Professional Documents
Culture Documents
Commentary On José Zalabardo's The Tractatus On Unity'
Commentary On José Zalabardo's The Tractatus On Unity'
ABSTRACT
José Zalabardo’s view of the aims of the Tractatus limits the options available to us for
reading and understanding the book. I argue that an alternative kind of reading is
possible, if we don’t take for granted Zalabardo’s view of Wittgenstein’s aims. A
necessary condition for any alternative reading is that it can make sense of central
features of the Tractatus. I start from the development of Wittgenstein’s
understanding of philosophy, and show that such an alternative reading is possible.
There are at least three advantages which a reading of this sort, an undoing or inter-
rogatory or transformative reading, has.1 One advantage I have already suggested. There
are passages in the Tractatus which appear to be meant to help us to engage in philo-
sophical activity, including the activity of diagnosing philosophical confusion. One of
the most striking is the one I quoted earlier, where Wittgenstein [TLP 3.325] says
that in order to avoid the fundamental confusions of which philosophy is full, we
must use a symbolic notation, which avoids the kind of sign-symbol ambiguities of
ordinary language, which underlie those confusions. If the author of the Tractatus
thought that philosophical proposition-making could be shown to be incoherent in
one fell swoop, there wouldn’t seem to be anything for a symbolic notation to do.
And there are other passages in the Tractatus which also seem to suggest that Wittgen-
stein took there to be an activity of philosophical clarification which involved particular
sorts of tools, and which could not simply be identified with some fell-swoop argument
put forward in his book. I don’t want to suggest that a fell-swoop reading cannot
account for Wittgenstein’s remark about the need for a good symbolic notation, or
for his other remarks that appear to be guides to philosophical activity. My point is
simply that undoing readings have an apparent advantage here.
A second advantage of a reading of the Tractatus which takes the book to provide
ways in which we can engage in the activity of philosophy, rather than a general fell-
swoop demonstration of the nonsensicality or incoherence of philosophical thinking,
comes into view once we notice that a disadvantage of fell-swoop readings is that
they involve the drawing of conclusions from the supposed doctrines of the Tractatus
(in Zalabardo’s case, from the account of propositional representation), and so there are
questions how anything can be justified by doctrines which are themselves incoherent
or nonsensical. I am not suggesting that fell-swoop readers cannot come up with replies
to this sort of question, but only that it is an advantage of undoing readings that they
don’t need to do any such thing. An undoing reading involves the use of methods of
philosophical clarification to which one may be led by reading the Tractatus, but a
method of clarification is justified by its helpfulness, not by what might have led one
to it in the first place. If one comes to see propositions of the Tractatus as nonsensical,
this doesn’t call into question the helpfulness of whatever philosophical methods one
might have been led to by reading the book.
That second advantage of undoing readings is connected with a third advantage,
which can be seen in the case of Anscombe’s [1963: 151] example. People fail to
mean anything by what they say in various sorts of context in ordinary life, and this
is frequently the result of being taken in by a misleading analogy of some sort. It is
sometimes the result of being taken in by features of the grammar of a language—
and often enough, these are features that can be brought into view by comparison
with some other language, with different grammar. We may sometimes wind up
saying nothing coherent because (for example) we have got muddled about the differ-
ence between relative and interrogative uses of ‘what’, or between exclusive and inclus-
ive uses of ‘or’. It is an advantage (I think) of undoing readings of the Tractatus that, on
such readings, the diagnosis of philosophical confusion, of nonsensicality or failure to
mean anything definite, is similar to the ways in which we can sharpen our eyes to con-
fusion in perfectly ordinary sorts of cases (see Conant [2014: 46–7]).
1
For more on the significance of such readings, see Floyd [2007].
AUSTRALASIAN PHILOSOPHICAL REVIEW 277
None of these advantages would count for anything, though, if only a fell-swoop
reading could make sense of central features of the Tractatus, including the significance
it gives to the general form of proposition. One might, that is, ask a proponent of an
undoing reading, why Wittgenstein should have bothered to specify what all prop-
ositions have in common, if one can get rid of philosophical confusion by simply
improving and making deeper the techniques we have for uncovering illusions of
sense, with the addition of techniques derived from the work of Frege and Russell.
There are various ways in which one might respond to such an objection. Thus, for
example, Oskari Kuusela [2011] reads the Tractatus as providing a kind of symbolic
notation through its presentation of propositionhood; and then that notation is
intended to be an essential tool for the achieving of philosophical clarity and the reveal-
ing of logical unclarity. I have a different suggestion, which I set out in Parts 2 and 3.
was closer to Hertz’s thinking about our wanting to ask ‘What is the nature of force?’
What has happened in that case, as Hertz [1899: 7–8] saw it, is that our conceptual
unclarity about physics gets expressed in an unclear question about the nature of
force. If we resolve the conceptual unclarity, the question would disappear. I’m not
suggesting that we simply take Wittgenstein to be expressing Hertz’s view, but that
we take seriously the idea that Wittgenstein is not trying to show that the nature
of propositional representation rules out the asking or answering of philosophical
questions; it’s rather that he thinks that the propositional shape that our philosophi-
cal questions have, the shape that we take as entirely obvious in our usual under-
standing of philosophy (as well as the particular propositional shape that the
puzzlement takes in particular cases), is already a reflection of misunderstanding
of language. Putting this point another way: when Wittgenstein says that the way
our problems get put rests on misunderstanding, he suggests that, in putting our pro-
blems as we do, in giving them their shape, we mis-shape them into being or appear-
ing to be a particular kind of problem. This is a point that has been emphasized by
Juliet Floyd [2007: 186–90]: the way we put our philosophical questions, the Frage-
stellung, may itself reflect a ‘need for clarification’. She takes this to be essential also
to an understanding of how Wittgenstein’s early philosophy is connected with the
Philosophical Investigations.
It may seem that, in speaking of the way we may ask ‘What is a proposition?’ despite
its being in a sense before our eyes, I have been making the Wittgenstein of the Trac-
tatus too close to the later Wittgenstein. But Wittgenstein [1979: 182–3] himself pointed
out to Waismann in 1930 that already in a manuscript of the Tractatus, he had written
that the answers to philosophical questions must never be surprising; and that in phil-
osophy you cannot discover anything. When he commented on this in 1930, he took it
to reveal a problem in the Tractatus, since he had also held, back then, that at some later
time the application of logic would reveal the elementary propositions. That was his
example of how he had gone against his claim that in philosophy you cannot discover
anything. In the background to that choice of example is his idea that the other things in
the Tractatus weren’t ‘discoveries’. So, for example, he did not take it to be any kind of
discovery that there were elementary propositions. This is in some sense already given
with any proposition: the entirety of logical space is already given by every proposition.
That’s not ‘hypothetical’, so it doesn’t count as a discovery. By excluding from what
counts as discovery everything which is not ‘hypothetical’, he could think of his own
innovative logical moves as not involving discoveries, but rather as bringing out what
is already present in our thought or in our language. The idea of philosophy as clarifica-
tion is important in understanding what he took himself to be doing in the Tractatus,
but it’s not as if clarification is one thing: one thing that he is himself doing, and that he
is giving as what philosophical activity consists in. Further, while it is clear that there is a
contrast between philosophy-as-clarification and philosophy conceived as progressing
through improved solutions to problems like those on which Zalabardo focuses, it is
not obvious how that contrast should be drawn. The problem-and-improved-solution
understanding of philosophy is entirely compatible with the idea that we might need to
clarify the formulation of our questions. In ‘On Scientific Method in Philosophy’,
Russell [1914 (1932)] gives a problem-and-improved-solution account of how philos-
ophy should progress, which many later analytic philosophers would recognize as
expressing a version of their understanding of the subject, and which indeed helped
to shape that understanding (and which may have helped to shape, as a kind of reaction,
AUSTRALASIAN PHILOSOPHICAL REVIEW 279
the very different understanding that is expressed in the Tractatus).3 But Russell’s
account clearly recognizes the importance of clearing up the formulation of philosophi-
cal problems, which could reflect confusion that might be revealed by logical analysis.
(And Russell had also emphasized the role of clarification in philosophy. See, for
example, his essay on the nature of truth and falsehood [Russell 1910 (1966)].)
The familiar problem-and-improved-solution conception, as we see it in Russell’s
[1914 (1936)] ‘Scientific Method in Philosophy’, and in later analytic philosophers
including notably Carnap, is modelled on an understanding of how science progresses.
Both problem-formulations and possible solutions are meant to have the kind of imper-
sonality of scientific problems and solutions. The Tractatus starts off from an apparent
foray into ontology, and might seem to promise the kind of impersonality of philoso-
phical method as described by Russell, but the structure of the book reflects a quite
different sort of impersonality. In the opening statements about what the world is, Witt-
genstein is (I believe) responding to Russell, who had said that what the existing world
consists of is many things with many qualities and relations, and that a complete
description of it would have to include a catalogue of the things, as well as mentioning
their qualities and relations.4 That way of thinking seems to call for such propositions as
‘(∃ x).x = a”, as part of our catalogue of things; it also (as Wittgenstein thought) was
connected with unclarity about logic and the supposed need for logical laws. But Witt-
genstein’s ‘The world is the totality of facts, not of things’ [TLP 1.1] is not an alternative
solution to the philosophical problem, impersonally formulated, ‘What does the world
consist of: facts, or things, or things together with qualities and relations, or what?’ Nor
is it a response to a confusion that is, as it were, personal to Russell. I quoted Juliet
Floyd’s phrase ‘a need for clarification’; and the point here is that Russell’s remark
about what the world consists of can be understood (and was understood by Wittgen-
stein) to show that he had gone down one of the paths that lead us deeper and
deeper into ‘an immense network of easily trodden false paths’—with no easy and
obvious way back out.5 If there is a way back out, it’s the whole of the Tractatus that
is supposed to help us to find it—to interrogate, and rethink, the apparent ineluctability
of the questions and the apparent necessities of the answers. I will turn to the respon-
siveness of the book in Part 3, but first I want to note the limits of a fell-swoop reading in
relation to the opening remarks of the Tractatus. Such a reading can take those remarks
to have a significant argumentative role in the development of the Tractarian account of
propositional representation; it can also maintain that the Tractarian passages that
appear to provide an answer to the question what the world is are shown to be incoher-
ent by the Tractarian account of propositional representation, and that the question
itself is similarly shown to be incoherent. The fell-swoop reading also allows for
someone to reject the supposed Tractarian account of propositional representation,
and then to investigate whether the apparently ontological remarks about what the
world is can advance our understanding of metaphysical questions about what there
is. Such an investigation could certainly take up some of the problems and issues to
3
On whether Wittgenstein read ‘On Scientific Method in Philosophy’ during World War I, see Diamond [2014a:
148–9].
4
See Russell [1929: 54]. The first edition of the book, with Russell’s Lowell Lectures, was published in 1914, and
mentioned to Wittgenstein by Keynes in a letter early in 1915. Wittgenstein asked Keynes to send him a copy. I
believe Wittgenstein received a copy in March or April of 1915.
5
The image of the immense network of easily trodden false paths comes from Wittgenstein’s writings in the early
1930s, but fits both his earlier and his later ideas about philosophy. See Wittgenstein [2005: 312].
280 CORA DIAMOND
which Wittgenstein draws attention elsewhere in the Tractatus, and could reformulate
them in the terms of contemporary analytic philosophy. What a fell-swoop reading,
whether it takes the Tractatus to be successful or to be a failure, does not do is what
a ‘need for clarification’ reading, of the sort I am suggesting, does do: take seriously
the idea of the book as reshaping how we see things, in the light of what it takes us
not to have seen, what it takes to be ‘clouded over’. On the particular version of this
reading that I am recommending, one main thing that the book takes to be clouded
over, and especially clearly to be clouded over in Russell’s thinking and that of Frege,
is what it is for logic to be present in our thought. (My focus in what follows on
Russell is not meant to suggest that the cloudings-over in Frege’s thought are not
equally Wittgenstein’s concern.)
The contrast Wittgenstein is making in the Notes between propositions and names
(where only propositions have sense in his sense) is tied to the grammatical point
about what kind of sign signs expressive of sense are. What the ‘One must not say’
formula brings out is that, in a proposition, signs go proxy for things, but the logic
of the facts (here two-term relationality) is exemplified.7 I have put these two points
in something like the form in which they are put at TLP 4.0312; but they needn’t be
put that way. In Wittgenstein’s spelling out of how we should speak, you can see
signs going proxy for things (the ‘a’ going proxy for a, the ‘b’ for b); you can see the
shared logic in the two ‘that’-formulae, you can see the shared logic not being
proxied. The ‘One must not say’ formula was an early expression of what Wittgenstein
later called his Grundgedanke [TLP 4.0312]. The reference to ‘logical constants’ in the
Grundgedanke was meant to cover all propositional forms, including (as in the particu-
lar case in Wittgenstein’s example) dual relationality. The interest of the example as a
case of unclouding is also that it provides an example of what was previously clouded
over. You can’t clearly see what it is for logic to be present in propositions when you say
‘The complex sign “aRb” says that a stands in the relation R to b’.8 I am not suggesting
that suddenly in 1913, Wittgenstein got propositionhood clear. He had an insight that
6
Wittgenstein [1961a: 105]. The first italics are mine; they correspond to the use of italics in the corresponding
passage in the Tractatus.
7
On exemplification and the connection with what shows itself, see Narboux [2014].
8
I am grateful to one of the readers of my commentary for pointing out that TLP 3.1432 is also cited by Zalabardo. I
am not concerned here to argue for my reading as preferable to Zalabardo’s; my aim was to show how the
AUSTRALASIAN PHILOSOPHICAL REVIEW 281
could be, and was, probed further and thought through in the following years. But there
was also the way the ‘One must not say’ remark had responded to a particular ‘need for
clarification’—a different kind of insight. In her book on the Tractatus, Anscombe
[1963] speaks about how Wittgenstein’s conception of the picture-proposition makes
its logical character ‘extremely intelligible’, but she can do this precisely because she
had previously brought out how negation can appear mysterious. The responsiveness
of philosophy as Wittgenstein conceived it is mirrored in the structure of her expo-
sition, and we should see an early occasion of what such responsiveness can be in
the 1913 Notes.
Wittgenstein’s thinking often moved from a particularly clear simple case, in which
what is logically significant is open to view, to a generalization of that case, which can
then be taken to reveal something present in all thinking. He also sought rewritings and
redescriptions, in which thought is made more perspicuous. Both of these ways of phi-
losophizing are important in the 5’s of the Tractatus, and important also for the con-
nection with the general form of proposition. At 5.101, he gives a way of setting out
the truth-functions of an arbitrary number of elementary propositions, illustrated by
the case of two propositions; and he uses a kind of propositional sign, a sequence of
T’s and F’s, introduced earlier at 4.442. In such a setting out, what is brought out
clearly is what the common truth-grounds (if any) are of the various resulting prop-
ositions. The subsequent remarks in the 5.1’s make use of what can thus be seen in
some cases, namely that all the truth-grounds that are common to a number of prop-
ositions may be truth-grounds of some other proposition. As we read what Wittgenstein
says there about truth-functional inference, we have before us (as we look back at the
schematism at 5.101) the simple case of truth-functional inferability, from ‘p.q’ to ‘p’:
we can see the containment of truth-grounds. Thomas Ricketts [1966: 84] has remarked
about this, that ‘Wittgenstein audaciously maintains that this understanding of logical
connectedness is exhaustive’: he means the understanding dependent on containment
and non-containment of truth-grounds as between propositions constructed truth-
functionally from elementary propositions. As Ricketts notes, this audacious move
depends on Wittgenstein’s being able to treat quantificational generality in truth-func-
tional terms, and on his treatment of identity, that is, on the 5.5’s [Ricketts 1996].9
Among the central moves in the 5.5’s are the introduction at 5.5 of the operation
that forms the negation of all the propositions that are values of some variable; and
the introduction at 5.501 of a kind of description of the propositions ‘represented’ by
a variable. Among other forms of description, there is the specifying of a formal law
governing the construction of propositions. At 5.503, Wittgenstein says that we can
express how propositions may be constructed with the operation introduced at 5.5,
and how they may not be constructed. Here he has in view a formal law governing
the construction of propositions, of the sort mentioned at 5.501. If we are given a
description of the general form according to which propositions are constructed, using
the operation introduced at 5.5, we can see not only what may be in the series of prop-
ositions constructed in that way, but also what isn’t in it. This may change our under-
standing of what it is we want, if we come to see that the series gives us what we had
taken ourselves to want, but had thought of differently. That’s to say, in seeing what
passage could illustrate the overall story I tell about the task of philosophy being to ‘uncloud’ what, prior to phil-
osophy, is clouded over.
9
I have drawn here also on my account of the Tractatus on inference, in Diamond [2014b].
282 CORA DIAMOND
is in the series, we come to see differently what we had thought we needed, like the kinds
of apparent propositions that would figure in a Russellian catalogue of what there is in
the world. Such a change in our understanding undoes the apparent need for logical
tools through which we could get beyond the generality internal to the ordinary prop-
ositions of our language.10 It makes it possible to engage critically with the propositions
we had thought were necessary, makes it possible to see that they involve words to
which no meaning has been given. That is, at any rate, what it is meant to do. The
kind of transformation of Russell’s understanding that is at stake here is more
evident in the Prototractatus, where the remarks about the limits of my world are num-
bered so that they are comments on the use of the identity sign. I am not here trying to
defend this way of reading the Tractatus, but simply to argue that a ‘transformational’
reading of the book can take seriously the role in it of the description of the general form
of proposition.
Michael Kremer has argued that we can see in the Tractatus how Wittgenstein’s
thought was continuing to develop over the years during which he was working on
it. The book is substantially different in significant respects from his earlier writings,
including also his notebooks and even also the Prototractatus. There is a contrast
between Wittgenstein’s ideas about philosophical method (even before they become
explicit in the Prototractatus and the Tractatus) and Russell’s [1929: 63] understanding
of how the role of logic in philosophy will enable us to obtain results which ‘must
command the assent of all who are competent to form an opinion’. How to do philos-
ophy as Wittgenstein wanted to do it is something that took him a lifetime to think
through; and we see some of this development during the war years, and in the
writing of the Tractatus. The new emphasis on the context principle in the Tractatus
goes with a deepened understanding of how to engage in the activity of philosophy. I
think the power and interest of the book lie in considerable measure in how it relates
to its readers—in how it is meant to transform their understanding. But what I am
arguing against here is Zalabardo’s presentation of the options we have in reading
the Tractatus. He says that you don’t have to accept that Wittgenstein succeeded in
undermining all philosophical problem-posing in one fell swoop, and can instead
take the book to have failed in that fell-swoop aim. You can then, he says, take the
posing of philosophical problems to involve legitimate forms of propositional represen-
tation, and you will then regard the Tractatus as providing discussable solutions to a
range of such problems. This understanding of the available options reflects, I think,
Zalabardo’s [2018] idea that philosophy, if it is indeed possible, involves propositional
representation in the posing of its problems; and it hence appears that there is a weighty
question whether what we need for the posing of our problems is genuinely available.
But that question is not ineluctable; it does not have to shape our understanding of
Wittgenstein’s aims; and the possible ways of reading the Tractatus do not have to
be specified in terms of success or failure in the supposed aim of showing that the
correct account of propositional representation rules out the posing of philosophical
problems, in one fell swoop. Zalabardo suggests that there are ‘rules’ that define the
enterprise of philosophy, and one might take such ‘rules’ to include the idea that the
enterprise of philosophy involves posing problems for which propositional represen-
tation is required. That view of philosophy might appear to provide a quick way of
10
Again, there are connections with Russell [1929] (which I take Wittgenstein to have been reading in 1915). See
especially Russell’s [ibid.: 59] discussion of generality.
AUSTRALASIAN PHILOSOPHICAL REVIEW 283
eliminating some sorts of reading from the context of discussion of the Tractatus; but
only if Wittgenstein did think of the philosophical enterprise that way. Whatever one
may make of the significance of Hertz’s influence on Wittgenstein, he certainly provides
a way of understanding what it might be to engage in philosophical thinking, not gov-
erned by supposed rules about the need for propositional representation in the posing
of problems. Zalabardo takes Wittgenstein to be working with an at least implicit rec-
ognition of the ‘rules’ of the enterprise of philosophy. In thinking about what the space
is of possible ways to read the Tractatus, one should recognize that ideas about the sup-
posed ‘rules’ and Wittgenstein’s supposed (at least implicit) recognition of them, should
not be taken to exclude some sorts of readings, since they are themselves part of read-
ings like Zalabardo’s, and not an independent constraint.11
Disclosure Statement
No potential conflict of interest was reported by the author(s).
References
Anscombe, G.E.M. 1963. An Introduction to Wittgenstein’s Tractatus, 2nd edn, London: Hutchinson
University Library.
Conant, James 2014. Mild Mono-Wittgensteinianism, in Wittgenstein and the Moral Life, ed. Alice
Crary, Cambridge MA: MIT Press: 31–142.
Diamond, Cora 2014a. The Hardness of the Soft: Wittgenstein’s Early Thought About Skepticism, in
Varieties of Skepticism: Essays after Kant, Wittgenstein, and Cavell, eds James Conant and Andrea
Kern, Berlin: de Gruyter: 145–81.
Diamond, Cora 2014b. Addressing Russell Resolutely?, Philosophical Topics 42/2: 13–43.
Floyd, Juliet 2007. Wittgenstein and the Inexpressible, in Wittgenstein and the Moral Life, ed. Alice
Crary, Cambridge, MA: MIT Press: 177–234.
Goldfarb, Warren 2011. Das Überwinden: Anti-metaphysical Readings of the Tractatus, in Beyond
the Tractatus Wars, eds Rupert Read and Matthew A. Laverty, New York and Abingdon:
Routledge: 6–21.
Hertz, Heinrich 1899. The Principles of Mechanics Presented in a New Form, trans. D.E. Jones and J.T.
Walley, London: Macmillan and Co.
Kremer, Michael 1997. Contextualism and Holism in the Early Wittgenstein: From Prototractatus to
Tractatus, Philosophical Topics 25/2: 87–120.
Kuusela, Oskari 2011. The Dialectic of Interpretations: Reading Wittgenstein’s Tractatus, in Beyond the
Tractatus Wars: The New Wittgenstein Debate, eds Rupert Read and Matthew A. Laverty, New York
and Abingdon: Routledge: 121–48.
Narboux, Jean-Philippe 2014. Showing, the Medium Voice, and the Unity of the Tractatus,
Philosophical Topics 42/2: 201–62.
Ricketts, Thomas 1996. Pictures, Logic and the Limits of Sense in Wittgenstein’s Tractatus, in The
Cambridge Companion to Wittgenstein, eds Hans Sluga and David G. Stern, Cambridge UK:
Cambridge University Press: 59–99.
Russell, Bertrand 1929. Our Knowledge of the External World, 2nd ed., New York: W.W. Norton and
Company.
Russell, Bertrand 1914 (1932). On Scientific Method in Philosophy, in Mysticism and Logic and Other
Essays, London: George Allen and Unwin.
Russell, Bertrand 1910 (1966). On the Nature of Truth and Falsehood, in Philosophical Essays,
New York: Simon and Schuster: 147–59.
Sullivan, Peter 2011. Synthesizing Without Concepts, in Beyond the Tractatus Wars: The New Wittgenstein
Debate, eds Rupert Read and Matthew A. Lavery, New York and Abingdon: Routledge: 171–89.
11
I am very grateful for suggestions from Michael Kremer, Alice Crary, and the two readers for this journal.
284 CORA DIAMOND
Wittgenstein, Ludwig 1961a. Notes on Logic, in Notebooks 1914–1916, eds G.H. von Wright and G.E.M.
Anscombe, Oxford: Basil Blackwell: 93–106.
Wittgenstein, Ludwig 1961b. Notes Dictated to G.E. Moore in Norway, in Notebooks 1914–1916, eds
G.H. von Wright and G.E.M. Anscombe, Oxford: Basil Blackwell: 107–18.
Wittgenstein, Ludwig 1963 (1921). Tractatus Logico-Philosophicus, trans. D.F. Pears and B.F.
McGuinness, London: Routledge and Kegan Paul.
Wittgenstein, Ludwig 1971. Prototractatus, eds B.F. McGuinness, T. Nyberg and G.H. von Wright,
trans. D.F. Pears and B.F. McGuinness, London: Routledge and Kegan Paul.
Wittgenstein, Ludwig 1979. Wittgenstein and the Vienna Circle, conversations recorded by Friedrich
Waismann, ed. Brian McGuinness, trans. Joachim Schulte and Brian McGuinness, Oxford: Basil
Blackwell.
Wittgenstein, Ludwig 2005. The Big Typescript, TS 213, ed. and trans. C. Grant Luckhardt and
Maximilian A.E. Aue, Oxford: Blackwell Publishing.
Zalabardo, J.L. 2018. The Tractatus on Unity, Australasian Philosophical Review 2/3: 250–71.