela also makes a case for a most needed change in contemporary analytic philosophy, through a renewal of its historical and metaphilosophical self-consciousness, as well as the adoption of Wittgensteinian logical methods. As he puts it, “this book . . . aims to rewrite parts of the history of analytic philosophy in order to uncover paths to the future that previous histories have covered up.” (Kuusela 2019, 1)[1]

2. Kuusela devotes most of the first half of his book to the Tractatus. As he himself acknowledges, his non-metaphysical, logically-oriented reading belongs to the family of resolute interpretations, for it rejects the ascription of “any ineffable truths or nonsensical theses to the Tractatus” (Kuusela 2019, 39), i.e., the familiar (though hardly intelligible) story that takes it to put forward a theory that specifies the conditions of sense, which itself implies the nonsensicality of the sentences that seemingly articulate it—sentences that are nonetheless supposed to convey ineffable truths.

According to Kuusela, the Tractatus’ “two key insights” (Kuusela 2019, 37), both responding to problems Wittgenstein saw with Frege’s and Russell’s accounts of logic, are: 1) logical necessity finds its proper expression through the encoding of the logical principles that govern our thought and talk into the structure of a logically perspicuous notation, not by means of (purportedly true) philosophical doctrines or theses; 2) logic does not prescribe what can and cannot be said, for we, speaking beings, possess a pre-theoretical understanding of the principles of logic and the ability to distinguish sense from nonsense, and so it “must take care of itself” (TLP 5.473). On this view, logic (in the sense of “the object of investigation of the discipline of logic”) cannot be an object of factual statements[2], for it is internal to any referring expression (see Diamond 1991), nor can it be justified with reference to facts, for its principles are already relied upon in making any sort of statements and giving justifications (cf. Sheffer’s logocentric predicament[3]).

This is to say that logic (in the sense of “the discipline of logic”) cannot be a substantial science[4]: it does not provide us with any new (propositional) knowledge, but seeks to remind us of the know-how involved in our activities qua speaking beings, by rendering it perspicuous using a proper Begriffsschrift.[5] Logic can thus be understood not as the systematization of reason, as Russell wanted it to be, but as its clarification (or explicitation). As for its “justification”, the most that can be said here is that a logical notation is justified insofar as it proves reliable in its application as a clarificatory tool, without giving rise to anomalies and contradictions; these are the terms in which correctness is to be assessed in logic, not as a matter of correspondence between logical statements and some alleged logical facts.[6]

3. In order to bring out the significance of Kuusela’s reading among contemporary debates on the Tractatus more clearly, let me place it alongside another resolute reading, Conant’s, and draw attention to a few affinities and contrasts between them. In his essay “Wittgenstein’s Later Criticisms of the Tractatus” (2005), Conant presents three different lists, each containing a different set of views found in the Tractatus, as a means of illuminating Wittgenstein’s philosophical development, including what he later came to see as flawed in his earlier work. The first list includes metaphysical views often attributed to the Tractatus, which Conant sees instead as pieces of self-consciously employed nonsense that are supposed to help the reader in recognising the illusory nature of the external standpoint envisaged by metaphysical theorising.[7] The second list presents views Conant takes Wittgenstein to have been committed to when writing the Tractatus, not as explicit doctrines or theses but rather as preconceptions implicit in his conception of philosophical elucidation, notably by privileging a particular canon of analysis; it “illustrates the extent to which, from the standpoint of his later thinking, there was an entire metaphysics of language tacitly embodied in his earlier method of clarification” (Conant 2005, 190). As for the third and final list, it contains views that, when purged of all the “hidden [Tractarian] dogmatism” (ibid.), can be unproblematically ascribed to the later Wittgenstein, too.

As far as I can see, Conant and Kuusela largely agree about the second and third lists.[8] There is, however, a fundamental disagreement between them about the first one, namely on what concerns the particular status and function of its seemingly metaphysical statements. (In fact, according to Kuusela’s reading, some of that list’s statements are not to be found in the Tractatus at all.) For while Conant distinguishes between two kinds of Tractarian sentences, i.e., deliberate elucidatory nonsense, on the one hand, and sentences by means of which Wittgenstein outlines his framework of analysis and characterises his own activity, Kuusela draws no such distinction. Rather, he sees the majority of the book’s sentences, “though perhaps not all” (Kuusela 2019, 90), as contributing to the task of introducing, “by using a [transitional] material mode of speech, logical concepts and principles constitutive of a logical language that enables us to philosophize in the formal mode” (Kuusela 2019, 10), as he explains in Carnapian terms. This is also to say that they disagree about Wittgenstein’s use of “nonsensical” [unsinnig] in the much debated 6.54 passage[9]: neither takes it to be a technical term, and certainly not as denoting a property possessed by certain sentences which makes them illegitimate according to a supposed Tractarian theory of sense, but while Conant takes it straight (or “austerely”), i.e., as indicating “simply, and open-endedly, a failure to make sense” (Sullivan 2004, 34), Kuusela treats such sentences in a way more akin to how Frege thought of some of his elucidations, i.e., as imprecise modes of expression, untranslatable into a proper notation, that may nevertheless play a positive role in helping the reader make the transition from the logically defective medium of ordinary language to a logically perspicuous one.

Seeking to illustrate his view of Tractarian elucidatory sentences, Kuusela considers, among other examples, the book’s opening “ontological story” (Ryle 1997, 101):

When in the beginning of the Tractatus Wittgenstein states that the world is a totality of facts, not things (TLP 1.1), he is saying that, from . . . the point of view of his logical notation this is what the world is. . . . [His] apparently metaphysical account of the nature of reality is really a component of an account of language and thought . . . whose proper expression is his notation. (Kuusela 2019, 91; my emphasis)

He then goes on to add that the “core of this account is an idea of representation and reality possessing an identical logical structure” (Kuusela 2019, 91-2), but even this, the so-called isomorphism thesis, found among Conant’s first list as a piece of plain nonsense to be overcome, he still sees as part of Wittgenstein’s way of explaining features of his notation: in short, a framework that treats ordinary declarative sentences, contingent representations of reality, as truth-functions of elementary ones, which are in turn treated as concatenations of names that stand for objects. (Notice that treating A as B is quite different from claiming that A is B.) Ultimately, what committed the Tractatus to a metaphysical doctrine was rather its dogmatic adoption of such a framework of analysis as a canonical one, i.e., as the framework.

Kuusela also points out that Carnap himself had begun to pursue a similar line of thought in trying to make sense of Wittgenstein’s puzzling metaphysical-looking statements: in Logical Syntax, he remarks that “many . . . sentences of his which at first appear obscure become clear when translating into the formal mode of speech” (Carnap 1967, 303), such as 1.1 when translated as “Science is a system of sentences, not of names” (ibid.).[10] Though not all of Wittgenstein’s elucidations admit of such translations, Kuusela thinks Carnap was on to something, which he seeks to clarify and further elaborate.[11]

Now, how plausible is this? On the one hand, Kuusela’s proposed interpretation has the advantage of deflecting the vexata quaestio—faced by reading’s such as Conant’s—of what are exactly meant to be the (plainly) nonsensical sentences that form rungs of the ladder to be (resolutely) thrown away.[12] (It may also turn out to be easier to gather external evidence in its support.) On the other, however, as the qualification “though perhaps not all” quoted above indicates, some work remains to be done in order to account for the status and function of the sentences least likely to be seen as elucidations of a notation’s features (the remarks on ethics or mysticism, for instance), as well as on how these are supposed to fit together with the rest of the book. Moreover, one may sense a tension between Kuusela’s understanding of “nonsensical” in 6.54 and the remarks (such as 5.473-5.4733) that seem to advocate a so-called austere view of nonsense. Though it remains to be seen whether such potential difficulties can be addressed in a fully satisfactory way, Kuusela’s overall account of the Tractatus has managed to take the conversation a step further, which is itself an impressive achievement.

4. Most of the book’s remainder is devoted to Wittgenstein’s post-Tractatus development. Chapter 4 focuses on the transitional period in the 1930s, and chapters 5 and 6 mainly on the Investigations. There Kuusela explains Wittgenstein’s turn from a methodological monism to a methodological (and logical) pluralism, in an attempt to break free of the dogmatic commitments that were still present in his earlier philosophy. In particular, he offers an account of Wittgenstein’s later conception of clarification as both preserving logic’s ideals of simplicity and exactness and doing justice to the actual complexity and messiness of the phenomena under investigation. This involves an expansion of logical methods beyond calculus-based ones, which, contrary to what has often been supposed, are not altogether rejected by the later Wittgenstein, but simply recognised as insufficient for certain, though not all, clarificatory purposes. In order to avoid falling back into dogmatism, the “crystalline purity of logic” (PI §107) is to be regarded as a feature of a form or mode of investigation, not to be projected onto its object as something it must in reality possess, even if at an underlying level.[13] That is, we can fruitfully compare our objects of investigation, such as linguistic expressions, with calculi and other logical models in order to render, through similarities and dissimilarities, some of its features perspicuous, notably the ones that are relevant to our dealing with particular philosophical problems, though without going as far as forcing such comparisons into doctrines or theses.

Wittgenstein’s later conception of logic, Kuusela argues, is neither aprioristic, conventionalist, nor empiricist. It rejects the a priori as “a grammatical illusion” (apud Kuusela 2019, 133), springing from the tendency to confuse statements regarding modes of representation, expressing exceptionless necessities, with statements about the objects being represented. Still, this does not make Wittgenstein a conventionalist, for his view is simply that linguistic conventions, such as grammatical rules, are the proper expression of logical necessity, not its source (see pp. 135-6). Here the continuities with the Tractatus are unmistakable, as the earlier Wittgenstein, taking the logocentric predicament seriously, had already sought to dislodge the frame of mind from which metaphysical questions concerning the source (or ground) of necessity arise, or show that necessity is not properly expressed by means of factual statements. Besides, Wittgenstein now recognises that, given that logic is chiefly concerned with clarifying uses of words (concepts) and that such uses are an empirical matter, our natural history is after all important for logical investigation. This is what he had in mind when remarking: “We are talking about the spatial and temporal phenomenon of language, not about some non-spatial, non-temporal phantasm” (PI §108). But how does this not entail a collapse into a form of empiricism (such as Ramsey’s psychologistic logic)? Again, the key here is to recognise Wittgenstein’s natural historical remarks as part of his modes of investigation, not as empirical statements.

In this vein, Chapter 5 provides an account (and illustration through concrete examples) of the method of language-games as “a method that enables a logician or philosopher to isolate for study, by means of simple or simplified models of language use, facets of complex uses of linguistic expressions embedded in human activities” (Kuusela 2019, 142), thus constituting an expansion of logic to areas where Fregean-Russellian or model-theoretic logics are of limited application. And Chapter 6 characterises an even more radical expansion, i.e., the introduction of what Kuusela calls the quasi-ethnological method, which employs as clarificatory models “real or fictional natural history—for example, accounts of language learning or descriptions of fictional tribes of language users” (Kuusela 2019, 183). Just as the Tractarian notation, all these different models are to be judged according to their contribution to specific clarificatory tasks, not in terms of empirical accuracy.[14] So, despite his conception of language as (part of) a form of life, with its emphasis on primitive reactions, as well as on activities and cultural circumstances, Wittgenstein should not be seen as committed to any doctrines or theses about language. At most, he can be said to be committed to holding that our objects of logico-philosophical (or “grammatical”) investigation are usually, though not always necessarily, better apprehended from the point of view of such a conception.

It is also worth emphasising a short section (6.4) where Kuusela criticises readings that take Wittgenstein to articulate conditions of possibility of language use, hence falling back into a form of apriorism (and dogmatism). There we see him at his best both as a scholar, invoking textual evidence indicating that Wittgenstein was himself concerned to criticise such Kantian-style projects[15], and as a philosopher, arguing, in a Wittgensteinian spirit, that the appeal to such purported conditions is empty:

Insofar as something is an essential feature of a language-game, the feature cannot be appealed to as a condition of possibility of the game, because we have no grasp of what is to be explained independently of the feature that was meant to explain it. . . . Ultimately, to say of an essential feature that a thing must possess it in order to qualify as such and such has no more normative force than uttering a tautology such as ‘in order to be one metre long a stick must be one metre long’. (Kuusela 2019, 210)

We have, then, an outline of Wittgenstein’s “non-empiricist naturalism” (Kuusela 2019, 192), a novel paradigm in philosophical logic in which attention to concrete instances of our life with words is key. Very importantly, as Kuusela emphasises in a short epilogue (pp. 245-6), this revolution is a peaceful one, in the sense that, from its perspective, philosophical doctrines and theses may be rehabilitated qua clarificatory models or “objects of comparison” (PI §130) that can still serve a positive function. (Think of Elizabeth Anscombe's remark that “there is surely something right about . . . [the Tractarian account of sentences] if one could dispense with ‘simples’ and draw the limits of its applicability” (Anscombe 1971, 77), despite being too simplistic if taken as a generally applicable model.) For though philosophers often go astray with their wild generalisations, this is far from saying that valuable insights cannot be found in the midst of their theorising. “There is therefore no need for burning books” (Kuusela 2019, 246), unlike what other revolutionaries, such as Hume or some of the logical positivists, had claimed.

On this account, it also becomes clear that it is somewhat misleading to regard the later Wittgenstein as an ordinary language philosopher, for his approach to logico-philosophical clarification can be seen to incorporate features of both the ideal and ordinary language traditions, such as, respectively, the employment of formal methods or concerns with simplicity and exactness, on the one hand, or the importance of attending to (and describing) the uses of natural language (and its contexts) in their actual complexity, on the other. Wittgenstein’s example thus helps to fulfil Kuusela’s aim in Chapter 7, i.e., to dissolve the enduring dispute between these two seemingly opposed traditions of analytic philosophy, by showing them to be largely complementary, not opposed.[16]

5. The aforementioned chapters on Wittgenstein’s later writings all raise interesting (and intricate) issues. These include, for instance, the far from straightforward empirical-conceptual distinction, Wittgenstein’s take on necessity, his relation to the so-called linguistic turn, or the question of how successful his later struggle against dogmatism actually was. The very issue of empirical accuracy in logical models and the idea of offering not just negative criticisms but also positive contributions, as raised by Martin Gustafsson in his review of Kuusela’s book (see Gustafsson 2019, 7), is also a pressing one, among various others. Alas, I shall have to leave any discussion of these matters, which would all require careful treatment, to another occasion, for I cannot even begin to do them justice within the scope of the present review. I can nevertheless, as a conclusion, draw attention to a general feature of Kuusela’s approach that seems to me to encapsulate both some of its main virtues and limitations.

Overall, Kuusela’s book is an excellent piece of scholarship, but it is more than that. It is a work of “history of philosophy done philosophically” (Burnyeat 2006, xiii), whose author emerges as a lucid and realistic thinker, not at all given to metaphysical speculation.[17] As he makes clear right away in the introduction, his aim is not simply to provide what he takes to be a viable reading of Wittgenstein: he clearly believes that Wittgenstein’s approach, as he understands it, constitutes a viable way of doing philosophy. He hence seeks to elucidate Wittgenstein’s writings as a means of outlining a metaphilosophical program to which he is himself committed. And, interestingly, it is as if he ends up applying this very program to the task of interpreting Wittgenstein. Kuusela’s object of investigation—Wittgenstein’s body of work—is a highly complex and rather messy one, and, in order to throw light into it, he develops a particular model, Wittgenstein-the-logician, which allows us to attain a perspicuous overview of an otherwise winding landscape. (I cannot praise Kuusela’s clarity of exposition enough.) Such a model, of course, does not capture the full landscape, but provides a useful map to explore it, among other possible ones. For instance, with Kuusela’s Wittgenstein, to whom simplicity and exactness are key, gone is much of the Investigations’ (far from accidental) pathos, masterfully grasped by someone like Stanley Cavell. But given the nature of Kuusela’s project, it would be unfair to say that he fails to do justice to the actual complexity of its object of investigation. Besides, it has the advantage of being helpful not only to devoted Wittgensteinians but also to those whose temperament is not quite suited to that kind of pathos.[18]

[1] Some of Kuusela’s main concerns as a reader of Wittgenstein are thus in important respects close to those of earlier ones such as Rush Rhees or Peter Winch. In his 1966 review of George Pitcher’s The Philosophy of Wittgenstein, Rhees had argued against the dominant two-Wittgensteins view and emphasised that, contrary to what most have assumed, the Investigations, too, “is a book on the philosophy of logic” (Rhees 1970, 37). Starting with his introduction to Studies in the Philosophy of Wittgenstein (1969), Winch took a similar path, identifying Wittgenstein’s logical interests as a fundamental continuity in his thought, which he also saw as opposed to the “disastrously mistaken [view] . . . that we are dealing with two different philosophers” (Winch 1969, 1), each with his own doctrines. In particular, he took the much broader scope of Wittgenstein’s later investigations to reflect his new conception of logic.

[2] Kuusela also draws attention to the fact that Wittgenstein came to hold this view before developing his account of sentences, which is to say that his conception of logic as insubstantial should not be seen as following from that account. (See p. 43.) Rather, his “picture theory” is better understood as part of his general attempt of elucidating the nature of logic, namely by contrasting logical statements with factual ones.

[3] “In order to give an account of logic, we must presuppose and employ logic” (Sheffer 1926, 228). In Chapter 2, Kuusela expounds the early Wittgenstein’s solution to such a predicament.

[4] This point is connected with Wittgenstein’s Grundgedanke, i.e., that the so-called logical constants, unlike names, do not stand for anything in reality (TLP 4.0312), as Kuusela explains (see specifically pp. 52-54). And by allowing us to rewrite sentences in a way that makes connectives disappear, his notation embodies this particular insight, as it were. Wittgenstein’s intention here was not to argue against Frege and Russell, claiming that, say, “negation is the name of a second-level function” or “logical forms are abstract objects” are false. Rather, he sought to dissolve the very problems that such theses were trying to address: from the point of view of a connective-free notation, such problems no longer have the point they seemed to have.

[5] This distinction between knowledge-that and knowledge-how concerning logic is essentially what the Tractarian saying-showing distinction is about. The point is that, contrary to what proponents of irresolute readings have held, what (meaningful) sentences are supposed to show (but not say, i.e., represent) is not some ineffable content or quasi-content one is supposed to grasp somehow but that sort of practical, pre-theoretical understanding presupposed in our practices of inference and representation. (Kuusela explains this, in slightly different terms, in pp. 100-1.)

[6] Now, notice that nothing that has been said in the last couple of paragraphs is supposed to constitute a doctrine or thesis (i.e., a true-false statement or a body of such statements) about logic. These are rather insights or principles constitutive of the notation that the early Wittgenstein took to be satisfactory, before eventually recognising inconsistencies in it, such as the so-called colour-exclusion problem (see Wittgenstein 1993).

[7] See Conant 2002 for an account of this method.

[8] Actually, the second list can serve as a useful summary of much of Kuusela’s account of the Tractatus and its unwitting dogmatic commitments, so I find it worth quoting it here at length:

1.     The logical relations of our thoughts to each other can be completely shown in an analysis of our propositions.

2.     These relations can be displayed through the employment of a logically absolutely perspicuous notation.

3.     Through the employment of such a notation, it is possible for propositions to be rewritten in such a way that the logical relations are all clearly visible.

4.     A proposition must be complex.

5.     Every proposition can be analyzed.

6.     Logical analysis will reveal every proposition to be either an elementary proposition or the result of truth-operations on elementary propositions.

7.     All inference is truth-functional.

8.     There is only one logical space and everything that can be said or thought forms a part of that space.

9.     There is such a thing as the logical order of our language.

10.  Antecedent to logical analysis, there must be this logical order – one that is already there awaiting discovery – and it is the role of logical analysis to uncover it.

11.  By rewriting them in such a notation, what propositions our propositions are will become clear.

12.  By rewriting them in this way, it will also become clear what all propositions have in common.

13.  There is a general form of proposition and all propositions have this form.

14.  In its thus becoming clear what propositions are, it will also become clear how misleading their appearances are – how much the outward form disguises the real hidden logical structure.

15.  A logically perspicuous notation is the essential tool of philosophical clarification.

16.  Through our inability to translate them into the notation, despite their resemblance in outward form to genuine propositions, certain strings of signs can be unmasked as nonsense, i.e., as strings in which signs to which no determinate meaning has been given occur.

17.  All philosophical confusions can be clarified in this way.

18.  By demonstrating the significance of this tool and its application in the activity of clarification, the problems of philosophy have in essentials been finally solved. (Conant 2005, 189-90)

[9] “My sentences are illuminating [in that] anyone who understands me eventually recognizes them as nonsensical” (TLP 6.54; amended translation).

[10] It is here worth noticing that, likewise, Gilbert Ryle, an unfortunately neglected ancestor of non-metaphysical readings of the Tractatus, had suggested, in a posthumous paper, that “Wittgenstein’s ontological remarks are approximations in the material mode to what he is going to say later in the formal mode.” (Ryle 1997, 105)

[11] See chapter 3, “The Tractatus’ Philosophy of Logic and Carnap”, which contains some of Kuusela’s most original, if controversial, work. His account of the Tractatus-Carnap relationship also includes, for instance, an interesting comparison between the saying-showing distinction and Carnap’s distinction between object-language and syntax-language (see pp. 99-104).

[12] As Conant himself recognises, “the plausibility of a [Diamond-Conant] resolute approach to reading the book will depend partly upon how satisfying an answer to this question can be given” (Conant 2005, 179), and so it should definitely be pressed. Notice, however, that the way I have formulated it, in imitation of many critics of the resolute program, involves a misunderstanding. Given Conant’s account of the Tractatus’ method and its view of nonsense, there can be no such thing as a definitive determination of what are exactly meant to be the ladder’s rungs. The very idea of such a determination, as if one could approach the book from a point of view detached from the imaginative activity it is supposed to trigger, is itself part of the illusion that the Tractatus, on this account, seeks to undermine. And to be fair to Conant, his first list was already a fine attempt to provide at least an outline of how the issue can be addressed. Of course, much more can be done, and actually should be done, though without assuming that such a conversation can ever come to an end, or even to a point of overall stability.

[13] The Tractatus was guilty of precisely this kind of error. (See Chapter 4, “Ideality and Reality: Beyond Apriorism, Empiricism, and Conventionalism”, especially sections 4.1 and 4.2.)

[14] For instance, think of how the so-called shopping language-game (PI §1), which is not an empirically faithful description but an idealised model, manages to bring so clearly into view the different functions of the words “apples” (an object-word, used label-like), “red” (a property-word, used comparatively) and “five” (a number-word, used correlatively). As simple as this example is, it already does quite some work in dispelling the temptation to treat the meaning of every word as its reference. (See Kuusela 2019, 154.)

[15] Here are a couple of examples: “If humans were not in general agreed about the colours of things, if underdetermined cases were not exceptional, then our concept of colour could not exist.” No: our concept would not exist” (Z §351); “If someone says “If our language had not this grammar, it could not express these facts—it should be asked what “could” means here” (PI §497).

[16] And remember that someone like J. L. Austin, the ordinary language philosopher par excellence in the eyes of most, also employed idealised models, such as in his essay “How to Talk”. (See Austin 1961.)

[17] That is, speculation concerning purportedly ultimate foundations or justifications, as well as absolute generalities. This sense of “metaphysics” is thus quite different from that of, say, Aristotelian “first philosophy”.

[18] My thanks to Raimundo Henriques for his helpful comments on an earlier version of this review.

REFERENCES

Anscombe, Elizabeth. An Introduction to Wittgenstein’s Tractatus. South Bend, Indiana: St. Augustine’s Press, 1971.

Austin, John. “How to Talk: Some Simple Ways”, in Philosophical Papers. Oxford: Oxford University Press, 1961.

Burnyeat, Myles. “Introduction”, in Bernard Williams, The Senses of the Past: Essays in the History of Philosophy. Princeton, NJ: Princeton University Press, 2006.

Carnap, Rudolf. The Logical Syntax of Language. London: Routledge and Kegan Paul, 1967.

Conant, James. “Wittgenstein’s Later Criticisms of the Tractatus”, in A. Pichler and S. Säätelä, eds., Wittgenstein: The Philosopher and his Works. Frankfurt: De Gruyter, 2006.

Conant, James. “Wittgenstein’s Methods”, in O. Kuusela and M. McGinn, eds, The Oxford Handbook of Wittgenstein. Oxford: Oxford University Press, 2011.

Conant, James. “The Method of the Tractatus”, in E. G. Reck, ed., From Frege to Wittgenstein: Perspectives on Early Analytic Philosophy. Oxford: Oxford University Press, 2002.

Diamond, Cora. “Throwing Away the Ladder: How to Read the Tractatus”, in The Realistic Spirit: Wittgenstein, Philosophy, and the Mind. Cambridge, MA: MIT Press, 1991.

Gustafsson, Martin. “Book Review: Wittgenstein on Logic as the Method of Philosophy: Re-examining the Roots and Development of Analytic Philosophy, by Oskari Kuusela”, Mind, Vol. 0 . 0 . September 2019.

Kuusela, Oskari. Wittgenstein on Logic as the Method of Philosophy: Re-examining the Roots and Developments of Analytic Philosophy. Oxford: Oxford University Press, 2019.

Rhees, Rush. “The Philosophy of Wittgenstein”, in Discussions of Wittgenstein. London: Routledge, 1970.

Ryle, Gilbert. “Ontological and Logical Talk in Wittgenstein’s Tractatus”, The Linacre Journal, n. 3: 101-107, 1999.

Sheffer, Henry. “Review of Principia Mathematica, Vol. I, Second Edition, 1925”, Isis 8 (1): 226-31, 1926.

Sullivan, Peter. “What is the Tractatus About?”, in M. Kölbel and B. Weiss, eds, Wittgenstein’s Lasting Significance. London: Routledge, 2004.

Winch, Peter. “Introduction: The Unity of Wittgenstein’s Philosophy”, in P. Winch, ed., Studies in the Philosophy of Wittgenstein. London: Routledge, 1969.

Wittgenstein, Ludwig. Philosophical Investigations, Second Edition. Translated by G. E. M. Anscombe. Oxford: Blackwell Publishers, 1997. Fourth Edition. Translated by P. M. S. Hacker and Joachim Schulte. Oxford: Wiley, 2009. (PI)

Wittgenstein, Ludwig. “Some Remarks on Logical Form”, in Philosophical Occasions. Indianapolis: Hackett, 1993.

Wittgenstein, Ludwig , Tractatus Logico-Philosophicus. Translated by C. K. Ogden. London: Routledge & Kegan Paul, 1951. (TLP)

* PhD candidate funded by FCT (SFRH/BD/147507/2019). Program in Literary Theory, School of Arts and Humanities, University of Lisbon. joao.esteves.silva@campus.ul.pt

REFERÊNCIA:

Kuusela, Oskari. Wittgenstein on Logic as the Method of Philosophy: Re-Examining the Roots and Development of Analytic Philosophy. Oxford: Oxford University Press, 2019.