Академический Документы
Профессиональный Документы
Культура Документы
ABSTRACT. In this paper, I try to uncover the role played by Wittgenstein's context
principle in his criticism of Russell's theory of types. There is evidence in Wittgenstein's
writings that a syntactical version of the context principle in connection with the theory
of symbolismfunctions as a good reason for his dispensing with the theory of types.
1. INTRODUCTION
The aphorism 3.3 of the Tractatus has been the source of some debate
about the interpretation of Wittgenstein's early philosophy:
Only the proposition has sense; only in the nexus of a proposition does a name have
meaning. (1961, sec. 3.3)
The majority of Wittgenstein's commentators recognise in this aphorism
the influence of Frege's context principle, which is most clearly stated
at the beginning of The Foundations o f Arithmetic as one of Frege's
fundamental methodological principles:
[N]ever to ask for the meaning of a word in isolation, but only in the context of a
proposition. (1953, p. x)
The bias according to which commentators have mainly interpreted
this central thesis in the Tractatus is semantic, i.e., it has almost always
been seen as suggesting a theory of meaning or of object-construction.
It is not my aim here to deprive this interpretative tendency of its
legitimacy, but it seems clear that the semantic question is not among
Wittgenstein's primary motivations at the very beginning of his philos-
ophy. In my opinion, if the semantic bias today seems to be one of the
most attractive aspects of the Tractatus, this is because of the great
emphasis placed on questions concerning language in contemporary
philosophy.
What I intend to do here is to elaborate the role played by the context
principle in the criticism of Russell's theory of types, as formulated by
Wittgenstein in his pre-Tractatus writings. The context principle as a
central thesis of the very early Wittgenstein has the negative role of
showing the dispensability of the theory of types, as formulated in
follows that a function cannot have among its values an object that
presupposes the function itself; in other words, expressions of the form
" , ~ ( ~ ) " are not well formed, for in this case we could not consider the
totality of its values as intelligible before the function itself is intelli-
gible, thus generating a vicious circle.
Russell's second argument is based on the explanation, by means of
direct inspection of the corresponding "logical intuition", of the nature
of what corresponds to the word "function" and of what differentiates
it from the nature of individuals. This difference in nature between
function and individuals implies that not only expressions of the type
"p(~2)" are nonsensical, but also expressions of the form "0(~z)",
when "0~" and " ~ " denote any functions for which there exists an
argument a such that °' q)a" and "~a" were both meaningful (i.e., when
and p belong to the same logical type). Russell's approach to the
nature of a function is one of the most important points of this chapter
of Principia. It contains observations like the following:
The question as to the nature of a function is by no means an easy one. It would seem,
however, that the essential characteristic of a function is ambiguity. (Ibid., p. 39)
and later:
A function, in fact, is not a definite object, which could be or not be a man: it is a mere
ambiguity awaiting determination, and in order that it may occur significantly it must
receive the necessary determination, which it obviously does not receive if it is merely
substituted for something determinate in a proposition. (Ibid., p. 48)
We may regard the function itself as that which ambiguously denotes, while an undeter-
mined value of the function is that which is ambiguously denoted. (Ibid., p. 40)
function itself. And this direct inspection involves regarding the func-
tion as having a kind of thing-hood.
Finally, Russell presents in Section V of this chapter the famous
hierarchy of logical types for functions and propositions. The rigorous
observance of this hierarchy would suffice to avoid the emergence of
anfinomies in the construction of logical systems.
The very idea of a theory of types as Russell conceives it raises
crucial questions about the nature of its method and its epistemological
character. What, after all, is it talking about? And is its procedure
legitimate? As regards the theory itself, it gives the impression of telling
us something new, i.e., presenting new information. But from where
does this form of knowledge get its appearance of novelty and, at the
same time, of necessary truth? Russell's position with respect to this
point is to be linked with his general ideas on epistemology of logic. A
good synthesis of Russell's thought on this topic can be found in his
'Logical Data' of 1913. There, Russell provides an account of what he
calls "acquaintance with logical objects". He sees acquaintance, in a
formalised way, as a relation established between two objects, one of
which is the knowing subject and the other the object of knowledge:
[E]xperience is a dual relation of subject to an object, though it is not necessary to assume
that we experience either the subject or the relation, but only the object and (sometimes)
the complex subject-experiencing-object. Assuming that this analysis is correct, we called
the dual relation in question "acquaintance". (Russell, 1979a, p. 100)
According to the kind of object that occupies the second position, in the
experience-relation we shall have also different kinds of acquaintance:
We then considered various kinds of acquaintance. The first classification is according to
the logical character of the object, namely, according as it is (a) particular, (b) universal,
or (c) formal, i.e. purely logical. (Ibid.)
The existence and the nature of the third kind of acquaintance men-
tioned in the quotation, i.e. with logical objects, is the main preoccupa-
tion of the text in question, and it is made possible on the basis of two
transcendental arguments presented by Russell. First, for our discourse
on terms such as "function", "relation", "logical form", etc., to be
intelligible, as it in fact is, it is necessary that objects correspond to
these notions with which we are already familiar:
It is not at all clear what is the right logical account of "form", but whatever this account
may be, it is clear that we have acquaintance (possibly in an extended sense of the word
CRITICISM OF RUSSELL'S THEORY OF TYPES 405
"acquaintance") with something as abstract as the pure form, since otherwise we could
not use intelligently such a word as "relation". (Ibid., pp. 98-99)
If we are acquainted with a and with similarity and with b, we can understand the
statement "a is similar to b ' , even if we cannot directly compare them and "see" their
similarlity. But this would not be possible unless we knew how they are to be put together,
i.e., unless we are acquainted with the form of a dual complex. Thus all "mental
synthesis", as it may be called, involves acquaintance with logical form. (Ibid., p. 101)
Besides the forms of atomic complexes, there are many other logical objects which are
involved in the formation of non-atomic complexes. Such words as or, not, all, some,
plainly involve logical notions; and since we can use such words intelligently, we must
be acquainted with the logical objects im~olved. (Ibid., p. 99)
And further: every theory of types must be rendered superfluous by a proper theory of
symbolism . . . . What I am most certain is not however the correctness of my present
way of analysis, but of the fact that all theory of types must be done away with by a
theo13J of symbolism showing that what seem to be different kinds of things are symbolised
by different kinds of symbols which cannot possibly be substituted in one another's place.
(Wittgenstein, 1974, pp. 19-20)
If this objection does not yet seem completely articulated, the oppo-
sition between the recognition of a field of logical objects as what
confers significance on logical discourse and Wittgenstein's conception
of philosophy (already present in the 'Notes on Logic') is nevertheless
quite clear:
Philosophy gives no picture of reality.
Philosophy is the doctrine of the logical form of scientific propositions (not only ot:
primitive propositions). (Wittgenstein, 1979a, p. 106)
That logic should "take care of itself" here expresses the philosophical
demand that logic should not be made possible by recourse to some
kind of reality (be it empirical, psychical, or even logical). The impossi-
bility of error in logic which Wittgenstein mentions here is not due to
a special kind of rigour in its procedure or to the exactness of :its
method, but rather to the nature of its task. Logic should deal only with
the description of syntactical rules for the use of symbols:
In philosophy there are no deductions; it is purely descriptive. (Wittgenstein, 1979a, p.
106)
the 'Notes on Logic', but, there, it is still not in connection with the
criticism of the theory of types.)
The most general formulation of an answer to the question about
what symbolises can be found below:
What symbolizes in &~ is that ch stands to the left of a proper name and obviously this
is not so in - p . What is common to all propositions in which the name of a property (to
speak loosely) occurs is that this name stands to the left of a name-form, (Wittgenstein,
1979b, p, 116)
And further:
~b cannot possibly stand to the left of (or in any other relation to) the symbol of a
property. For the symbol o f a property, e.g., O/x is that q~ stands to the left o f a name
form . . . . (Ibid.; my emphasis, except for "that")
We can see from both quotations that what symbolises for Wittgenstein
are the signs "4~" or "~" in a certain position in relation to neighbouring
signs (those of a different logical type). In other words, what symbotises
is the perceptible sign plus its relative position in propositional signs.
Position here is understood as some kind of relation (spatial or tempo-
ral, according to the nature of the sign employed) that the sign main-
tains with other signs which are members of the complete propositional
sign. In my opinion, this is not more than the context principle in a
special formulation which I shall call syntactic, 4 i.e., only in the articu-
lated context of perceptible signs .forming a propositional sign does a
simple sign acquire symbolic nature. We can find a clear indication
about this point in the following passage:
The reason why, e.g., it seems as if "Plato Socrates" might have a meaning, while
"Abracadabra Socrates" will never be suspected to have one, is because we know that
"Plato" has one, and do not observe that in order that the whole phrase should have
one, what is necessary is not that "Plato" should have one, but that the fact that "Plato"
is to the left o f a name should. (Ibid.)
"Plato" as such is merely a perceptible sign, but the .fact that "Plato"
occurs to the left of another name is what has meaning.
For Wittgenstein, what plays the role of a theory of types is merely
showing, for a symbol, the relation that its perceptible part has to the
perceptible part of the other symbols of the propositional symbol (e.g.,
being to the left of a name, or between two names, or to the right of
a relation, etc.). The whole theory of types is, hence, reduced to talking
about the characteristic positions of certain symbols in the propositional
410 MARCO ANTONIO RUFFINO
We may see here how the famous doctrine of what can be said and of
what can be shown is a natural consequence of a correct answer to the
question of what symbolises. We need to make the distinction between
C R I T I C I S M OF R U S S E L L ' S T H E O R Y OF TYPES 411
This becomes clear when one considers that in "0a", for example, one
cannot say that " 0 " occurs to the left of "a" without it also being
implicit that "a" occurs to the right of " 6 " . The complementarity
between logical types is thus a natural consequence of the way in which
Wittgenstein conceives the theory of types, i.e. as a theory of symbolism
based on the context principle, and means simply that, in the partition
of the symbol of a complete proposition, complementary symbols are
obtained.
In Russell's theory, to the contrary, it seems that this idea of comple-
mentarity is somewhat problematic. For, as we have seen, Russell
allows himself to consider the nature of a function in itself, by direct
speculation on the "logical intuition" corresponding to the term "func-
tion". It follows that the thesis according to which "function" and
"individual" seem to denote complementary logical types is due to a
property of the arrangement of logical objects corresponding to these
terms.
A second remarkable feature of Wittgenstein's theory of symbolism
is that self-reference or self-predication is not only an error, but is also
actually impossible. For, if we write "(q~(q~))", even if we are using
the same letter for function and argument, the constituted symbols are
in fact different, since what symbolises is not only the sign but also the
fact that "q¢' occurs at the left of a name (or, if " ~ " symbolises a
412 MARCO ANTONIO RUFFINO
function of second-order, the fact that " 9 " occurs at the left of a first-
order function symbol). As Wittgenstein said later in the Tractatus,
what is common to both functions in the expression above is only the
letter q~, which
by itself signifies nothing, (1961, sec. 3,333)
4. CONCLUSION
NOTES
* I would like to thank Michael Wrigley (UNICAMP) and Gottfried Gabriel (Universit~it
Konstanz) as well as an anonymous journal referee for their valuable comments on a
previous version of this paper.
1 A detailed account of the evolution of Russell's opinions concerning the theory of types
is given in Cocchiarella (1980, 1989) and Consuegra (1989); the ontological implications
of Russell's logical doctrines is the main topic of Quine (1966).
2 For a reconstruction of the Wittgenstein-Russell dialogue during the period of the pre-
C R I T I C I S M OF R U S S E L L ' S T H E O R Y OF TYPES 413
Tractatus writings, see McGuinness (1972, 1974), Btackwell (t981), Hacker (1986, pp.
1-27), and Ishiguro (1981).
3 For this reason, Wittgenstein's general attitude towards the theory of types may seem
contradictory at first sight. Wittgenstein does indeed reject the possibility of a theory of
types in the Russellian style, but recognises the legitimacy of a theory of types restricted
to the different kinds of symbols. The contradiction is, therefore, merely apparent.
4 Here I am using the term 'syntactic' not in the modern sense of linguistics, i.e. as
opposed to 'semantics', but rather in the same sense that Wittgenstein uses in the
Tractatus (3.33) of the "logische Syntax" as a general descriptive theory concerning logical
properties of occurrences of symbols. 'Syntactic' seems to me an adequate term to
characterise a style of investigation that gives priority to formal features of linguistic
expressions and logical relations between sentences containing these expressions as rel-
evant data for the formulation and solution of philosophical questions.
REFERENCES
Blackwell, K.: 1981, 'The Early Wittgenstein and the Middle RusseIl', in Block (1981),
pp. 1-31.
Block, I. (ed.): 1981, Perspectives on The Philosophy of Wittgenstein, MIT Press, Cam-
bridge.
Cocchiarella, N.: 1980, 'The Development of the Theory of Logical Types and the Notion
of a Logical Subject in Russell's Early Philosophy', Synthese 45, 71-116.
Cocchiarella, N.: 1989, 'Russell's Theory of Logical Types and the Atomistic Hierarchy
of Sentences', in C. Savage and A. Anderson (eds.), Rereading Russell, Minnesota
Studies in the Philosophy of Science, Vol. XII, University of Minnesota Press, Minnea-
polis, pp. 41-62.
Consuegra, F.: 1989, 'Russell's Theory of Types, 1901-1910: Its Complex Origin in the
Unpublished Manuscripts', History and Philosophy of Logic 10, 131-64.
Frege, G.: 1953, The Foundations of Arithmetic, 2nd ed., trans. J. L. Austin, Blackwell,
Oxford (originally published in 1884).
Hacker, P.: 1986, Insight and Illusion, rev. ed., Clarendon Press, Oxtord.
Ishiguro, H.: 1981, 'Wittgenstein and the Theory of Types', in Block (1981), pp. 43-59.
McGuinness, B.: 1972, 'Bertrand Russell and Ludwig Wittgenstein's Notes on Logic',
Revue Internationale de Phifosophie 26, 444-60.
McGuinness, B.: 1974, 'The Grundgedanke of the Tractatus', in G. Vessey (ed.), Under-
standing Wittgenstein, Macmillan, London, pp. 49-61.
Quine, W. V.: 1966, 'Russell's Ontological Development', The Journal of Philosophy
LXIII, 657-67.
Russell, B.: 1984, 'Logical Data', in Russell, Theory of Knowledge: The 1913 Manuscript,
Vol. 7, The Collected Papers of Bertrand Russell, ed. by E. Eames, Allen & Unwin,
London, pp. 97-102.
Russell, B., and A. Whitehead: I973, Principia Muthematica, repr., VoL I, Cambridge
University Press, London (repr. of 1910 original).
Wittgenstein, L.: 1961, Traemtus Logico-Philosophicus, trans. D. Pears and B, McGuin-
ness, Routledge and Kegan Paul, London.
414 MARCO ANTONIO RUFFINO
Wittgenstein, L.: 1974, Letters to Russell, Keynes and Moore, ed. G. H. yon Wright,
Blackwell, Oxford.
Wittgenstein, L.: 1979, Notebooks 1914-1916, 2nd ed., ed. G. H. yon Wright and G. E.
M. Anscombe, trans, G. E. M, Anscombe, Blackwell, Oxford.
Wittgenstein, L.: 1979a, 'Notes on Logic', in Wittgenstein (1979), pp. 93-104 (originally
written in 1913).
Wittgenstein, L.: 1979b, 'Notes Dictated to G. E. Moore in Norway', in Wittgenstein
(1979), pp. 108-19 (originally written in 1914).
Department of Philosophy
University of California
405 Hilgard Avenue
Los Angeles, CA 90024-1451
U.S.A.