*"Sub-TRACTATUS"*

R. E. *Tully*
University of Toronto
Canada

ABSTRACT

Explicit references to Russell's ideas are numerous in the
TRACTATUS. I am less interested in discussing these than the
themes which lie just beneath the surface of the text where, I
suggest, Wittgenstein was engaged in either correcting or
modifying a number of the epistemological and metaphysical views
he had encountered in Russell's writings. Chief among these, of
course, is the concept of a proposition, a topic which they
approached in sharply different ways. But no less interesting are
the topics of logical form, internal relations, belief, the Self,
the correspondence between Realism and Solipsism, and even the
mystical attitude towards the world.



Although the following paper is as detailed as I can make it in a
brief space, it is better understood as merely sketching a larger
project which examines the TRACTATUS LOGICO-PHILOSOPHICUS*1*
through an unfamiliar lens: a theory of propositions once advanced
by Bertrand Russell but soon afterwards abandoned by him --
mainly, it seems, because of Wittgenstein's criticisms. I hope at
least to make my project clear and, more importantly, interesting
enough to be worth pursuing.*2*

    Wittgenstein's references to Russell in the TRACTATUS are
numerous and mostly critical. The theory of types, Russell's
treatment of quantification and the alleged inexactness of his
logical symbolism are familiar examples.*3* These references and
the frequent bracketing in the text of Russell's name with Frege's
help foster the impression that Wittgenstein's differences with
Russell mainly concerned formal logic and related issues of
logical theory. Yet, as important as these differences are, I
think that they are only one aspect of a fundamental criticism of
Russell's philosophy which Wittgenstein developed in the TRACTATUS
and that the source of his criticism was a bitter disagreement
between these two thinkers about the nature of a proposition.
Within the past decade important details of their discord have
come to light: during the Spring of 1913 Russell confidently began
to write a major work on epistemology and philosophical logic,
likely to have been entitled THEORY OF KNOWLEDGE, only to abandon
his plans after having completed two parts of the work because of
strong objections from his Austrian student.*4*  The criticisms
were leveled at Russell's theory of belief, which itself was
intimately tied to his concept of a proposition. A few months
later, Wittgenstein remarked in a letter to Russell that the only
way to remove the "paralysis" to which his teacher had succumbed
as a result of these criticisms was a "correct theory of
propositions".*5*

    I think the TRACTATUS aimed partly to provide such a theory,
one whose radically different approach to propositions was meant
either to solve or else to avoid the main difficulties
Wittgenstein had found in Russell's own account. Had THEORY OF
KNOWLEDGE been finished by Russell, or had he even proceeded to
publish the two parts which lay hidden for so many years, its
association with the TRACTATUS would have been obvious long ago.
Wittgenstein was obviously concerned to provide more that a purely
formal theory of propositions and their relation to facts, just as
Russell had been in THEORY OF KNOWLEDGE. The TRACTATUS contains
fragmentary features lying either at or just under the surface,
often puzzling in themselves but in fact closely related to
metaphysical and epistemological views that Russell had promoted
either in THEORY OF KNOWLEDGE or else in several other works with
which Wittgenstein was already familiar. Accordingly, an awareness
of the philosophical controversy between Russell and Wittgenstein
invites us to peer into the depths of the TRACTATUS and to
discover connections which help give unity to that work as well as
help clarify its philosophical doctrine. I shall characterize
their dispute concerning the nature of a proposition, mentioning
Russell's views first followed by Wittgenstein's replies, and then
I want to suggest that, despite the distance from Russell's
position which Wittgenstein sought to gain in the TRACTATUS, there
was yet a tacit agreement about both the nature of the phenomenon
to be analyzed and the problems to be solved.


I.
    Although Russell strove to free himself from the Idealistic
philosophy to which he had been exposed at Cambridge, there are
unmistakable traces of its influence (along with that of Meinong)
in his account of belief and proposi-tions. Like definite
descriptions, which -- unlike names -- he regarded as having no
meaning on their own, Russell classified propositions as
"incomplete symbols" whose meaning is supplied by the context of a
mental act. He viewed such acts as relations whose terms are
either objects of acquaintance or else describable with the help
of such objects. This doctrine took its most developed form in
THEORY OF KNOWLEDGE, although traces of it can be found in THE
PRINCIPLES OF MATHEMATICS, published a decade earlier, and more
clearly in "On Denoting" of 1905.*6*  Recall his claim in that
essay that "we do not necessarily have acquaintance with the
objects denoted by phrases composed of words with whose meanings
we are acquainted" [42], and the further comment near the end that
"in every proposition that we can apprehend . . . , all the
constituents are really entities with which we have immediate
acquaintance." [p. 56]  So, in apprehending that Scott was the
author of WAVERLY, a person brings various items together into a
complex cognitive relation whose terms might include at least two
objects (Scott, WAVERLY) as well as two relations ('being author
of' and 'being identical to'). Of course, the focus of Rus-sell's
essay was just the logical analysis of definite descriptions
rather than the structure of judgments which employ them. How much
more complicated he thought the epistemological issues were can be
guessed from the fact that, despite his lengthy discussion of
propositions in THEORY OF KNOWLEDGE, only names and directly
perceived relations were mentioned as values of the famous
propositional form, "aRb". Such topics as general propositions,
descriptions and molecular compounds were to have been covered in
the unwritten Part III.*7*

    I believe that two factors prompted Russell to continue
developing his theory of propositions along epistemological lines.
One was his previous dis-agreement with Meinong about the analysis
of mental contents and the nature of judgment.*8*  The other was
his uncertainty about choosing between two pos-sible analyses of
the concept of truth, prompted by his consideration of Harold
Joachim's work.*9*  Within a few years of "On Denoting", however,
the choice became clear. Russell fashioned an ingenious account of
truth which satisfied the demands of realism and yet preserved a
central role for the con-cept of acquaintance. His so-called
'multiple-relation theory' of judgment (or belief) distinguished
between facts and propositions without making the latter some kind
of mental intermediary between the person who makes a judgment and
the independent fact on which its truth depends. It also treated
propositions, and consequently whatever facts would make them true
or false, as essentially complex in nature rather than (contra
Meinong) as unitary objects of acquaintance. Corresponding to
their logical role as incomplete symbols, propositions were
described as sub-structures involving terms and relations within
the broader framework of judgment. Nowadays, the weakness in
Russell's account of judgment seems obvious: its dependency on
acquaintance, which (like that of judgment) he considered to be an
irreducibly mental relation. By the time of THEORY OF KNOWLEDGE
Russell was willing to project a hierarchy of mental acts, with
understanding now taken to be simpler than judging. He
distinguished between understanding and belief and even between
belief and disbelief, among other propositional attitudes.*10*

    The theory which attracted Wittgenstein's attention was first
presented in "The Nature of Truth and Falsehood" (1910) and then
restated in THE PROBLEMS OF PHILOSOPHY (1912), the latter version
being the one with which most philosophers today are familiar.*11*
The full epistemological and metaphysical aspects of his theory
were not evident in its first presentations, however, because
Russell's examples used familiar objects (Othello and Desdemona,
for example) instead of the more specialized objects of
acquaintance (sense-data) which he discussed in other contexts,
and in any case such features were not what drew Wittgenstein's
attention. Ostensibly, Russell's theory of belief sets out the
conditions under which the content of a belief or judgment -- the
proposition -- can be said to be true. Othello's belief that
Desdemona loves Cassio is true if and only if there exists a
relating relation ('loves') between Desdemona and Cassio which
relates them in that order. In Russell's theory, the belief or
judgment (as opposed to the fact that makes the belief true) is a
complex relation involving Othello, Desdemona, Cassio and the
relation 'loves': Othello's act of believing supposedly organizes
these components (with which he is acquainted individually) into
the right determinate order. The young Wittgenstein found this
theory inadequate: if 'loves' is merely another term related to
Othello's believing, how is one to distinguish his belief that
Desdemona loves Cassio from its converse? For that matter, what
makes 'loves' to be a relation in Othello's belief in the first
place, given that it was characterized by Russell in THE PROBLEMS
OF PHILOSOPHY as a mere brick cemented by Othello's believing? At
least two important problems confronted Russell's theory. Although
propositions involve facts in the case of true beliefs, there is
an important difference of logical type between a relating
relation in the fact and a relation taken as a mere term of
acquaintance. What accounts for this difference of type and how
does one's act of acquaintance distinguish between them? More
seriously, whenever a belief is false -- as in the Othello example
-- no fact is involved at all, though the ingredients of the
belief remain the same. So, if propositions are to involve facts
in the intimate way required by Russell's realism, the nature of
false beliefs remains a mystery.*12*

    Russell was alerted to such problems shortly after beginning
work on THEORY OF KNOWLEDGE. This may explain why he devoted so
much attention there to the analysis of relating relations
(particularly non-symmetric ones) and to identifying what he
called the "form of a relating relation". This was a new component
of Russell's analysis. He characterized it as an abstract entity,
evident to introspection (i.e., to a form of acquaintance), which
helped to determine the order of terms in a relation when one is
unacquainted with the complex described by the content of a
belief.*13*  Russell's principal aim, however, was to incorporate
this new analysis into a revised version of his multiple-relation
theory, which he unveiled in Part II of THEORY OF KNOWLEDGE. To
the terms of the cognitive relation found in the earlier version
he now added a new component which he called the 'form' of a
relating relation. His favorite example was the form of a dual
relation, symbolized as "R(x,y)". According to Russell's theory,
then, a simple case of understanding was to be represented as:

                   U{S, a, b, similarity, R(x,y)},

where 'S' represents the subject who understands ('U') that two
items ('a' and 'b') are similar.*14*  I think that Russell meant
that the form of a dual relation was logically presupposed in such
a case; he never claimed that S must discriminate this form
separately. All the same, he regarded the form as, somehow, a real
component in any case of understanding and that the above
formulation represented the structure of a fact.

    It is certainly no coincidence, however, that a great deal of
Russell's discussion in Part II dwelled on the topic of
propositions alone rather than on understanding and that portions
of two crucial chapters resemble a dialogue ('cross-examination'
might be a more accurate description) concerning details of his
revised version of the theory. Wittgenstein's opposition and
perhaps even certain of his criticisms are suggested by some of
the questions Russell confronted himself with in the text. To
illustrate, here are two which he posed:

   'Can we, by bringing in the "form" or in any other way, make
    the proposition an entity, i.e. not a mere incomplete symbol,
    but something which can subsist on its own account, and not
    only as a fictitious constituent of certain mental complexes?'

   'How can we be sure that acquaintance with the "form" is
    involved in understanding a proposition?'*15*

His discussion of these and other questions reveal an underlying
confrontation between two different theories about the nature of
propositions, the one his own, the other shadowy and unexpressed -
- probably not yet even formulated. Russell defended the view that
they are complexes -- specifically, substructures of complex
cognitional acts -- but he recognized that the "chief demerit" of
his approach was that "we cannot be sure that there are
propositions in all cases in which logic would seem to need them."
[TK, 115]  Two chapters later, he admitted that "it would seem
that we must find some non-psychological meaning for the word
'proposition'". [134]  Even after presenting his new theory of
truth a further two chapters later Russell remained haunted by the
objection that his account was too psychological in nature and
thus inadequate for the purposes of logic (which he intended to
explore in Part III of THEORY OF KNOWLEDGE): "It may be said that,
after all, there must be non-mental 'propositions' as opposed to
complexes, and that therefore beliefs had better be interpreted as
dealing with propositions". [153]  But Russell refused to succumb
to doubt. His theory of propositions was designed for
epistemology, not for logic, he contended.*16*  Besides, there was
nothing resembling a propositional entity which he was able to
apprehend. On the basis of his own introspection, then: logical
forms, yes; propositions, no. Russell adopted the stance of the
realist against Wittgenstein's views, just as he had done against
Meinong's.*17*

    Wittgenstein was apparently even more severe in criticizing
this new ver-sion and in a letter formulated an objection that is
as obscure to us today as it was apparently trenchant to Russell
in 1913. It seems to endorse the claims of logic over Russell's
epistemology. To Wittgenstein it was obvious that

    from the proposition 'A judges that (say) a is in the Relation
    R to b', if correctly analysed, the proposition 'aRb v -aRb'
    must follow directly without the use of any other premiss.*18*

Russell's theory failed to meet "this condition", Wittgenstein
asserted.*19*  Wittgenstein's objection has been taken by some
philosophers to formulate a refutation of Russell's theory.*20*
At the very least, it reaffirms the objection which Russell had
reported in THEORY OF KNOWLEDGE, namely, that his whole approach
was ill-suited to provide a theoretical account of logic, even by
means of the symbolism he had so far introduced. Russell wished to
use the expression "aRb" in two different senses: (i) as
representing a proposition (such as "a is before b"), and (ii) as
what he called a "complex name", stand-ing for a fact with which
one is presently acquainted. But when he asked him-self, "What is
meant by 'aRb consists of a and R and b united in the general form
of a dual complex?'", he confessed that such questions belong
either to a later portion of epistemology or not at all to it, or
else that, wherever they belong, he did not know the answer to
them. [128]

    Russell believed that his new version of the theory safe from
refuta-tion.*21*  Nevertheless, his confidence in it was clearly
ebbing also and, having completed Part II of THEORY OF KNOWLEDGE,
he abandoned further work on the entire project. Not until 1919
did he present a new account of proposi-tions and of truth,
different from the multiple-relation theory and wholly lacking any
role for either the form of a relating relation or, for that mat-
ter, the relation of acquaintance.


II.
    Part III of THEORY OF KNOWLEDGE -- called "Molecular
Propositional Thought" -- was supposed to have been a logical
investigation. Russell planned to examine compound propositions,
patterns of inference and our knowledge of such patterns.
Significantly, he intended to retain the same method of analysis
which intermingled questions of language, philosophical logic and
epistemology and, doubtless, he would have further described the
role of logical forms, which were a key ingredient in his newly
revised account of propositions. Russell's very concept of a
proposition seems very remote from our contemporary view: the
epistemological correlate of his claim that propositions are
incomplete symbols was that propositions themselves are dependent
on our thinking. His doctrine made it impossible to meet
Wittgenstein's demand that logic "must look after itself" [TLP,
5.473] and invited the danger of entan-gling logic itself in
"unessential psychological investigations" [4.1121]. While a
proposition for Russell was essentially connected with some mental
act or other (such as judgment), Wittgenstein saw it as
"essentially connected" with a "situation" [4.03] -- that is, with
a possible state of affairs, the sense of the proposition. The
semantic nature of propositions as signs, which Russell had
ignored, stands out prominently in Wittgenstein's claim that a
proposition not only signifies a state of affairs but also affirms
its existence [4.022]. The sense of the proposition that Desdemona
loves Cassio exists in logical space, only not as a fact.*22*
From a Russellian viewpoint, therefore, the TRACTATUS pursued two
major goals: to provide the "correct theory of propositions" that
Russell had failed to produce, and to conduct the logical
investigation that he had failed to begin.

    On the other hand, the TRACTATUS declared that propositions
    and thought are related intimately:

   "A propositional sign, applied and thought out, is a
    thought" [3.5];
   "A thought is a proposition with a sense" [4];
   "A proposition is a model of reality as we imagine it" [4.01].

Wittgenstein's paradoxical stance on the nature of propositions
and thoughts can be seen as a rigorous attempt to purify the
"Idealistic" framework within which Russell had interpreted
questions regarding philosophical logic. Revers-ing Russell's
method, Wittgenstein made propositions primary and subordinated
his philosophical account of thinking to them. He offered instead
a formal characterization of thought which lacked the most
familiar components of the Russellian scheme: the relation of
acquaintance and the act of understanding, the individuation of
objects by means of introspection, and the Self as the subject of
acquaintance and cognition. Objects in the TRACTATUS were assigned
both a metaphysical and a linguistic role, rather than a
psychological one based on acquaintance. They constitute "the
substance of the world" [2.021] and give to the world "an
unalterable form" [2.026]; they are also the meanings of names
[3.203] which, in a proposition, are the representatives of
objects. [3.22]  "The configuration of objects produces states of
affairs" [2.0272] in which "objects stand in a determinate
relation to one another". [2.031]  That names have meanings
reflects not so much a fact of experience but an objective feature
of the world: "If the world had no substance, then whether a
proposition had sense would depend on whether another proposition
was true", and so on ad infinitum. [2.0211]  While Russell's
objects of (per-ceptual) acquaintance were reclassified in the
TRACTATUS, the relations which he had described as further
constituents of facts were fundamentally reinter-preted. Just as
objects "fit into one another like the links of a chain" [2.03],
the elements of a proposition "stand in a determinate relation to
one another." [3.14]  "A proposition is articulated." [3.141]
Situations in which -- to use one of Russell's favorite examples -
- a precedes b can only be described, "but not given names".*23*
[3.144]  "Instead of, 'The complex sign "aRb" says that a stands
to b in the relation R', we ought to put, 'That "a" stands to "b"
in a certain relation says that aRb'." [3.1432]  In short,
relations among objects are shown by the propositions which
express them; relations themselves, however, like the senses of
propositions, are not named.

    A proposition and its sense have logical form in common,
although that fact too can only be shown but not said. [see 4.12]
At one point in THEORY OF KNOWLEDGE Russell also acknowledged the
special character of logical form. "It is obvious," he contended,
that "when all the constituents of a complex have been enumerated
there remains something which may be called the 'form' of the
complex, which is the way in which the constituents are combined
in the complex." [TK, 98]  Nevertheless, in a manner
characteristically platonic, Russell went on to insist that this
form can be apprehended by inspection as a simple though abstract
object, making it therefore an ingredient in the analysis of
propositional complexes. Later on, he referred to this object as
the "pure form" of a proposition which one is able to apprehend on
its own and which comprises part of the subject matter of
logic.*24*  For Wittgenstein, logical form was also a primitive
notion, although unable either to be directly apprehended or
defined in a metalanguage. Logical form makes its appearance
obliquely as something common to a proposition and its sense, more
exactly between a possible fact and any propositional sign which
correctly expresses it.*25*  "What expresses itself in language,"
however, "we cannot express by means of language," because
"propositions show the logical form of reality" -- "it is mirrored
in them". [4.121]  It is a well known doctrine of the TRACTATUS
that we cannot get outside the world to describe logical form, for
that would require going beyond the limits of logic itself.*26*
Wittgenstein transformed Russell's concept of logical form into a
metaphysical axiom intended to bond propositions to the world by
reckoning them as a sub-set of facts. Perhaps it is a contingent
matter that there are propositions, that the world is represented
in any manner at all. Given that there are propositions, however,
their relation to what they express is fixed by logic and is a
necessary one. The connection between a propositions and its sense
is "precisely that it is its logical picture". [4.03]

    From the world described by Wittgenstein, the world examined
by logic, the Self of Russell's epistemology has been extruded,
lingering only in the sense of being a "limit" of the world
[5.632] and not found among its objects. Ironically, Russell
himself had been of two minds about the metaphysical role of the
Self. In 1910 he had ranked it along with sense-data, universals
and relations as one of the terms of the cognitive relation of
acquaintance, but this view changed in THEORY OF KNOWLEDGE, where
he was willing to characterize the Self as something known only by
description, and possibly not by nature mental at all.*27*
However interpreted, the Self was required by his dualistic
epistemology. In less formal writings belonging to the same
period, on the other hand, Russell adopted a Spinozistic view of
the Self as something capable of transcending the bonds or
"prisons" of selfishness in order to achieve mystical union with
the universe. This rather different view might be called the 'non-
Self' conception of the Self. It is remarkably similar to
Wittgenstein's declaration in the TRACTATUS that the Self does,
after all, enter into philosophy as "the metaphysical subject, the
limit of the world" [5.641], the Self which views "the world sub
specie aeterni as a whole a limited whole".*28*  It is worth
noting that Wittgenstein was very familiar with -- and originally
scandalized by -- Russell's metaphysical speculations about the
non-Self*29*, and so it is ironic that he felt free to incorporate
remarks about the non-Self, once the need for a Self of any kind
within the Tractarian framework had been removed. (It is perhaps
even more ironic that the Self fared less happily at Russell's
hands. Once he had abandoned the con-cept of acquaintance, he
ceased to ruminate about it altogether as enjoying a more sublime
role.)

    The question of the correct analysis of belief which received
so much attention in THEORY OF KNOWLEDGE was curtly dismissed by
Wittgenstein in the TRACTATUS. He maintained that "theory of
knowledge is the philosophy of psychology," that psychology is "no
more closely related to philosophy than any other natural
science," and that the job of philosophy, which is not itself a
natural science, is to engage in the activity of logically
clarifying thoughts [4.111-4.1121]. Moreover, he discounted any
philosophical arguments on behalf of either philosophical monism
or the dualistic position which Russell then favored (4.128).
Wittgenstein's cryptic analysis of sentences expressing belief was
apparently meant to bear out these points. Elementary
propositions, in themselves independent of each other, constitute
bases for "truth-operations", despite the appearance of one
proposition being nested within another. The natural home of "p"
is not located within another proposition, since, according to
Wittgenstein, the form of a sentence like `A believes that p'
merely expresses a correlation between a propositional sign ("p")
and a proposition (p). [5.54-5.542]  (At this point in the
TRACTATUS (5.5422), Wittgenstein stated an objection to Russell's
theory of belief in a manner which echoes the more famous one
mentioned above.*30*) Whatever is of purely philosophical interest
regarding belief was supposed to be already pro-vided for by means
of elementary propositions and the compound forms generated from
them. To Wittgenstein, his teacher's interest in distinguishing
acts of understanding, belief, disbelief and other propositional
attitudes was scientific in spirit but not strictly philosophical.
Philosophical analysis does not produce new truths. "The totality
of true propositions" for Wittgenstein was "the whole of natural
science." [4.11]

    In the midst of his discussion of the Self, Wittgenstein
declared that Solipsism and Pure Realism coincide (see 5.64).
Commentators have puzzled over this cryptic remark. In fact,
Wittgenstein's comment points in the same direction in which
Russell's own philosophy was headed during the period of THEORY OF
KNOWLEDGE. Russell had clearly sympathized with the philosophical
stance taken by the so-called New Realists, whose spiritual
founders were William James and Ernst Mach, even to the extent of
adopting several of their views, but he ardently disagreed with
their behavioristic analysis of the concept of belief as well as
with their attempt to eliminate the relation of acquaintance from
the foundations of epistemology.*31*  The few illustrative remarks
which Wittgenstein made about objects in the TRACTATUS are
consistent with the views held by both Russell and the New
Realists concerning the basic data of experience.*32*  Given this
philosophical background, considerably less vivid today than it
would have been earlier in the century, it appears that
Wittgenstein's critical treatment of the Self and of belief,
together with his relegation of much of epistemology to the
science of psychology, are much closer in spirit to Neutral
Monism, the very doctrine that Russell himself would eventually
endorse, than to any other view of the time. The objects of the
TRACTATUS suggest the "neutral stuff" which the New Realists
regarded as common to both physics and psychology but, in itself,
as essentially neither mental nor physical. Perhaps this is what
Wittgenstein meant by the 'coincidence' of Solipsism and Pure
Realism. He may well have thought that his account of the logical
nature of objects revealed a deeper metaphysical side to the
doctrine of Neutral Monism.*33*


III.
    Unlike Russell, Wittgenstein was not exercised by the
difference between symmetric and asymmetric relations. In any
event, he obviously thought that the Russellian analysis of
relations was not needed to account for the specific order of
terms in a two-termed relation. The objects in a possible state of
affairs have the order which is fixed unambiguously by the
propositional sign that contains their names: the configuration of
names determines part of the proposition's sense.*34*  Evidently,
Wittgenstein looked to logical syntax to eliminate potential
ambiguities from these forms of expression, but while he
emphasized the requirements of syntax he did prac-tically nothing
in the TRACTATUS to illustrate them or to test their adequacy. In
contrast, Russell used the technique of introspection not only to
isolate and identify different types of relations (referring to
these forms as "logical data") but also to penetrate the sense of
actual relations, as given in experience, beyond what is suggested
by descriptions in ordinary language. Though their methods for
analyzing relations differed, as well as their results, Russell
and Wittgenstein were hunting the same quarry. A major difference
is that Wittgenstein distinguished between relations in an
'internal' as well as an 'external' sense and probably thought
that Russell had failed to meet the demands of philosophy by his
preoccupation with the latter, that is, with the experience of
relations instead of their essential features.*35*  The
distinction was nevertheless a difficult one for Wittgenstein to
make since internal properties themselves are not able to be
described but, like logical form, "make themselves manifest" in
propositions representing external properties.*36*  In paragraph
4.123 of the TRACTATUS Wittgenstein seems to have intended a close
comparison between a given pair of visual objects, which stand in
the relation of lighter and darker, and Tractarian objects in
general. Just as it is "unthinkable that these two objects" should
not stand in this relation, he wrote, so it is unthinkable that an
object should fail to possess its internal properties. However,
Wittgenstein was aware that the word 'object' in this comparison
had a "shifting use". The relation of 'being darker than' in his
example comes as near as any external property could to being an
internal one but the difference between them remained fundamental.

    Why stress that difference? The later Wittgenstein was well
aware of what had captured his attention years earlier. Logic
seemed to lie "at the bottom of all the sciences," he wrote in the
PHILOSOPHICAL INVESTIGATIONS. Logical investigations were driven
by "an urge to understand the basis, or essence, of everything
empirical".*37*  And that includes thought itself, whose essence,
logic, "presents . . . the a priori order of the world: that is,
the order of possibilities, which must be common to both world and
thought." [PHILOSOPHICAL INVESTIGATIONS, I:97]  It is evident that
Wittgenstein's former interests were not narrowly formalistic --
indeed, he might even have associated purely logical research with
mathematics and therefore as something to be distinguished from
philosophy proper. I think that the enterprise of logic
represented to the young Wittgenstein a pure method of discovery
independent of natural science. The facts it uncovers,
characterized in terms of objects and structure, logical form and
internal relations, define the nature of what there is in its
barest sense, prior to any meaning that is formulated in
scientific descriptions or based on experiences of the sort
promoted by Russell who, in Wittgenstein's eyes, had developed the
tools of logic well enough without ever grasping their proper use.
Although his teacher had "performed the service of showing that
the apparent logical form of a proposition need not be its real
one [4.0031], the real service to philosophy which logic could
perform of clarifying the essence of propositions was yet to be
rendered.

    Wittgenstein held that objects have an "internal relation" to
the states of affairs which they constitute (2.0123-2.0141;
4.123); also, that proposi-tions themselves have internal
relations to each other which express the internal relations
between possible states of affairs. He thus wished to emphasize
that relations have an inherently logical character in addition to
whatever "external" features of objects happen to be represented
by a proposi-tion.*38*  Similarly, Wittgenstein spoke of an
"internal relation" for a series of increasingly generalized
propositions derived from the basic proposition, such as 'aRb'
[4.1252]. Russell's quite different approach had been to treat
generalized forms of propositions as "pure forms" capable of being
apprehended by a subject; not only that, he had regarded the pure
form of a proposition as a simple object of acquaintance and
therefore able to be named, a view that would later be proscribed
in the TRACTATUS ("Propositions are not names of facts").
Ostensibly, Wittgenstein's aim was to give "the answer to the
vexed question 'whether all relations are internal or external'"
[4.1251], but I believe that he meant also to correct a deep flaw
in Russell's philosophical logic. Russell had once battled against
Bradley's Idealism over the question of whether relations are
internal or external, defending (as is well known) the latter
alternative. In rejecting internal relations, Russell had turned
his back on the very aspect of logic which defines its nature and
its independence from all those psychological aspects of logical
thinking which not only had retarded progress in this abstract
science but also prevented a proper understanding of the nature of
propositions. Preoccupied with the goal of showing how knowledge
(even logical knowledge) is grounded in experience, Russell had
failed to appreciate the objective nature of logical necessity.
This is perhaps the most significant part of the hidden message of
the TRACTATUS.


Conclusion.
    However tempting it may be to look upon the TRACTATUS as an
utterly singular creation of genius, the evidence I have given
shows that it did not spring unaided from Wittgenstein's forehead.
History can unmake some myths. Despite its originality and
breadth, and despite its striking advance-ment of the cause of
logic, the TRACTATUS was in the first instance a work of rejection
and reorientation based on logical intuitions vastly different
from those which Wittgenstein had encountered in the writings and
teaching of Rus-sell. The work aimed to change the course which
Russell (and Frege) had given to modern logic. It redefined the
central concepts of Russell's philosophical logic, above all what
it is to be a proposition and how propositions should be
symbolized. In place of Russell's sense-data the TRACTATUS offered
merely unanalyzable objects. It conceived logical forms and
relations entirely in terms of structure rather than as objects of
acquaintance. Wittgenstein renounced Russell's analysis of belief
and thereby succeeded in bypassing the corrosive problem of false
beliefs. He banished the Self from philosophical analysis but also
gave new prominence to the idea of logical necessity by means of
the concept of internal relations.

    Such a catalogue of contrasts makes it seem that Wittgenstein
had turned his back completely on Russell's philosophy. Indeed,
the list becomes even longer when we recall that the TRACTATUS
included remarks about probability, hypotheses, causality,
generalizations, self-evidence, the nature of value, death, the
ground of ethics, and mystical contemplation -- topics which
Russell too had examined. From one perspective, then, the
TRACTATUS is a work of stark contrast, a post-Great War revision
of THE PROBLEMS OF PHILOSOPHY in the spirit of Eliot's "The
Wasteland". Nevertheless, not all the differences between two
things are complementary. Apart from the doctrine of names and
logical simples, there is one central area in which the doctrine
of the TRAC-TATUS in fact blends with and continues Russell's
philosophy. Despite his repudiation of the psychological
tendencies in Russell's epistemology, the doctrine of acquaintance
lived on, I believe, in Wittgenstein's picture theory of
propositions and in his concept of showing. Russell's cognitive
acts were transmuted by Wittgenstein into the picturing of facts
and the understanding of propositions into the showing of possible
states of affairs. For both philosophers, propositions were
structures of objects dependent on thinking, and although their
views about these structures quickly diverge when placed side by
side, the fact that this was their common starting point suggests
that Wittgenstein broadly agreed with Russell that to give an
account of propositional thinking was a job reserved for
philosophy. That "we picture facts to ourselves" [2.1] is the
phenomenon to be explained, though in logical terms. This helps
explain why their analyses of the concept of truth did not
diverge; both of them advocated a correspondence theory between
the sense (or content) of a proposition with the fact on which its
truth depends. The nature of this relation of correspondence is no
less puzzling in the TRACTATUS than it was in THEORY OF
KNOWLEDGE.*39*

    The word "logic" in the TRACTATUS displaced "epistemology" but
only by acquiring some of its associations. When he declared that
propositions are thoughts "applied and thought out" [3.5],
Wittgenstein had in mind the product (namely, signs) of the same
type of process (understanding) that had engaged Russell's
attention, but by focusing on the vehicle of representation
instead of experience he was able to avoid the psychological
emphasis which in fact would always permeate Russell's philosophy.
While impressive, Wittgenstein's achievement was not supreme,
however. The pursuit of what is abstract, the relentless endeavor
to penetrate beyond all "empirical cloudiness" in search of "the
purest crystal" [PI, I:97], can result in obscuring some of the
ordinary facts most in need of analysis, as Wittgenstein himself
would eventually realize. "But first we must learn to understand
what it is that opposes such an examination of details in
philosophy," he would declare in the INVESTIGATIONS [I:52]. In the
TRACTATUS he thought that the need to grasp the essence of
phenomena, "the truth itself in its entirety" [5.5563], was not
only a correct answer but also the only correct one left for
philosophy to give.


                            ENDNOTES

*1*
Ludwig Wittgenstein, TRACTATUS LOGICO-PHILOSOPHICUS, trans. by D.
F. Pears and B. F. McGuiness (London: Routledge & Kegan Paul,
1963). (Quotations from the TRACTATUS throughout this 'electronic'
article are from this translation. (PLEASE NOTE: Italicized words
in these quotations have not been preserved in this version of the
article.) Where appropriate, the title will be abbreviated as
"TLP".

*2*
This is a revised and considerably enlarged version of the paper
which I gave at the 16th International Ludwig Wittgenstein
Symposium Kirchberg am Wechsel, Austria: 15-21 August, 1993, and
which was published in PHILOSOPHY AND THE COGNITIVE SCIENCES, vol.
I, ed. Roberto Casati and Graham White (Kir-chberg am Wechsel: The
Austrian Ludwig Wittgenstein Society, 1993), pp. 537-541.

*3*
See (following the order of the three areas mentioned) TLP: 3.331-
3.333; 5.521, 5.525; 3.325, 4.1272-4.1273, 5.42, 5.452, 5.5302,
5.5351.

*4*
The unfinished work was published posthumously as THEORY OF
KNOWLEDGE, ed. E. R. Eames in collaboration with K. Blackwell
(London: George Allen and Unwin, 1984). (The title of this work
will be abbreviated in references as TK.)

*5*
See TK, Introduction, xxviii.

*6*
B. Russell, LOGIC AND KNOWLEDGE, ed. R.C. Marsh (London: George
Allen & Unwin Ltd., 1956), 41-56.

*7*
See TK, Appendix C, p. 201.

*8*
See the 1904 article, "Meinong's Theory of Complexes and
Assumptions" as well as two later reviews of Meinong in B.
Russell, ESSAYS IN ANALYSIS, ed. Douglas Lackey (New York: George
Braziller, 1973).

*9*
See his review of Joachim's The Nature of Truth in MIND, n.s. 15
(1906), 528-33 and his essay "On the Nature of Truth and
Falsehood", PROCEEDINGS OF THE ARISTOTELIAN SOCIETY, n.s. 7, 1906-
7, 28-49.

*10*
See TK, Part II, Chapter 1, "The Understanding of Propositions".
While remaining critical of Meinong's views in this work, Russell
was nonetheless more accommodating than previously.

*11*
See PHILOSOPHICAL ESSAYS (London: George Allen & Unwin Ltd., rev.
ed., 1966 [orig. pub. 1910]), pp. 147-159 and THE PROBLEMS OF
PHILOSOPHY (Oxford: Oxford University Press, paperback edition,
1967, Chapter 12.

*12*
Russell seems to be recalling this problem in his Logical Atomism
lec-tures when he told his audience that Mr. Wittgenstein had
discovered "a new beast for our zoo". See LOGIC AND KNOWLEDGE, p.
226.

*13*
I have discussed this and related parts of Russell's account in
"Three Studies of Russell's Neutral Monism", in the journal,
RUSSELL (Winter, 1993, pp. 3-21).

*14*
See TK, p. 117. (I have altered the symbolism slightly here by
rendering the two similar items in lower case, which conforms to
Russell's practice elsewhere in THEORY OF KNOWLEDGE.)

*15*
See p. 113; for further examples, see p. 128 and 153-5.

*16*
See 155; there is evidence in TK that Russell revised the work by
adding a new chapter which stressed the distinction between logic
and epistemology. See TK, Introduction, pp. xxxviii-xxxix. In a
territorial sense, perhaps Russell was abandoning the field of
logic to his student. The TRACTATUS attacked Russell's
epistemology on a broad front, nonetheless.

*17*
See "On the Nature of Truth and Falsehood", pp. 150-3.

*18*
For emphasis, Wittgenstein italicized the phrase 'without the use
of any other premiss'.

*19*
Quoted from the Introduction to TK, xxvii (minor editorial changes
made). I have discussed Wittgenstein's objection in "Thinking of
Propositions", I have discussed Wittgenstein's objection in
"Thinking of Propositions" (the third of my "Three Studies of
Russell's Neutral Monism") in RUSSELL, Winter, 1993, pp. 3-21.

*20*
See Nicholas Griffin, "Was Russell Shot or Did He Fall?",
DIALOGUE, Vol. XXX, No. 4 (Fall 1991), 549-553.

*21*
See TK, p. 135.

*22*
It is worth recalling that, for Wittgenstein, "true and false" are
not "relations of equal quality between signs and what they
signify" [4.061], nor is it possible to "make ourselves understood
with false propositions" [4.062]. Although a false proposition has
no fact corresponding to it, still it has a sense.

*23*
Exactly this is what Russell had tried to do in the third chapter
of Part II of TK.

*24*
See TK, 113-4 and (esp.) 132-3.

*25*
A propositional sign itself is also a fact [see 3.14], a claim
which makes Wittgenstein's correspondence theory of truth appear
to be a straightforward matter of relating one overt structure,
consisting of names, to another one of objects. I think that the
appearance is deceiving for reasons discussed briefly at the end
of the paper.

*26*
See 5.61.

*27*
See "Knowledge by Acquaintance and Knowledge by Description", in
MYSTICISM AND LOGIC (London: George Allen and Unwin Ltd.), pp.
209-32 (regard-ing the Self, see esp. p. 224 and accompanying
footnote 2; also TK, pp. 36-7.

*28*
6.45. The Latin phrase alludes to Spinoza and, I think, to
Russell's Spinozism.

*29*
See details about Russell's contemplative piece, "Prisons", and
his essay, "The Essence of Religion", in THE COLLECTED PAPERS OF
BERTRAND RUSSELL, Volume 12, ed. by R. A. Rempel, A. Brink and M.
Moran (London: George Allen and Unwin, 1985). Some of the passages
in "Prisons" found their way into the closing chapter of THE
PROBLEMS OF PHILOSOPHY, where Russell described "philosophic
contemplation" as the enlargement of the Self which "achieves some
share in [the] infinity" of the universe." [p. 92]  Russell's
expression "the not-Self" designates the logical complement of the
Self. The aim of contemplation is to eliminate self-assertion,
thereby reducing the Self to virtual non-existence. It is the
latter, not its logical complement, that I mean by "non-Self" in
the paper.

*30*
The "requirement" which Russell's theory of judgment is claimed
not to fulfill is that it is "impossible for a judgment to be a
piece of nonsense." This seems to differ from the requirement
which Wittgenstein had communicated to Russell in June of 1913,
during the period of their dispute about certain portions of
THEORY OF KNOWLEDGE. Wittgenstein may be alluding in the TRACTATUS
only to an earlier version of Russell's theory, such as published
in THE PROBLEMS OF PHILOSOPHY and which Russell himself admitted
during his logical atomism lectures to have been "a little unduly
simple", because it had not provided for the type difference
between relations and other objects of acquaintance (see LOGIC AND
KNOWLEDGE, p. 226). It is worth noting that Russell had no
solution to offer during those lectures to the problem of belief.

*31*
I have discussed Russell and the New Realists (also called Neutral
Monists) in "Russell's Neutral Monism", (in ANTINOMIES AND
PARADOXES, ed. Ian Winchester and Kenneth Blackwell [Hamilton:
McMaster University Library Press, 1989], pp. 209-224).

*32*
See 2.0131 and 2.0251.

*33*
By the time the TRACTATUS was published, of course, Russell
himself had already adopted a form of Neutral Monism.

*34*
See 3.21.

*35*
Wittgenstein also used the word 'material' to describe external
proper-ties: ". . . it is only by means of propositions that
material properties are represented only by the configuration of
objects that they are produced." [2.0231]

*36*
4.122; also, see 4.124.

*37*
Ludwig Wittgenstein, PHILOSOPHICAL INVESTIGATIONS, tr. G. E. M.
Anscombe (Oxford: Basil Blackwell, 1963), I:89.

*38*
See 4.023 and 4.122.

*39*
The correspondence in the case of a belief cannot be between a
mere propositional sign and a fact (as Wittgenstein suggested at
5.542), but between a propositional sign "applied and thought out"
and its sense.



Prof. R. E. Tully
St. Michael's College
81 St. Mary Street
Toronto, ON
M5S 1J4
CANADA

e-mail: tully@epas.utoronto.ca
FAX: 416-593-1041
TEL: 416-926-1300 ex. 3223