Collingwood and Sellars

This impression of a difference between the ideals
of a scientific vocabulary and a philosophical is only
deepened by observing that many of the greatest
philosophers, especially those who by common consent have written well in addition to thinking well,
have used nothing that can be called a technical
vocabulary. Berkeley has none ; Plato none, if consistency of usage is a test; Descartes none, except
when he uses a technical term to point a reference
to the thoughts of others ; and where a great philosopher like Kant seems to revel in them, it is by no
means agreed that his thought gains proportionately
in precision and intelligibility, or that the stylist in
him is equal to the philosopher.

A general review of the history of philosophy
compared with the equally long history of mathematics, would show that whereas exact science has
from the first been at pains to build up a technical
vocabulary in which every term should have a rigid
and constant meaning, philosophy has always taken
a different road: its terms have shifted their meaning
from one writer to another, and in successive phases
of the same writer’s work, in a way which is the
exact opposite of what we find in science, and would
justify the assertion that, in the strict sense of the
word technical, philosophy has never had anything
that deserved the name of a technical vocabulary.
Before concluding that this is a state of things
calling for amendment, it may be well to ask what
technical terms are, and why they are needed in the
expression of scientific thought.

Technical terms are terms not used in ordinary
speech, but invented ad hoc for a special purpose,
or else they are borrowed from ordinary speech but
used ad hoc in a special sense. They are needed
because it is desired to express a thought for whose
expression ordinary speech does not provide. Hence,
because they are essentially innovations in vocabulary,
and artificial or arbitrary innovations, they cannot
be understood and therefore must not be used unless
they are defined: and definition, here, means ‘verbal’
as distinct from ‘real’ definition.

It has sometimes been maintained that all language consists of sounds taken at pleasure to serve as
marks for certain thoughts or things : which would
amount to saying that it consists of technical terms.
But since a technical term implies a definition, it is

impossible that all words should be technical terms,
for if they were we could never understand their definitions. The business of language is to express or
explain; if language cannot explain itself, nothing
else can explain it; and a technical term, in so far
as it calls for explanation, is to that extent not language but something else which resembles language
in being significant, but differs from it in not being
expressive or self-explanatory. Perhaps I may point
the distinction by saying that it is properly not a
word but a symbol, using this term as when we speak
of mathematical symbols. The technical vocabulary
of science is thus neither a language nor a special
part of language, but a symbolism like that of mathematics. It presupposes language, for the terms of
which it consists are intelligible only when defined,
and they must be defined in ordinary or non-technical
language, that is, in language proper. But language
proper does not presuppose technical terms, for in
poetry, where language is most perfectly and purely
itself, no technical terms are either used or presup-
posed, any more than in the primitive speech of
childhood or the ordinary speech of conversation.

Thus the technical element in scientific language
is an element foreign to the essence of language as
such. So far as scientific literature allows itself to
be guided by its natural tendency to rely on technical
terms, scientific prose falls apart into two things:
expressions, as a mathematician speaks of expressions, made up of technical terms, which signify
scientific thought but are not language, and the
verbal definitions of these terms, which are language
but do not signify scientific thought.

Philosophical literature shows no such tendency.
Even when, owing to the mistaken idea that whatever is good in science will prove good in philosophy,
it has tried to imitate science in this respect, the
imitation has been slight and superficial, and the
further it has gone the less good it has done. This
is because the peculiar necessity for a technical
vocabulary in science has no counterpart in philosophy.
Technical terms are needed in science because in
the course of scientific thought we encounter concepts which are wholly new to us, and for which
therefore we must have wholly new names. Such
words as chiliagon and pterodactyl are additions to
our vocabulary because the things for which they
stand are additions to our experience. This is possible because the concepts of science are divided
into mutually exclusive species, and consequently
there can be specifications of a familiar genus which
are altogether new to us.

In philosophy, where the species of a genus are
not mutually exclusive, no concept can ever come
to us as an absolute novelty; we can only come to
know better what to some extent we knew already.
We therefore never need an absolutely new word for
an absolutely new thing. But we do constantly need
relatively new words for relatively new things: words
with which to indicate the new aspects, new distinctions, new connexions which thought brings to light

in a familiar subject-matter; and even these are not so
much new to us as hitherto imperfectly apprehended.

This demand cannot be satisfied by technical
terms. On the contrary, technical terms, owing to
their rigidity and artificiality, are a positive impediment to its satisfaction. In order to satisfy it, a
vocabulary needs two things: groups of words nearly
but not quite synonymous, differentiated by shades
of meaning which for some purposes can be ignored
and for others become important; and single words
which, without being definitely equivocal, have
various senses distinguished according to the ways
in which they are used.

These two characteristics are precisely those
which ordinary language, as distinct from a technical
vocabulary, possesses. It is easy to verify this statement by comparing the scientific definition of such
a word as circle with the account given for example
in the Oxford English Dictionary of what the same
word means or may mean in ordinary usage. If it
is argued, according to the method followed elsewhere in this essay, that since technical terms are
used in science something corresponding to them,
mutatis mutandis, will be found in philosophy, the
modifications necessary to change the concept of a
technical term from the shape appropriate to science
into the shape appropriate to philosophy will deprive
it exactly of what makes it a technical term and
convert it into ordinary speech.

The language of philosophy is therefore, as every
careful reader of the great philosophers already
knows, a literary language and not a technical.
Wherever a philosopher uses a term requiring formal
definition, as distinct from the kind of exposition
described in the fourth chapter, the intrusion of a
non-literary element into his language corresponds
with the intrusion of a non-philosophical element
into his thought: a fragment of science, a piece of
inchoate philosophizing, or a philosophical error;
three things not, in such a case, easily to be dis-
tinguished.

The duty of the philosopher as a writer is therefore to avoid the technical vocabulary proper to
science, and to choose his words according to the
rules of literature. His terminology must have that
expressiveness, that flexibility, that dependence upon
context, which are the hall-marks of a literary use
of words as opposed to a technical use of symbols.

A corresponding duty rests with the reader of
philosophical literature, who must remember that
he is reading a language and not a symbolism. He
must neither think that his author is offering a verbal
definition when he is making some statement about
the essence of a concept—a fertile source of sophistical criticisms—nor complain when nothing resembling such a definition is given; he must expect
philosophical terms to express their own meaning
by the way in which they are used, like the words
of ordinary speech. He must not expect one word
always to mean one thing in the sense that its meaning undergoes no kind of change; he must expect
philosophical terminology, like all language, to be
always in process of development, and he must
recollect that this, so far from making it harder to
understand, is what makes it able to express its own
meaning instead of being incomprehensible apart
from definitions, like a collection of rigid and therefore artificial technical terms.

R.G. Collingwood, “Philosophy as a Branch of Literature” (1933)

The habits of any formal scientist, like those of the mathematician in particular are tautology-habits. We can urge their adoption;
we can point to the practical consequences of not adopting them. The same is
true of justification. Thus, a “justification of induction” is either a tautology
in pragmatics; or else it is a recommendation of a set of tautology-habits for
“law,” “confirmed-to-degree-n,” “evidence,” etc.

“Are you not saying that, after all, the pragmatist has the last word?”, I
shall be asked. In a sense this is true. But the pragmatist must take the bitter
along with the sweet; for the “last word” is not a philosophical proposition.
Philosophy is pure formalism; pure theory of language. The recommendation of
formalisms for their utility is not philosophy. Hume’s scepticism was a consequence of his mistake in supposing that the philosophical questions he asked
in the study were sweeping questions of fact, and that therefore outside the study
he took an unquestioning attitude towards factual propositions questioned in
the study. The truth of the matter, and I speak in the tradition of Hume, is
very opposite. There are no factual statements which become philosophical
in the study (though there are non-factual statements which are philosophical
outside the study); and in philosophy, scepticism is a self-contradictory position.

Wilfrid Sellars, “Pure Pragmatics and Epistemology” (1947)

Though seemingly in opposition, aren’t the two of them effectively making the same point about technical discourse in philosophy?

3 thoughts on “Collingwood and Sellars

  1. Collingwood agrees, he just doesn’t like it!

    Seriously, I suspect a Bourbonian analysis could resolve how tautologies of formalism can add up to “literature,” no?

  2. It seems Wittgenstein resolved much of this though. He would contend the point that “Philosophy is pure formalism; pure theory of language.” For Wittgenstein, philosophy is not theoretical, but therapeutic, an activity. The distinction between the boundaries of science and philosophy is further complicated with his late work. Wittgenstein eventually denies the “essence” of language as Collingwood appeals to, and instead, posits science as a language-game with, perhaps, more inflexible rules than ordinary language. Collingwood’s claim that, “The technical vocabulary of science is thus neither a language nor a special part of language” is opposite of Wittgenstein’s view in both early and late. In his early work, the tautological propositions of language – that is, the language of natural science – was the only meaningful language. In his late work, he proposed that science is but a particular variety of language. Although, they abide by different rules, they are both still practices of language. It is true philosophy generally has more malleability within the language; that what Kant called “categories” and “a priori synthetic propositions” might be brought to more contemporary psychological terminologies. But no less, Kant’s philosophy was of a particularly formal and technical kind that I do not think we can so easily dislodge as philosophy “proper” – instead, we must allow philosophy and language more plurality of forms. Collingwood notes – contra scientific formalism – that, “he must expect philosophical terms to express their own meaning by the way in which they are used,” but mathematics is carried out in much of the same way. It is a practice of use.

Leave a Reply