Frege and Russell on the concept of “Number”: A Critique

Seldom do we find in the history of philosophy one giant owing his fame largely to another. But in
the field of philosophy of mathematics and logic, had it not been for Russell’s discovery of Frege’s
work, the latter might have remained unnoticed. The two together laid the foundations of what has
come to be called Analytical philosophy. However Frege’s project of logicizing mathematics which he
believed he was on the verge of accomplishing failed to take the last step due to a major flaw that
Russell discovered, the famous “Russell’s paradox”. But in the quest of this aim, Frege had advanced
philosophy of mathematics tremendously.

Frege was interested in the fundamental laws of arithmetic, especially the concept of a number. He
had been engaged with this question for a long time and in his Foundations of arithmetic, after
undertaking a historical survey of the various ways in which the concept of number had been
conceived, proposed a new approach towards understanding the number . He rejected the prevalent
theories regarding number, especially the empiricist conception which subscribed to the view that it
is an abstraction from our experience of objects. Nor did he have much sympathy with the
psychologistic school which attempted to explain arithmetic in terms of mental processes of
thinking. As he put it, “Number is not anything physical but nor is it anything subjective or an idea.”
(1960:58). Instead, he tied the notion of a number to an “assertion about a concept”. To quote
Frege,

‘ If I say “Venus has 0 moons”, there simply does not exist any moon or agglomeration of moons for
anything to be asserted of; but what happens is that a property is assigned to the concept “moon of
venus”, namely that of including nothing under it.’ (1960:51)

It might appear from the above that a number is a property of a concept but it would be incorrect to
maintain this view. Rather Frege believed that a number was an object, not a
property(Passmore:149). Whether an object or a property, one cannot help suspecting that certain
nagging questions remain unanswered in the background. To begin with, in the above example, the
concept of a number occurs in an expression in a sentence in ordinary language. However numbers
are obviously the fundamental entities in mathematics and if so, how are we to make sense of
numbers occurring in mathematical expressions. There does not seem to be a ready translatability of
the above concept of number to a mathematical expression. But actually Frege had no hesitation in
treating number in a mathematical expression in almost the same way as he does in ordinary
language. This is what he accomplishes in his seminal paper Functions and Concepts. In it, Frege
demonstrates by the help of several illustrative functions how the number -1 and 1 could be defined.
Further, he shows how this method of defining a number has its counterpart in logic as the “identity
of the extension of concepts.”. By doing this, Frege wanted to show the analogous nature of a
mathematical object like a number and a logical object like the extension of a concept. However, this
mode of conceptualising numbers still remains far from convincing. Apart from the fact that it is
highly counterintuitive, it also seems to make numbers somewhat less substantial than they really
are in the world of mathematics. Frege’s approach suggests that functions have a logical priority
over numbers whereas the working hypothesis of a mathematician is just the reverse.

This attempt to pin down the meaning of a number in terms of the context in which it occurs,
whether in a linguistic expression or a mathematical one, has been described as the “linguistic turn”
which was to influence much of 20 th century philosophy of language. Frege’s later distinction
between sense and reference perhaps further strengthened his view that ordinary language and
mathematics can be made to mirror each other with the aid of logic. Infact the Fregean
commandment that meaning is always dependent on the context is an extension of the idea of the
concept of number to the entire arena of language. It would not be wholly incorrect to add that
perhaps this unnatural intermingling of the world of mathematics with the world of language
created more troubles than it solved, atleast in the philosophy of mathematics. The maxim he
developed, “never to ask for meaning of a word in isolation, but only in the context of a
proposition”, becomes a wholly unreliable guide when we introduce it to the world of numbers.
What counts for a context in the case of a number is the function of which it is a part. Herein lies the
chief source of confusion since a number is an independent object in its own right, something which
Frege himself acknowledges. However the symbolism of algebra where a variable stands for a
number becomes a more important puzzle to solve for Frege than the existence of the number itself.
Because he gave so much importance to the form, he inadvertently mistook the woods for the trees.
Something deeper in the nature of mathematics may have led to this oversight. Mathematics is both
discovered and invented. While we freely choose the operations to which we can subject a number,
the ensuing properties are discovered. If it were not so, the vast and exciting field of number theory,
with theorems and conjectures like Fermat’s last theorem which remained unsolved for centuries,
would be reduced to a trifle. So if we are to assess Frege’s theory of a number, we have to admit
that it illuminates some parts of algebra more than the metaphysics of a number. Algebra is to
arithmetic what equations are to physical world. An algebraic statement can assume the form of an
equation ,an inequality , a theorem as well as several other forms. Frege’s method of reducing these
various forms to the two truth values of True and False, and then redefining everything backwards
does not appear either to be justifiable or pragmatic. In the end, we remain as mystified about the
significance of a number as before.

Does Russell have a different view on this? At first glance it appears like Russell too follows a similar
line, especially in his treatment of functions, where his account is expressed in a different
terminology though the central idea is the same. Russell comes up with the term “Propositional
function” to show how logic underlies both language and mathematics. A propositional function is
an expression containing an undetermined constituent from which we can obtain a proposition by
substituting a determinate value for the undetermined constituent. (Stebbing:130) So “x is a man” is
a propositional function which becomes a proposition when we substitute a proper name, say
Newton,for x. According to Russell, a mathematical function is similar to a propositional function in
that it too has variable but differs from it in that it is descriptive in nature which the latter is not.
Thus a function like Log x is a mathematical function which for a definite value of x gives us a number
which is described by the function “logarithm of”. Thus in Log 100 is 2, the “logarithm of 100”
describes the number 2. The number described has to be distinguished from the description of it.
The same number can be described in various ways , potentially infinite ways.

However we find that there is an important and crucial difference between Frege’s and Russell’s
account of a function. Russell says that a propositional function, which is of the form “x is man”
becomes a proposition when “x” is substituted for it. It does not describe anything . On the other
hand , a mathematical function like log x becomes a description of an entity when a substitution is
made for x. Thus there is nothing analogous to a description in a propositional function.
(Stebbing:132) On the other hand, Frege is so keen to show the equivalence of language and
mathematics that he tries to read a linguistic sense into a mathematical expression which leads him
to strange ways of using the “equal to “=” sign. Instead of the distinction that Russell rightly makes
between a propositional function and a mathematical function, Frege claims that even a
mathematical function is actually a description of a proposition, not merely the description of an
entity. Thus log 100=2 is a description of the proposition which is True. So we can see that what
Frege is proposing is that a mathematical proposition only describes the logical objects, True or
False, not any entities. That is the reason why he claims that when we say 2^4 = 4^2 we are also
simply describing the truth value of the sense of both sides of the equation. Hence ,according to
Frege, it is possible to write that (2^4=4^2) =(3>2). (Geach & Black:21). The underlying assumption is
that it is the sense of an expression whose identity is being expressed.

Thus both Frege and Russell arrive at the notion of a function through logical concepts. However
Russell parts ways from Frege when it comes to deriving a conception of a number from function.
Frege’s antagonism to the empiricist conception of number forced him to seek a definition of
number through concepts of logic alone. Russell was temperamentally an empiricist and he did not
feel similar qualms about empiricism in mathematics. Infact, he wanted any conception of number,
however logicist in nature, to resonate with our commonsensical empiricist view of it. He says,

“We want our numbers not merely to verify mathematical formulae,but to apply in the right way to
common objects. We have already some knowledge of what we mean by “1” and “2” and so on, and
our use of numbers in arithmetic must conform to this knowledge.” (Russell: 194).

Frege’s account,on the other hand, is deficient in this respect because it is hard to square with our
ordinary use of numbers. Russell, in any case, was far too inspired by Mill, his godfather, to overlook
the farsightedness that Mill had shown in his empirical view of mathematics which had been
brushed aside by Frege in a cavalier manner even though he may not have subscribed to the
psychologism that formed a part of it. Though posterity has largely sided with Frege, many have
belatedly come to appreciate the depths of Mill’s philosophy of mathematics. (Scruton:230).
Russell’s definition of a number as a “class of all classes similar to the given class” is therefore rooted
in an empirical understanding of number because a class of a given number is ostensively sought
( though its logicist variant need not rely on the ostensive property). Thus Russell’s conception of a
number was able to answer both to our empirical intuition as well as to its apriori analytic nature.
This also shows that Russell avoided the Fregean trap of privileging the function over a number
simply because a function is more amenable to logical analysis than a number. Russell’s account of a
number reasserts the independence of a number as well as restoring it its lost dignity and mystery.

Thus there is an uneasy fit between the concepts of number that Frege arrived at and the
commonsensical view of numbers as independent entities. It is not clear whether one can account
for the immense depths of theorems in number theory through a Fregean view of number. How are
we to account for the prime numbers in this view? It seems more of a classificatory tool than a
substantial advance in our understanding of numbers. How does the information that a number may
be described in various ways help us in uncovering any deep truth about relationship between
numbers? The fact that numbers acquire properties which remain hidden from us is still
unanswered.

The Fregean theory of number because of its reliance on functions is also open to challenges on
several other fronts. For example how are we to account for Transcendental numbers, i.e., numbers
which are not expressible in the form of a polynomial function. Such a number should not exist if
Frege is to be strictly followed. But Transcendental number theory is a well respected branch of
mathematics. Similarly there is the concept of algebraic independence of number, similar to a
transcendental number, which also poses important questions regarding Frege’s view. Ultimately
the Fregean notion of number does not do justice either to our empirical understanding of number
nor to the immense richness of properties that numbers display. Russell, on the other hand, viewed
mathematics more in terms of science and was fully cognisant of the elusiveness of its ultimate
origins.
References:

Frege,Gottlob.(Tr.J.L.Austin) 1960. The foundations of Arithmetic HarperTorch Books NewYork.

Frege,Gottlob (Ed. Peter Geach and Max Black) 1960 Translations from the Philosophical Writings of
Gottlob Frege Basil Blackwell Oxford

Passmore,John.1966 A hundred Years in Philosophy Penguin Books

Russell,Bertrand 1919 Introduction to mathematical philosophy George Allen and Unwin

Scruton,Roger 1995 A short history of modern philosophy Routledge

Stebbing,L.Susan 1933 A modern introduction to logic Methuen &Co. London