An elucidatory interpretation
of Wittgenstein’s Tractatus: A
critique of Marie McGinn’s reading of Tractatus
6.54
R.Read and L. Mulhall
The penultimate ‘proposition’ of Wittgenstein’s Tractatus Logico-Philosophicus, proposition 6.54, (in-)famously reads:
My propositions serve as elucidations in this way: he who understands me eventually recognises them as nonsensical, when he has used them—as steps—to climb up over them. (He must, so to speak, throw away the ladder after he has climbed up it.) He must overcome these propositions, and then he will see the world aright.
Any satisfactory reading of the Tractatus must be able to comprehend Wittgenstein’s enigmatic concluding insistence that his ‘elucidations’ have one and all been nonsense. There currently prevail two ways of interpreting this proposition, along with the rest of the Tractatus to which it promises to be an interpretive guide.
According to one of the two interpretations,[1] which could with justice be called the ‘traditional’ or ‘metaphysical’ interpretation, when one throws away the Tractatus, one throws away the text (i.e. one throws away its elucidatory propositions), but one holds onto what those propositions have elucidated for one, i.e. certain unstateable ‘truths’ about the way language works and how it is related to the world. According to this view, in 6.54 Wittgenstein is adding the finishing touch to a theme that has been lurking in the background since at least 3.221, but that really first gets stated as such at 4.115-4.1213: the distinction between what can be said and what can only be shown.
This interpretation can be characterised in somewhat more detail in the following way.[2] According to the Tractatus, (non-nonsensical) propositions have a sense because they are capable of picturing the facts: i.e. the situations which objects are in, in relation to other objects. As long as there is more than one thing in the world, the propositions of language work by mirroring the relationships in which those things stand. We can call the order in our propositions that makes them capable of thus picturing the facts ‘logic’. So logic can be seen as the order of possible situations in which things can find themselves, as reflected in language. Now the distinction between what can be said and what can only be shown arises as follows. Logic, as the order of possibilities for objects standing in relation to one another when they combine to form facts as represented in language, is not itself a fact.[3]
Rather, it is more like the condition by virtue of which there are facts at all—as 6.13 says, “Logic is not a theory, but a reflection of the world. Logic is transcendental”. Propositions work—they have sense—by virtue of the fact that they are capable of depicting how things stand in the world. But logic is not something, some thing or state of things, in the world, and so cannot be depicted in sense-bearing propositions. Rather it shows itself by and in the fact that we have sense-bearing propositions to begin with.
In trying to tell his readers all of this, in trying to elucidate the logical isomorphism that must exist between language and the world, the metaphysical interpretation of the Tractatus argues, Wittgenstein will have been committing himself to writing something that is self-conscious nonsense; something that, if it is read with understanding, can only be thrown away once it has been finished, leaving its readers with a set of philosophical insights that they previously lacked … but that cannot be articulated in words. Thus Wittgenstein’s insistence at 6.54 that the whole book has been nonsense and must be discarded. Nevertheless, according to this interpretation, reading the Tractatus leaves us with a (true but ineffable) residue of doctrine even after this discarding. This interpretation of the Tractatus, then, can also suitably be labelled the ‘ineffabilist’ interpretation.
Now, as a Frege-or-Russell-style philosopher[4] – an absolutely central representative of Wittgenstein’s intended audience -- reads the Tractatus, he finds the things that he would expect to be elucidated or theorized in a logico-philosophical treatise -- like the ontological background to the propositional calculus, the role of formal concepts and the nature of inference, all of the things in fact that could give him a handle on how language ‘hooks onto’ the world -- consigned to the category of the ‘ineffable’, or ‘shown but unsayable’. The ineffabilist claims that the central logico-philosophico questions have been answered by the adducing of ‘substantial’ or ‘profound’ or ‘illuminating’ nonsense (– that is, one has come to realise that these answers cannot, strictly, be expressed in words. Rather, perhaps, one displays one’s insights by the ways in which one will be inclined and disinclined to use language in the future.). Thus the ineffabilist interpretation can rightly be called ‘traditional’ in that it reads the Tractatus as being concerned to give answers to the type of questions posed by the Frege-and-Russell-style philosopher, albeit in an unorthodox way.[5]
But, as advocates of the newer [non-traditional] interpretation of the Tractatus have pointed out (see below for detail), this leaves ineffabilist readers in a deeply problematic position (unless they are prepared to find Wittgenstein himself guilty of a fundamental, unavoidable and horrendous incoherence), a position where they have failed to live up to the real challenge posed by 6.54. For one thing, the distinction between what can be said and what shows itself is developed in the main body of the text, so if one holds onto the idea that the Tractatus has helped one to grasp ineffabilia of one kind or another, one will not really have thrown the entire body of Wittgenstein’s propositions away. And in failing to throw the entire body of the propositions away, one will have lumbered oneself with the deeply suspect notion that one can have insights into reality that cannot be expressed in words, but that can be ‘pointed towards’, ‘gestured at’, ‘elucidated’ by nonsensical uses of language. But the fantasy that one can have any cognitive or linguistic access whatsoever to things that lie beyond the reach of language is one of which Wittgenstein is explicitly critical in the Preface to the Tractatus, where he maintains that the limit to what can be thought can “only be drawn in language and what lies on the other side will be simply nonsense”. [6] Here it appears that nonsense is not something that can indicate things that lie outside the reach of language but can be thought or displayed none the less. No; it is simple nonsense (einfach Unsinn).
The case against the traditional, ineffabilist interpretation is strengthened if we turn with a little more attention to the wording of the beginning of 6.54: “My Satze serve as elucidations in the following way: he who understands me eventually recognises them as nonsensical…” In 6.54 Wittgenstein does not invite his reader to understand his sentences—that would be impossible, they are simply nonsense—but to understand his intentions in writing them.[7] So what, then, were Wittgenstein’s intentions in writing the simply nonsensical sentences that make up the Tractatus? If nonsense cannot achieve by elucidation what we expect a logico-philosophical treatise to achieve, why engage in it? These questions can best be answered by taking a look at where all of this
leaves the Frege-and-Russell style philosopher.
As we mentioned above, a Frege-and-Russell-style philosopher, if he has followed what he takes to be Wittgenstein’s arguments, should (towards the end of the Tractatus) have realised that the position that he wanted to get out of his thinking about the relationship between language and the world can only be realised as a kind of position ‘outside of language’,[8] comprising insights that are (purportedly) real enough, but that can’t be expressed in words. That is to say, if a Frege-and-Russell-style philosopher agrees with what he thinks he reads in the Tractatus, he thinks he has arrived at the definitive philosophical position, the ultimate realisation of the philosophical project (only he can’t tell anyone about it). He will think that this is what Wittgenstein must have meant in the Preface to the Tractatus when he wrote, “I am…of the opinion that that the problems have in essentials finally been solved”. But 6.54, taken together with the rest of the Preface to the Tractatus, can most plausibly be read as intimating that to have a position that for principled reasons can’t be put into words is not to have a position at all. The insights that the Frege-and-Russell style would-be ineffabilist thought he could attain turn out to have been a mirage; the position he wanted (and perhaps thought the Tractatus had achieved) nothing more than the illusion of a position. What one comes away with after throwing Wittgenstein’s propositions away, then, is not a set of logico-philosophical insights, not a philosophical position, but rather the realisation that T L-P yielded only the illusion of a position, and that one has a tendency to be deluded -- led astray -- by certain ways of thinking and aspirations that turn out to be false when pursued rigorously. The tendency to read the Tractatus in this alternative (newer, anti-traditional) way can therefore be called the ‘therapeutic’ interpretation, in part because it aims thoroughgoingly to question the practice of a certain powerfully-influential philosophical mode of thinking, rather than, in however an orthodox or unorthodox fashion, to contribute to it.
In her recent paper in this journal [PQ], ‘Between metaphysics and nonsense: elucidation in Wittgenstein’s Tractatus’, Marie McGinn says that she is dissatisfied with both of the interpretative approaches to reading the Tractatus outlined above. She shares the worries about the traditional, metaphysical, ineffabilist approach that we have already laid out, which constitute the central motivation behind the therapeutic interpretation -- but she finds the novel alternative suggested by the therapeutic reading unsatisfactory as well. How, she asks, can a work of supposedly unintelligible nonsense be read as containing the insights necessary for us to realise that it is not communicating anything? She herself puts this ‘paradox’ as follows: “…[I]f the ladder by which we climb from unselfconscious nonsense to self-conscious sense turns out to be an illusion, how can we have got anywhere by climbing it?”[9] What is needed, McGinn argues, is an interpretation
which avoids the suggestion that there are ineffable truths about reality, but which allows that there is something behind Wittgenstein’s remarks; which permits these remarks to fall away completely, but which allows that the remarks accomplish something important; which avoids committing Wittgenstein to any metaphysical doctrines, but which does not fall into the paradox of self destruction.[10]
To this end, most of McGinn’s paper is dedicated to outlining a third way, which she calls an “elucidatory, or clarificatory, interpretation”[11].
Now, let us let common ground be common ground. For much of the above quotation from McGinn is not necessarily objectionable, even to someone very keen on the ‘therapeutic’ reading. For example, if by there being “something behind” Wittgenstein’s remarks is meant simply that the Tractatus is not a Post-Modernist sub-Existentialist joke,[12] then we are with McGinn on that point. The expression “fall away” is one we shall express a worry about, but the notion that Wittgenstein’s elucidations “accomplish something important” is unproblematic, if it means only (roughly) that this is a vital or even world-historical philosophical text, which consists essentially of elucidations (cf. T L-P 4.112).
Our paper seeks to test McGinn’s ‘elucidatory interpretation’, asking, among other things, whether McGinn’s practice actually cashes out her intentions as glossed here, and whether McGinn’s interpretation of the Tractatus actually warrants the title (‘the elucidatory reading’) that she wishes to give it.
There are two primary features of McGinn’s proposed ‘elucidatory’ reading that can serve to define it.
i) The first, which can be deduced from the desiderata for her reading cited above, is that it joins the ineffabilist interpretation in making saying vs. showing the central distinction of the work. McGinn sees the main work of the Tractatus as being ‘elucidatory’, and she seems to understand elucidation as a process by which an order is brought to the way in which we think about linguistic phenomena; an order with a particular aim in mind. The aim is to enable one “to see the phenomena of language in a new way, a way which no longer gives rise to philosophical problems; in which, for example, philosophical puzzlements concerning the status of logic, the relation between language and the world, or the relation between thought (the mind) and language, ‘completely disappear’ (PI 133).”[13]
McGinn wants to retain the saying vs. showing distinction, then, because she thinks that the Tractatus has elucidated something for its readers, although she makes it clear that what is elucidated should not be thought of as adding anything to the sum total of our knowledge in the sense that the ineffabilist readers of the Tractatus suppose. That is, on McGinn’s reading, Wittgenstein’s elucidations are not meant to give one access to otherwise ineffable ‘truths’; they do not add to one’s knowledge in that way. Rather, in bringing an order to the way in which one thinks about linguistic phenomena, the Tractatus should not be read as trying to tell one anything that one didn’t already know, but as bringing out a dimension of ‘knowing one’s way around’ what one already did know.[14]
Now, this ‘knowing one’s way around’ that McGinn reads Wittgenstein’s ‘elucidations’ as bringing about should not be thought of as contributing to the development of any theories. (In fact, the concept of elucidation, McGinn says, is “one that Wittgenstein intends to oppose to the concepts of explanation and theory construction”[15].) Rather,
It is in just this sense that Wittgenstein’s remarks are to be understood as elucidatory: their utility and significance are exhausted by their power to get the reader to see something familiar and everyday in a new light. Once the change in the reader’s perception…has been brought about, the remarks drop away, for they have no factual or descriptive content to sustain them.[16]
McGinn thus seeks to distance herself from the ineffabilist interpretation, which seems (paradoxically) to read the saying vs. showing distinction as indicating that Wittgenstein’s remarks do have a content of sorts, albeit inexpressible in propositions with sense.
ii) The second defining feature of the reading that McGinn proposes comes about as a direct result of the first. The importance that she places on the notion of elucidation as outlined above leads her to distinguish between what she calls the elucidatory “core” of the Tractatus, and those elements that cannot easily be read on the ‘elucidatory’ model that she proposes. Those remarks in which Wittgenstein “articulates his idea of philosophy, makes use of the comparison between saying and showing,…explores the role of logic”, and looks at ‘A believes that p’, solipsism and ethics, all “exemplify the idea of philosophy as ‘essentially elucidations’ (TLP 4.112)”[17] in McGinn’s sense. Excluded from legitimacy according to her reading are Wittgenstein’s (apparent) commitments to “the determinacy of sense, to a logically perspicuous symbolism, to simple symbols, to the logical independence of elementary propositions, [and] to the idea that all logical truths are tautologies”[18].
The idea behind this distinction that McGinn draws between the elucidatory “core” of the Tractatus, and those parts listed above that she sees as being essentially tangential (or worse), seems to be that Wittgenstein was misled by a certain picture of what logic must be like when he wrote the Tractatus; by a ‘mythology of symbolism’. And that when one removes those parts of the Tractatus that are expressive of his misguided commitment to this mythology from the elucidatory “core” of the work, one is left with something like Wittgenstein’s ‘real interests’ or ‘real views’, which span the Tractatus and the Philosophical Investigations. For example, one of the results of Wittgenstein’s elucidations in the Tractatus, McGinn concludes by saying, is that one is brought to see that logic cannot be separated from its application; another is that “what is essential to language is disconnected from the concepts of truth or falsity and agreement or disagreement with reality”[19]. So it is evident that, according to McGinn’s picture, once Wittgenstein abandoned the ‘mythology of symbolism’ and began to pay more attention to “the spatial and temporal phenomenon of language” (PI 108; see McGinn p.497), all of the pieces were in place for him to develop the position that has been called ‘the autonomy of grammar’, whereby “the potential for [ordinary language] to reveal ‘what kind of object anything is’ becomes increasingly significant philosophically”[20].
It will prove to be significant that McGinn justifies her division of the Tractatus into legitimate and illegitimate strands not on the grounds that the latter are nonsense, but because “they are expressive of certain theoretical preconceptions and are therefore not properly elucidatory”[21]. We agree with McGinn that Wittgenstein’s propositions are meant as elucidations, and that elucidations do not add up to any theory. (We also agree that there are deep and significant lines of continuity between the Tractatus and Wittgenstein’s later work, although we disagree with McGinn about what they are and where they are to be found – see below.) But in the present essay we want to take issue with McGinn’s understanding of what Wittgenstein means when he talks about elucidation.
We intend to argue that, despite the name that she gives her interpretation, McGinn shows herself to have an inadequate understanding of the role that Wittgenstein’s concept of elucidation plays in the Tractatus,[22] failing, in the end, to give a satisfactory reason why one should follow Wittgenstein’s injunction at 6.54 and throw his propositions away. We will argue that, while McGinn can perhaps hold onto a fairly stable reading of the main body of the Tractatus, this will yield a reading which altogether fails to do justice to Wittgenstein’s concluding insistence in 6.54 that the whole thing has been nonsense and should be thrown away (III). And we will conclude (IV) with some remarks about what doing justice to this insistence would involve.
We will suggest, in short, that McGinn’s approach runs the risk of being guilty of ‘quietism’, in the sense that she wants prematurely to quiet or stifle or shut-up philosophical desires (e.g. for a theory, ineffable or otherwise, of how language and world are related) which we suspect Wittgenstein wanted rather to get one self-consciously and happily to give up, if one can, if the wish to philosophise in certain ways dissolves within one. That, we suspect, is the meaning of 6.54.
What is McGinn’s account of Wittgenstein’s concluding insistence that the whole thing (including ‘saying vs. showing’) has been nonsense and should be thrown away?
We quoted McGinn above as saying that “Once the change in the reader’s perception [achieved by elucidations]…has been brought about, the remarks drop away because they have no factual or descriptive content to sustain them”. Her reason for excluding those parts of the Tractatus that are expressive of Wittgenstein’s (supposed) naïve commitment to a ‘mythology of symbolism’ (when he wrote T L-P) from legitimacy is not, on her interpretation, because they are nonsense, but because they are “expressive of certain theoretical preconceptions and are therefore not properly elucidatory”. So two reasons have been given by McGinn for why one should be inclined to throw propositions from the Tractatus away: either (1), they are elucidatory, and therefore have no descriptive content to sustain them (their value “depends entirely on their ability to induce [a] sense of clarified vision in us”[23]), or (2) they are expressive of “certain theoretical preconceptions”, and so illegitimate anyway.
We think that reason (1) is, in a way that we shall seek to fill out, reducible to the philosophically problematic ‘anti-theoretical’ stance implied by reason (2).
Towards the end of her paper, McGinn gives an account of a class of propositions that don’t have descriptive content according to her reading of the Tractatus, but should not be thrown away none the less, viz, scientific laws. The value of scientific laws does not depend on their content, but rather
their significance depends upon their application, that is, upon their being used as a means for constructing the propositions of science. The laws of mechanics, seen in this way, do not express necessary truths about the world, but are akin to synthetic principles which guide our construction of descriptions of the world, and regulate the transition from one form of description to another. It is not the system of laws itself that is important, but the precise way in which it helps us to construct true descriptions of the world.
On this view of matters, our accepting a particular set of views is comparable to our adopting a procedure for generating descriptions of the world, which can then be tested for truth. Clearly the question of truth, correctness or incorrectness, does not apply to the procedure itself, but only to the descriptions it generates.[24]
So what, for McGinn, distinguishes Wittgenstein’s content-free-and-therefore-disposable elucidations from the content-free-but-retained scientific laws? It seems reasonable to suppose that the answer must be something like that whereas the latter are, in a sense, genuinely and healthily theoretical, constituting the principles for a procedure that generates descriptions that can be tested for their truth or falsity, the former are not. The aim of the Tractatus, on McGinn’s reading, is to help dissolve our philosophical puzzlements by clarifying the nature of logic and of the logical isomorphism that must exist between language and the world. Its aim, then, is to clarify something prior to descriptions, and so not capturable in descriptions.[25] This might seem to be compatible with Wittgenstein’s reasons for saying in 4.112 that a philosophical work consists essentially of elucidations given above: if our elucidations are successful, one should realise that one doesn’t need to go on to form a theory on the back of them.
Elucidations, in McGinn’s sense, could perhaps be read as being akin (on one reading of the Investigations at least) to the “reminders assembled for a particular purpose” that Wittgenstein mentions in his later work.[26] They do not aim to make any additions to knowledge,[27] as philosophical theories would, but to bring about a clarity of vision that one previously lacked by allowing one to see an order to the things that one already knows. Once the clarity of vision that they aim to introduce has been attained, the elucidations have nothing more to offer and can be allowed to fall away. Wittgenstein’s propositions “drop away”[28] once one has worked through them, then, because they are content-free. This does not make them nonsensical, by the same token that one would not want to say that scientific laws are nonsensical (recall that McGinn wants to make saying vs. showing the central distinction of the book rather than (say) sense vs. nonsense), but their disposability lies rather in the fact that they do not add up to a procedure for generating propositions that do have content; i.e. they are not empirical, they are not theoretical; they are not parts of ‘science’, even in a very broad sense of that word.
But is this really good enough? Take another hard look at 6.54:
My propositions serve as elucidations in the following way: he who understands me recognizes them as nonsensical, when he has used them—as steps—to climb up over them. (He must, so to speak, throw the ladder away after he has climbed up it.) He must overcome these propositions, and then he will see the world aright.
Wittgenstein’s language here of “throwing away”, of actively “overcoming” his propositions stands in pretty stark contrast to McGinn’s talk about the elucidatory passages in the Tractatus “falling away” when they have done their elucidatory work. If Wittgenstein’s propositions are comparable to “reminders assembled for a particular purpose” as McGinn understands them, i.e. ‘reminders’ designed to bring an order to things that one already knows, to show-as-opposed-to-say something, then it may be the case that one would be prepared to allow them to fall away once they have shown one whatever they were meant to show one, but why would one want to actively throw them away? To overcome them?[29]
IV. Conclusion: The meaning of 6.54
It is time for us to lay our cards quite openly on the table, to state the point of view for our work as the authors of this essay. McGinn’s account of elucidation may be at least somewhat plausible as an account of what appears to be going on in the main body of the text, on one way of reading it (that wants to retain the saying vs. showing distinction). But ultimately, we would like to argue, to take what appears to be going on in the main body of the text on face value is still quite to fail to live up to the challenge posed by Wittgenstein in 6.54. McGinn fails, that is, to comprehend Wittgenstein’s writing strategies -- his dialectical tactics and techniques -- in the Tractatus, strategies notably different from those of the later work (even if the ultimate end in sight is, we think, identical).
In 6.54, Wittgenstein writes that “My propositions serve as elucidations in this way: he who understands me eventually recognises them as nonsensical…”. The emphasis here seems quite different from McGinn’s: Wittgenstein’s propositions are to be understood as serving as elucidations precisely because they are nonsensical.
As
Wittgenstein puts it in his Letters to Ogden, correcting his English
translator’s misapprehension of the point of his own use of the term
‘elucidation’:[30]
“I didn’t mean to use “elucidate” intransitively:
what I meant to say was: My propositions elucidate – whatever they do elucidate
– in this way: [he who understands me recognizes them as nonsensical…] //
Similarly I might have said “My propositions clarify in this way . . .” meaning
“My propositions clarify whatever they do clarify – say, the propositions of
natural science – in this way: . . .”. Here clarify is not used
intransitively...”
Something is elucidated – but something ordinary. Wittgenstein’s elucidations thus have, one might say, no philosophical content. And they don’t show philosophical things either. One gains lucidity about the character of nonsensicality, by the following means: what are actually made more lucid by elucidations are only humdrum things/words which make sense[31] – if one reads 6.53 (and 4.112) [32] again with this in mind, it is obvious that McGinn’s reading is unfaithful to Wittgenstein, and thus that McGinn’s paper has really nothing to contribute to our understanding of ‘elucidation’ in Wittgenstein. All one has, apart from the elucidation of the ordinary, is the (vitally important) realisation of the nonsensicality of efforts to do more than that. Whereas McGinn makes it sound as though we are still elucidating (imaginary) entirely unhumdrum things – e.g. 'the nature of picturing', and 'language-world relations'.
There is a way in which McGinn -- like us, following Conant and Diamond -- is 'pro' the ordinary, against metaphysics. But what McGinn has in common with ineffabilist readings of the Tractatus is thinking that the elucidations are getting us to see certain 'things'. We think therefore that she has got 6.54 exactly wrong: she is in practice precisely holding onto the elucidatory propositions, not really throwing them away! In fact, she sometimes seems to think it easier to hold onto the elucidations than standard ineffabilism does: she seems in a sense to think that the elucidations aren't really nonsense at all, but rather something like senseless, because quite trivially true. Instead, her interpretation bears certain resemblances to the 'formalist' rendering of the Tractatus suggested by Max Black:[33] McGinn would perhaps therefore respond to our remarks in the paragraph above by saying that we ought to think of the elucidations as themselves -- when seen aright -- being 'ordinary', humdrum, trivial remarks. For example, she would perhaps, we take it, say that to say that 'Language mirrors the world' is such a remark.
But the dream of a contentful philosophical grammar or of the assemblage of the forms of our life in the form of allegedly tautologous truths – a dream that in one way or another McGinn surely shares, even as she castigates the Tractatus for still wanting a “logically perspicuous symbolism” [34](p.513) – was a dream that Wittgenstein had, evidently, abandoned at the time of publication of the Tractatus.[35] Really taking that in means overcoming the urge to say (e.g.), 'Language mirrors the world', to think that one would be saying something by uttering that string, or even to think that it is a harmless (because tautologously true) thing to say.
To understand the work that Wittgenstein’s propositions are supposed to do as elucidations, and (what is closely related) why one should actively throw them away when one has finished with them, seems to require a more detailed (and different) account of the Tractarian conception of nonsense than that given by McGinn. Such an account is given perhaps its fullest expression at 5.4733:
Frege says: every legitimately constructed proposition must have a sense; and I say: every possible proposition is legitimately constructed, and if it has no sense this can only be because we have failed to give a meaning [Bedeutung] to some of its constituent parts.
(Even if we think we have done so.)
Thus “Socrates is identical” says nothing, because we have given no meaning [Bedeutung] to the word ‘identical’ as an adjective. For when it appears as a sign for identity it signifies in an entirely different way—the signifying relation is a different one—therefore the symbols are entirely different in the two cases; the two symbols have a sign in common only by accident.
Here, it seems fair to say, Wittgenstein is employing the string, “Socrates is identical” in the course of an elucidation in exactly the way ‘prescribed’ in 6.54. It looks as though “Socrates is identical” could be construed as a perfectly legitimate sentence with a subject-predicate form, exactly like the proposition “Socrates is dead” in that regard. This is the impression that has to be overcome. For in the case of “Socrates is dead”, we know pretty well what it would be for this proposition to be true or false. Only a proposition has sense, but in order for one to be able to decide what sense it has, one needs to know what referential work each of the signs that make it up is doing. A key clue to this referential work is given by the symbolising role that the sign plays in the proposition. In the case of “Socrates is dead”, ‘Socrates’ is a symbol for a subject and ‘dead’ for a predicate, for example. But in the case of “Socrates is identical”, it is difficult to know what work is being done by ‘identical’ (and equally, therefore, by ‘Socrates’ and ‘is’) because we haven’t given it a ‘Bedeutung’ that could give it a role as a predicate. What Wittgenstein’s consideration of “Socrates is identical” has elucidated for his receptive reader is the functioning of propositions like “Socrates is mortal” or “Socrates is dead”, and, by contrast, what it is for a proposition to be simply nonsensical; and it has done so, obviously, by means of a kind of example. If a proposition is nonsensical, it is because we have not given a meaning (in the sense of reference/Bedeutung) to some of its constituent parts. And of course language can deceive here: just because a sign has an established reference in one symbolising role does not mean that it will continue to have one in another.[36]
McGinn fails to respect the centrality in Frege’s and Wittgenstein’s work of ‘the context principle’, the elementary recommendation – operative, crucially, in the example just considered -- not to seek for sense except in the context of a proposition.[37] She thus tries to hold onto some sense of sense in Wittgenstein’s elucidatory ‘propositions’ – i.e. in ‘what’ they supposedly elucidate – after he himself has urged one above all to abandon such an illusion. Wittgenstein resolutely insists that it is only an illusion of sense that one has when one reads the sequence of Tractarian strings as if they are (sensical) propositions, or even as would-be tautologies. Their resemblance to such propositions, their being constructed out of perfectly respectable words, ought not to blind one to that illusion.
McGinn tells her reader (p.497) that the main task of her paper “is to develop an understanding of Wittgenstein’s idea that his remarks ‘serve as elucidations’ (TLP 6.54).” Now, we have seen that in 6.54 Wittgenstein tells his reader that his propositions serve as elucidations, if one comes to recognise them as nonsensical. If they are successful in that role – a process, an event, that one would do better imagining taking years or decades rather than seconds to occur -- what they will have elucidated for one, as Wittgenstein remarks to Ogden, is “philosophic matters”.[38] That is to say, they will have enabled one to act on the realization that Wittgenstein has failed to give a meaning to certain signs in his ‘philosophic’ propositions – that is, has failed really to say anything at all -- and to realize through and in practice that acted-upon self-consciousness about such matters is the goal of philosophical struggle. McGinn’s paper does not achieve this desideratum, and thus fails in its main aim.
Epilogue: The elucidatory
reading of the Tractatus
We think that McGinn’s reading could survive as an interpretation were it not for the important (impression of a) theme running through the Tractatus that culminates in 6.54: the distinction between sense and nonsense,[39] which is much closer to being a master-theme of the Tractatus than is that between saying and showing.[40] There are parts of her paper that give a plausible rendition of what seems to be going on in parts of the main body of the text. And some of this is very refreshing: it is refreshing, for instance, to see someone trying to actually make some sense out of ‘picturing’ (as she does on p.501f.), rather than subliming it into a positivistic ‘Picture Theory’ or into an ineffabilium). But it is because of the (partial but real) plausibility or attractiveness of her reading, as containing positive views that ‘right minded’ philosophers may actually want to align themselves with (and as more promising and less obviously-wrong than its metaphysical predecessors[41]), that she fails to give a satisfactory interpretation of the whole. Her paper thus serves to perpetuate, albeit in a more developed (and therefore perhaps insidious) ‘anti-theoretical’ form, exactly the kind of impulse (to think that there is some thing for philosophy to be about and to achieve, over and above insight into one’s own desires to do metaphysics and theory) that it was Wittgenstein’s ultimate and consistent aim to help us overcome. McGinn does not help us to elucidate ourselves – and ‘the peculiar practice of elucidating oneself’ might be one happy (elucidatory?) way of putting what it is that the Conant/Diamond/Kremer reading of T L-P is, above all, about.
But, we might be asked, can the latter reading stand, if there is not more to say vis-a-vis the main specific criticism (of the therapeutic reading) that McGinn makes in her paper—how can a work of supposedly ‘simply’, unintelligible nonsense be read as containing the insights necessary for us to realise that it is not communicating anything? “…[I]f the ladder by which we climb from unselfconscious nonsense to self-conscious sense turns out to be an illusion, how can we have got anywhere by climbing it?” (p. 496).
But there is a revealing and (we think) fatal error in the formulation of McGinn’s question, an error which should be relatively unsurprising now that we have clarified McGinn's complete failure to understand that the crucial thing is to overcome the elucidations, the nonsense, of TLP. The progress through TLP of the reader who understands Wittgenstein crucially involves climbing from unselfconscious nonsense to selfconscious NONSENSE, not to selfconscious sense! One throws away the ladder, when one manages to hear the elucidations precisely NOT as sense any more, and thus overcomes and jettisons them. They don’t ‘contain insights’. Rather, one engages in a certain (surprising) practice with them – that’s it.
Wittgenstein’s aim is to bring one from a piece of disguised nonsense, a seeming-proposition, to something that is patently nonsense, and that involves the idea that one can think one can understand something – disguised nonsense – when in fact one does not; when, in fact, the whole ‘proposition’ has as yet no meaning. But in order to realise that one cannot, one needs to ‘know’ inside out the illusion that one can. And that process may – or rather, will -- take some time, and have various ‘intermediate’ – transitional – steps.
To a charitably reformulated version of McGinn's question, now heard as “…If the ladder by which we climb from unselfconscious nonsense to self-conscious nonsense turns out to be an illusion, how can we have got anywhere by climbing it?”, the answer, perhaps, is: we haven't got anywhere: but we may have learnt a lot -- about ourselves in relation to our language -- on the 'journey', on the ‘road’ (to nowhere)…. A bit like the way you can learn a lot by listening to a car engine idling. (But the engine – is you.)
And thus, rather than McGinn’s interpretation, it is actually the alternative (Conant/Diamond) interpretation of the Tractatus, the therapeutic reading of that work (and of Wittgenstein’s philosophy in general), which perhaps deserves the title of the elucidatory interpretation. (For it is that reading which manages to make the role of elucidations central to the functioning of the Tractatus in one's philosophising.) So long as one is clear, with Wittgenstein, on the crucial point that the elucidations themselves are ultimately to be thrown away.[42] That the goal of Wittgenstein’s philosophizing is to get one to overcome the hopeless and needless hope that there was ever really anything (of the kind that we wanted and ‘required’) for our elucidations to elucidate in the first place.
This interpretation of the Tractatus, unlike that of the positivists or the ineffabilists or even the would-be ‘elucidatory interpretation’ given by McGinn, can (we think) finally (and resolutely) comprehend the way the Tractatus ends.[43]
ANSCOMBE, G.E.M; An Introduction to Wittgenstein’s Tractatus; Thoemmes Press, 1996.
CONANT, James; ‘Must we show what we cannot say?’, in The Senses of Stanley Cavell, ed. R.Flemming and M.Payne, Bucknell University Press, 1989.
“On Wittgenstein’s Philosophy of Mathematics”, Proceedings of the Aristotelian Society, LXXI, 1997.
‘Elucidation and nonsense in Frege and early Wittgenstein’, in Crary and Read, below.
CRARY, Alice, & READ, Rupert, eds.; The New Wittgenstein, Routledge 2000.
DIAMOND, Cora; ‘Ethics, imagination and the method of the Tractatus’, in Crary & Read, above.
‘Throwing away the ladder: How to read the Tractatus’, in The Realistic Spirit, MIT Press, 1991.
FREGE, Gottlob; The Foundations of Arithmetic, ‘On Concept and Object’, ‘Logic in mathematics’, all in The Frege Reader, ed. M. Beaney, Blackwell 1997.
HACKER, P.M.S; Insight and Illusion, Thoemes Press, 1997.
ISHIGURO, Hidé, ‘Use and reference of names’, in Studies in the Philosophy of Wittgenstein, ed. P. Winch, Routledge 1969.
McGINN, Marie, ‘Between metaphysics and nonsense: elucidation in Wittgenstein’s Tractatus’, in ‘The Philosophical Quarterly’, 1999.
WITTGENSTEIN, Ludwig, Tractatus Logico-Philosophicus, trans. C.K. Ogden, Routledge 1922; and trans. D.F.Pears & B.F. McGuinness, Routledge 1961.
Philosophical Investigations, trans. G.E.M. Anscombe, Blackwell 1958.
Zettel, trans. G.E.M. Anscombe, Blackwell 1981.[44]
[1] A complication: there exists ‘another’ interpretation of the Tractatus, which once upon a time was very popular, and which aims to be anti-metaphysical: namely, the positivist interpretation, of the Vienna Circle et al. We do not focus on this ‘additional’ interpretation in the body of the text for two reasons: firstly, because of its current unpopularity, at least ostensibly, as a way of taking the Tractatus; secondly, because it is in the very final analysis arguably not substantially different from the “metaphysical” interpretation, as both endorse a “substantial” conception of nonsense (for further explication, see Section 5 of Conant’s (2000)). We cannot explore the positivist reading of the Tractatus further here; suffice to say that we would argue that the metaphysical interpretation focuses on thought and world, and is ‘Realist’; while the positivist interpretation focuses on language and is ‘Anti-Realist’; neither Realism nor Anti-Realism captures Wittgenstein’s thought, we believe, even in T L-P. (See also the notes below on the affinities of Marie McGinn’s (later) Wittgenstein to positivism, and of her (early) Wittgenstein to an apparent discrete variant of ineffabilism, namely 'formalism'.)
[2] It should be pointed out that what we offer here is only a general characterisation of this interpretation. Examples of readings that follow this line in detail can be found in Hacker’s Insight and Illusion and (perhaps best of all) in Anscombe’s Introduction to Wittgenstein’s Tractatus. (See also note 2 of M. McGinn’s (1999).)
[3] This is how Michael Kremer, a critic of the traditional interpretation, prefers to cast this point, on p.4 of the m.s. of his “To what extent is solipsism a truth?” (forthcoming in B. Stocker and G.Guzey (eds.), Post-Analytic-Tractatus (London: Ashgate, 2001)): “[I]n discussing logical form we are attempting to state the essential presuppositions of any meaningful language whatsoever. To treat this as a fact which could be represented in a language is to suppose a language in which we could at least envisage the possibility of this fact’s not obtaining. But such a language, Wittgenstein would argue, would be an “illogical language” – which is to say an impossibility.”
[4] We intend here a stock figure, not necessarily to be identified with, especially, the real historical Frege. We find at least a strand in Frege that is considerably more sophisticated and non-traditional than this stock figure – see for instance Joan Weiner’s work.
[5] It is very clear, for example from Russell’s “Introduction” to the Tractatus, that Russell himself interpreted the Tractatus in this ineffabilistic way. (Like many such readers, he found ineffabilism to say the least an unorthodox vehicle for putting forward more or less traditional philosophical views, and was not convinced by the vehicle: “What causes hesitation is the fact that, after all, Mr Wittgenstein manages to say a good deal about what cannot be said…” (p.xxi (Pears and McGuinness)).
[6] Here, as everywhere in this paper except where marked otherwise, we use the ‘authorised’ (Ogden) translation of T L-P. But this point at least is equally clear in the Pears-McGuiness translation: “[T]he aim of the book is to draw a limit to thought, or rather – not to thought, but to the expression of thoughts, for in order to be able to draw a limit to thought, we should have to find both sides of the limit thinkable (i.e. we should have to be able to think what cannot be thought). // It will therefore only be in language that the limit can be drawn, and what lies on the other side will simply be nonsense.” (p.3) There is, we think, no room, here in the clear and plain speaking of the Preface, for ‘important nonsense’, for ‘gesturings at’ or ‘showings of’ alleged quasi-thinkable but not speakable deep truths. There is only what can be said clearly, on the one hand, and nonsense, on the other.
[7] Perhaps the best account of this telling point about 6.54 is James Conant’s, in “Throwing away the top of the ladder” (The Yale Review 79 (1991), 328-364). It is surprising and perhaps telling that M. McGinn fails even to mention this paper, in her (1999).
[8] Or a ‘God’s eye view’, or a ‘view as if from sideways-on’, or an ‘external point of view’ … with John McDowell and others, part of the burden of our argument here is that there is every reason to believe that Wittgenstein entirely gives up the fantasy of such such a position outside language – indeed, that he gives it up in the Tractatus (cf. T L-P 4.12, the Preface, etc.; cf. also n.3, above)
[9] McGinn, p. 496. Potentially a good question – and perhaps also a hint that McGinn wants to hold onto the illusion that we have actually got someplace when we climb Wittgenstein’s ladder. Contra Hacker, the final lines of his “Was he trying to whistle it?” (in The New Wittgenstein) arguably support rather than undermine the new ‘therapeutic’ reading. The ladder is a metaphor. McGinn is still stuck somewhere on its upper rungs if, like Hacker, she thinks one should have got somewhere through reading T L-P – she is still covertly an ‘ineffabilist’. She still wants elucidations to deliver (unspeakable) insights – see below, for more detail.
[10] McGinn, pp. 496-7.
[11] McGinn, p. 497.
[12] Which is how Hacker sees the therapeutic reading’s Tractatus – see his “Was he trying to whistle it?”, in The New Wittgenstein (op.cit.).
[13] McGinn, p. 504. There is a risk in McGinn’s phrasing here of an illusion of literal perspicuous vision, as if this ‘seeing’ were not just a metaphor, but rather one could literally chart the whole of grammar, as if from outside/above. (This risk, this illusion, we believe, is fully realized in the Analytical Commentaries on the Philosophical Investigations of Peter Hacker (and Gordon Baker, who has recently distanced himself in print from the Hackerian reading – see e.g. n.10 of Baker’s “Wittgenstein: Concepts or conceptions?”, in The Harvard Review of Philosophy IX (Spring 2001), 7-23).) For a reading of PI 133 which resists this danger, see R. Read’s “ “The real philosophical discovery” ” (Philosophical Investigations 1995). It is important to see the ‘seeing’ that is involved here, according to McGinn, as metaphorical through and through, or at most therefore as aspect-seeing... Compare McGinn p.508: she repeatedly writes as though we just manage to see lots of philosophical things/truths. (Cf. also n.14, below.)
[14] It is a dimension explored in some detail by Michael Kremer, who, for example, on p.7 of his (op.cit.), argues that “We should not read talk of “showing”, and correlatively of “perceiving”, “seeing”, “recognizing” that which is shown, on the model of a relation between a subject and some ineffable fact-like entity… This form of the idea of showing is exactly what the Tractatus wants to teach us to abandon. Rather, we should read talk of “showing”, and correlatively “seeing”, on the model of the demonstration of a technique, and the uptake required to understand the demonstration.” When one looks at how McGinn in practice explores these matters, one finds that, by contrast, she makes it look as though something is seen. See e.g. pp.507-8 & p.504 of her text, and her n.24, which reads in part, “[Wittgenstein] gets his reader to see that, at a certain level, there is no gap between language and world or between thought (the mind) and language. Neither the mind nor reality contribute anything to the sense of the sentences of our language from a position outside it… Rather, the world and thought are each of them mirrored in language.” (our italics) This is surely just ineffabilism by the backdoor.
[15] McGinn, p. 498.
[16] McGinn, p. 502.
[17] McGinn, p. 498. The reader may well wonder once again how one can take Wittgenstein to have elucidated something about these ‘things’, and not leave him still clinging to the upper rungs of his ladder.
[18] McGinn, p. 498. Evidently, therefore, McGinn’s Wittgenstein, like (say) Hacker’s, is (at least in the Tractatus) just pretty confused.
[19] McGinn, p. 513. These insights which we arrive at – are they really supposed to be tellable in plain, sensical English? One may well suspect otherwise; again, one may well suspect McGinn of harbouring here (a sophisticated version of) ineffabilism. Or perhaps there is an oscillation in her work between sophisticated positivism (Ed Witherspoon, in his “Conceptions of nonsense in Carnap and Wittgenstein” (in Crary and Read, 2000), reads McGinn’s rendition of On Certainty in her Sense and Certainty as a souped-up Carnap) and sophisticated ineffabilism (McGinn’s latest reading of On Certainty (against the Conantian reading of that work), presented at the Amsterdam ‘Skepticism and Interpretation’ Conference, June 2000 (and forthcoming in the volume Skepticism and Interpretation, edited by Conant and Kern), while emphasizing the importance of grammar as a term of philosophical art, backs away from the (Hackerian, ‘positivistic’) claim that, for Wittgenstein, grammar can really be explicitly described at all).
OR: Perhaps the best way to describe McGinn’s overall vision is in this fashion: She is a very sophisticated positivist, more than anything. Not meaning by that that she is scientistic, but simply that she wants to set out some relatively straightforward -- more or less Carnapian -- truths about linguistic meaning etc., truths which will yield us a set of tools with which to (dis)solve philosophical problems. She wants to preserve these truths which feature in T L-P and carry them forward into Wittgenstein’s later work. Whether in the early work or the later, she is inclined to say that Wittgenstein at his best has no position to give: and she then gives his position.
This is a refusal to allow that Wittgenstein might really be doing philosophy differently. A clear example is her footnote 17, where she writes that “[i]t is an unsatisfactory feature of the therapeutic interpretation that it is led to locate the main sense of TLP in the part of the work of which Wittgenstein is later most critical, referring here to the idea of the “general form of a proposition.” This is doubly unsatisfactory: firstly, McGinn misleadingly supposes that the therapeutic interpretation gives TLP a sense at all; secondly, and more important, it is of course the whole point of the therapeutic interpretation that the “general form of [the] proposition” is the lure, not the doctrine; the temptation, not the considered ‘position’. How can McGinn have missed this?
[20] McGinn, p. 513. As may already be guessed, this picture of the task of philosophy strikes us as too traditional by half.
[21] McGinn, p. 498.
[22] Whereas it is striking that the leading therapeutic readers of the Tractatus – and especially, in this connection, Conant – have an extremely detailed and sophisticated understanding of the role of ‘elucidation’ in the Tractatus. See, for example, his “Elucidation and nonsense in Frege and early Wittgenstein” (Crary and Read, pp.174-217), or similarly the longer manuscript “The method of the Tractatus” (forthcoming – this text, from which the former is excerpted and adapted, is cited by McGinn in note.4 of her (1999)). A close study of McGinn’s paper (see especially note 32) reveals that its rendition of ‘the therapeutic reading’ of Wittgenstein is virtually entirely constructed in relation to two papers of Diamond’s from 1991, not (e.g.) any of Diamond’s papers since, nor (crucially) that of the papers of even the ablest and most prolific of Diamondians – i.e. of Conant. Here, for example, is a passage from the closing pages (pp.195-7) of his “Elucidation…” paper: “…[What] should one take the aim of Tractarian elucidation to be? How [on the new/therapeutic reading of Wittgenstein] are we to make sense of the fact that the Tractatus takes itself to be engaged in an activity which is properly termed one of “elucidation”…? …If the aim of elucidation, according to the ineffability interpretation, is to reveal (through the employment of substantial nonsense) that which cannot be said, then, according to the [new] austere reading, the aim of Tractarian elucidation is to reveal (through the employment of mere nonsense) that what appears to be substantial nonsense is mere nonsense. While the aim of the former sort of elucidation was supposed to be the conferral of insight into inexpressible features of reality, the aim of the latter is not insight into metaphysical features of reality, but rather insight into the sources of metaphysics… // …Thus the elucidatory strategy of the Tractatus depends on the reader’s provisionally taking himself to be participating in the traditional philosophical activity of establishing theses through a procedure of reasoned argument; but it only succeeds if the reader fully comes to understand…that philosophy, as this work seeks to practice it, results not in doctrine but in elucidations, not in Philosophische Satze but in das Klarwenden von Satzen.” (How exactly Conant understands the latter – and this is crucially different from how McGinn understands it – is best explained in n.102 of Conant’s “Elucidation…” paper.) Our worry, expanded upon below, is that McGinn’s ‘elucidations’ still end up looking too much like philosophical explanations/ theorisations or like deliverances of ineffable insight.
If McGinn had taken account of Conant’s work, she would perhaps have been less likely to have felt entitled to call her reading of the Tractatus “the elucidatory reading” (or, indeed, the “clarificatory” reading – see again Conant’s remarks on “der Klarwenden von satzen”).
[23] McGinn, p. 502.
[24] McGinn, pp. 509-10.
[25] This rather seems to conflict with what McGinn writes on p.504, concerning there being for Wittgenstein “no propositions that are true a priori”.
[26] Wittgenstein, Philosophical Investigations 127. (See also Zettel, passim.) Try, then, another label for McGinn’s approach – the ‘Formalist Grammatical’ reading. She, like Max Black, attempts to place the elucidations of T L-P in a logically similar niche to the (sinnlos) propositions of logic; and she suggests that their role is very alike to the role of ‘grammatical remarks’ in Wittgenstein’s later work. See the final section of this paper (and n.35, below) for some more detail on this.
[27] Cf. Philosophical Investigations 89; and the footnotes on Kremer, above.
[28] McGinn, p. 502.
[29] It is striking that the German word here Wittgenstein uses here is “uberwinden” – the very word made famous by Nietzsche, in his struggles, in his attempts to persuade his readers to (self-)overcome. We suspect (though we cannot pursue the matter here) that this echo is not accidental, or at least not without resonance and potential significance – and that it is not compatible with McGinn’s approach.
[30] P.51. This passage, we believe, fairly decisively undermines the McGinn interpretation, just as thoroughly as it undermines the standard metaphysical interpretation. It gives the lie to McGinn’s effectively ineffabilist thought that Wittgenstein is still getting us to recognise something about ‘the harmony of language and reality’ by means of his elucidations. (If carpers still carp, asking whether we can rely on a text which is not within the body of TLP, then one need look no further than the oft-ignored 5.5563. 5.5563 gives the lie to something that McGinn appears to believe: namely, that Wittgenstein in TLP regrettably believes that ordinary language cannot be described itself as in logically perfect order. For while Wittgenstein in T L-P did (regrettably) think that ordinary language was invariably not ‘logically perspicuous’ – thus there is scope (see 3.325, 4.002, etc.) for its being made clearer what thought a thought is -- we are not thereby making the thought sharper/clearer. McGinn appears to us to elide – not, as needed to distinguish – these two points.
We are suggesting that if the Tractatus is read properly, it is not necessarily thoroughgoingly committed to a fatal mythology of symbolism (though see n.39, above) – unlike McGinn, who is committed to a residual linguistic ineffabilism. (See also p.50 of the Letters to Ogden,, alongside p.505 of McGinn).)
[31] Compare Wittgenstein’s Preface (see n.6, above). Unlike McGinn’s, our account is (we believe) faithful to Wittgenstein’s injunction to draw ‘limits’ only in language, and to be clear that what is on the other side of those limits is (only) nonsense. That is not to say, of course, that the Tractatus does not have any covert metaphysics. In believing that it does, we do not differ from McGinn or indeed from ineffabilism in general – but the therapeutic reading of T L-P argues that you will find this metaphysics in the wrong place if you do not take seriously Wittgenstein’s aim in T L-P of overcoming all metaphysics. Note, for example, that ‘Ineffabilism’ and ‘the picture theory’ are not attacked by the later Wittgenstein. But some things are: Such as the Russell/Frege subliming of logic and the idea of ‘logical space’, neither of which Wittgenstein entirely overcame even by prop.7 of T L-P.
[32] We note further that the term “elucidation” first enters T L-P at 3.263. McGinn, surprisingly, says nothing about this occurrence of the term. We hope in future work to argue that McGinn’s reading is weakened by having nothing to say about 3.263, while we believe the therapeutic reading of T L-P can make good sense of it.
[33] See n.19 of Conant’s “The Method the Tractatus” (forthcoming) for devastating criticism of the ‘formalist’ interpretation of TL-P; see also n.35, below.
[34] PI (and OC, etc.) features grammatical remarks: these may at various moments in one's philosophical struggle seem to one to be senseless, nonsensical, stipulational, sensical (seemingly true-or-false), and perhaps more besides. TLP engages (one) in its analogue to grammatical investigation -- namely, the activity of elucidation. This much of what McGinn evidently supposes, it would be hard to dissent from. But her quasi-formalist rendering of elucidations makes it look like they really should be part of the symbolism (like tautologies) -- and this just IS a mythology of symbolism, we have suggested. This CANNOT be an adequate reading of 6.54, which requires that we throw away the nonsensical, not that we (e.g.) let the merely senseless ‘fall away’ when we recognise it as senseless -- and the production of an adequate reading of 6.54 was expressly McGinn's MAIN AIM in her paper (see again p.497).
(It may seem an unsatisfactory aspect of our interpretation of McGinn that we seemingly cannot decide whether she is an ineffabilist, a positivist, or a formalist reader of TLP. But we think this is actually a symptom of her confusion -- SHE cannot decide, and oscillates unstably between all of these and the therapeutic reading. Her effort to generate an elucidatory reading which is not simply an unstable pot-pourri of all FOUR of these reading-genres is, we have argued, unsuccessful. (It is quite odd, however, that she does not mention the Blackian 'formalist' possibility, which on the central question of the status of elucidations seems possibly the reading closest to her own (insofar, at least, as we are capable of judging that there is anything stable in her interpretive activity). Perhaps the reason she does not favour ‘formalism’ (besides its well-known catastrophic defects) is her insistence, remarked earlier, that “the idea that all logical truths are tautologies” is itself to be excluded from the alleged “elucidatory core” of T L-P.))
[35] What does harm is to think that one has grasped and can lay out the grammar of our language, so as to refute metaphysicians, solipsists, sceptics etc. . This, as we have said, unfortunately seems to remain McGinn’s wish/dream.
[36] A fuller treatment of the matters discussed in this paragraph would bring out the importance to 5.4733 also of Wittgenstein’s remarks concerning ‘internal relations’ and concerning ‘the context of significant use’ (as emphasized by Conant in his (2000)). But we think that our text above is sufficient for present purposes.
[37] It is important to bear in mind that the CP is above all a methodological recommendation, not a quasi-metaphysical assertion. (In any case, even if taken as an assertion, it will ‘shape-shift’ in the manner depicted by Conant in n.102 of his (2000)) The failure to see its deep relevance throughout Wittgenstein's work links McGinn and Hacker once more – see for example the argument of Witherspoon’s paper, and especially pp.333-8, wherein Witherspoon shows exactly where in her (pre-1999) work she goes wrong, and falls into a sophisticated version of positivism. McGinn’s failure to take proper account of the Context Principle in her thought on any stage of Wittgenstein’s career, together with her wish to have Wittgenstein enable us to say or at least ‘see’ things about language, eclipses the – very secondary -- question of which particular variant of positivism or ineffabilism or formalism (or what hybrid of the three) she ascribes to T L-P, and which to PI or OC. (For more detail on (the centrality of) the context principle, see the work of Conant, Palmer, pp.106-7 of Ricketts’s paper in Sluga and Stern, and the 1997 exchange between Goldfarb and Diamond.)
[38] Letters to Ogden (op.cit.), p.51 – and see also the passage from p.51 quoted above. McGinn has, one might say, been deceived by the grammar of the word ‘elucidation’…
[39] It is important to understand that we do NOT intend this as a general, quasi-technical, quasi-positivist distinction which can be wheeled in to do philosophical work. On this score, the discussion on pp.70-71of Goldfarb’s (1997) is well-taken: “Wittgenstein’s talk of nonsense is just shorthand for a process of coming to see how [certain strings of] words fall apart when worked out from the inside.” (p.71, our italics)
[40] One reason why this must be so being that, crucially, the say vs show distinction as employed by the ineffabilists -- and once again we fear that on this point McGinn must backslide into their position (and/or else into some variant of positivism) -- is a mashed-together and incoherent rendition of two different distinctions. (I) There is what is said by ordinary propositions (something) and tautologies (nothing) versus what their form shows us (that they are propositions, or that they are tautologies) – a distinction made by Wittgenstein in body of the Tractatus. (II) Then there is what is said by nonsensical propositions (nothing) versus what is allegedly shown by them (e.g. the way language hooks up to reality), or what is allegedly elucidated for us by means of them – II is a ‘distinction’ not found in the Tractatus, but it has to bear the main weight of any more or less traditional reading of that work. For detail on why the say vs. show distinction is thus a hopeless basis for any genuinely elucidatory reading of that work, see notes 11, 19, 20, 26 & 68 of Conant’s (2000).
[41] By which we do not mean to say that McGinn actually succeeds in 'making sense of' more, quantitively, of the text than (say) Anscombian ineffabilism does -- we suspect that the reverse of that is true. What we mean is simply that her reading seems at least to give some hope of not turning out to be evidently irresolute, and of not requiring Wittgenstein to be a philosophical 'chicken', a fundamentally inconsistent and unintegritous figure. Plus that, as discussed above, she has at least the ambition, as we do, of sketching real and deep continuities between Wittgenstein’s early and later work.
[42] See again Conant (2000), n.102.
[43] Many readers, we suspect,
will however be more interested in the lessons, if any, which can be garnered
from this for the purposes of reading later Wittgenstein. Let us then briefly
note that, in the context of the Investigations,
Wittgenstein’s ‘elucidatory’ aim does not seem so paradoxical, and yet it too involves the idea of ‘disguised’
nonsense, the possibility of our thinking that we can understand something that
in fact we can’t. And it too,
arguably, involves itself being
(recognised to be) just such disguised nonsense. All Wittgenstein’s great
works, we believe, need to be taken in a thoroughly dialogical and transitional
fashion. As David Stern has put it, “On [the therapeutic, Cavell/Diamond]
reading, the aim of Wittgenstein's dialogues is not lead his reader to any
philosophical view, neither an idealised, frictionless theory of language, nor
a pragmatic theory of ordinary language, but rather to help us through such
ways of speaking and looking.” (“The availability of Wittgenstein’s philosophy”,
in Sluga and Stern (1996), p.444; though Stern means only to be talking about
later Wittgenstein.) This brings out again the importance of our topic: this is
no narrow debate over the interpretation of one text from 80 years ago; it
ramifies into the question of whether recent and contemporary philosophy have
tended to misplace Wittgenstein’s entire
contribution to philosophy. We believe that Wittgenstein’s elucidations etc.
throughout his career have to be recognised as of transitional, purpose-relative and ‘personal’ use only. This is why
we suggest that McGinn perpetuates in an insidious – because sophisticated, and
avowedly anti-theoretical -- form, exactly the kind of impulses that it was
Wittgenstein’s ultimate aim throughout his career to help us overcome. McGinn
makes it look as though Wittgenstein’s later work thoroughgoingly lets him do
(vis-à-vis language and grammar) what was illegitimate according to the Tractatus: namely, to simply say and
understand philosophical truths about how language works etc. . That is, people
think that one can simply say stuff about language, about ethics, etc., after
and according to the later Wittgenstein. We have suggested instead:
>>that Wittgenstein’s work never barred
one from doing anything, or at least from doing anything that actually is
anything (one might say);
>>that it is McGinn (and Hacker, etc.) who are at times quietists in a
problematic fashion (through telling us we mustn’t theorize without necessarily
enabling us to give up the wish to theorize, thus leading to a vacillation
between quietism and positivism, a vacillation hardly surprising against the
background of this essay and of Conant’s work – see especially his defence of
the therapeutic Wittgenstein as non-quietist in Section 5 of his (1997));
>> and, crucially, that, in his
later work quite as much as in his earlier work, Wittgenstein was endeavouring
to get one to give up the need, the wish,
the impulse, to say the kinds of things about language, grammar etc. that his
‘followers’ almost invariably say.
Contra McGinn’s intimations, ‘the
autonomy of grammar’, for example, is neither a controversial and true
philosophical thesis (as Hackerian quasi-positivism would have explicitly it)
nor an ineffable philosophical thesis (as McGinn would, we think, if against
her professed inclinations, have it). ‘It’ is no thesis at all. (Wittgenstein
NEVER says in so many words the more or less systematic things which Hacker and
McGinn attribute to him about grammar.) The remarks of Wittgenstein’s which
some have taken to reveal ‘the autonomy of grammar’, and thus for example
to show, via “the concept of grammar,
the potential for language to reveal ‘what kind of object anything is’”
(McGinn, p.513), are in fact only purpose-relative attempts to deflate certain
Metaphysical Realist philosophical illusions. They have no ‘affirmative’
content whatsoever
Crucially,
McGinn’s Wittgenstein gives one no real reason
not to theorize. It is not surprising therefore that she preserves too
straightforward – too traditional – a role for philosophy. Her attempted
reading of the Tractatus remains, one
might say, the shadow cast by metaphysics. An elucidatory reading of the Tractatus,
one that understands its full self-immolation -- in one’s
understanding-in-action, over time, of Wittgenstein’s purpose in writing it --
is needed, by contrast, if one is to have a chance of giving philosophy peace,
of attaining (philosophical) peace. Such an ‘elucidatory reading’ has we think
already been given in outline – it is
the ‘therapeutic’ reading.
[44] Our deep thanks go to Jim Conant and Simon Glendinning; thanks also to Kelly Dean Jolley, Dan Hutto, David Oderberg, Rob Deans, and to audiences at the U. Chicago Wittgenstein Workshop and the Notre Dame Philosophy Department.