On The Linguistic Turn in Philosophy

SOREN STENLUND
ON THE LINGUISTIC TURN IN PHILOSOPHY
Introduction
During the twentieth century, language enters the foreground of philosophy in an un-
precedented way. One speaks of a 'linguistic turn' that is to have taken place around the
turn of the century and which meant not only a greater interest in language as a subject
matter or one field of inquiry among others, but more importantly, a new way of dealing
with philosophical problems generally, even those which are not directly concerned with
language as such. 1 The concentration on language is commonly depicted in historical ac-
counts as characteristic of the kind of analytic philosophy that was predominant in the
United States, England and the Nordic countries during the twentieth century, but it is
also described as a distinctive feature of structuralism and hermeneutics, even if the role
of language in these traditions is somewhat different.
Before I enter into the question of the significance of this linguistic change of course,
I would like to note some difficulties attending the answer to that question as it is pre-
sented in standard philosophical historiography.
There is an inclination to portray this change as if it were internal to philosophy, a
development arising within the tradition itself, more or less independently of circum-
stances or factors external to the philosophical discussion. The new orientation in phi-
losophy is seen as having been initiated by the ideas propounded by Frege, Saussure and
others and by the influence that these ideas came to have. Thus Frege and Saussure have
become the intellectual heroes of their respective traditions. They are seen as having
made discoveries or having articulated new ideas and ways of thinking that are taken to
be central to the philosophy following in their wake. The question of how they came
upon their new ideas, the circumstances in and backdrop against which the ideas were
developed and how Frege and Saussure might have experienced them, are usually re-
garded as of secondary importance within the internal historiography of their respective
traditions, as if it were pointless to try to understand a 'hero'. Anders Wedberg intro-
duces his historical outline of Frege's philosophy by saying that Frege, in contrast to
Kant, wished "to prove conclusively that the propositions of arithmetic are 'analytic a
priori'" and that his thoughts about the philosophy of language were a part of that en-
Rorty. R. (ed.). The Linguistic Tum. Recent Essays in Philosophical Method. Chicago and London: The
University of Chicago Press. 1967.
11
M. Gustaf\'son and L Hertzberg (eds.), The Practice of Language, 11-50.
© 2002 Kluwer Academic Publi~·hers.
12 SOREN STENLUND
deavour. 2 But why would anyone make that his life's work, as one could reasonably say
that Frege did?
Is the linguistic turn in philosophy merely a new orientation or trend which happened
to become the fashion at philosophy departments during the twentieth century? It can ap-
pear that way if one looks at many current presentations which take as their starting point
a collection of methodological doctrines and slogans formulated from the perspective of
the current attitudes, rules and norms of various schools of contemporary academic phi-
losophy. This kind of historiography, however, tends to distort the conditions under
which the doctrines discussed were originally formulated, that is, it loses sight of the
problems which were crucial for the work of Saussure, Frege and others, and which lead
to the new interest in language within philosophy.
This kind of historiography is also anachronistic in a number of respects. A recent
work by Alberto Coffa is entitled "The Semantic Tradition from Kant to Carnap", which
gives the impression that the way of approaching language that has come to be desig-
nated 'semantics' and the sorts of problems taken up in that approach existed already be-
fore the latter half of the nineteenth century, when the term first came into use. 3 Coffa
speaks, for example, of "Kant's tacit semantic premises" and of the decline of semantics
since Descartes. But the questions concerning the meaning of statements and expressions
which we call 'semantic', I would argue, were just the sorts of questions that one began
to pose when the linguistic turn is supposed to have take place. Of course, philosophers
had been interested in the meaning of words and statements earlier as well. but, as I wish
to show in this essay. this interest was not in the same spirit as in our century. Similarly,
the questions were not formulated against the backdrop of the specific view of language
that is essential to twentieth-century semantic theories and discussion of meaning. 4
The terms 'theory of knowledge' and 'epistemology' are two further examples of
terms that are generally used in an almost anachronistic manner in philosophical histori-
ography. The corresponding terms in German and English apparently began to be used
around the middle of the nineteenth century as the various so-called special sciences
were detached from philosophy and came to be viewed as independent fields of knowl-
edge. 5 Previously. the subject matter of many of the questions thought to belong to the
theory of knowledge was not identifiable in this way. John Locke, for instance, speaks of
"human understanding" where we would say "human knowledge". When the physicist
was not a specialist in a delimited field, but was rather a natural philosopher, his ques-
2 Wedberg. A .• A History of Philosophy. Volume 3: From Bolzam) to Wittgenstein, Oxford: Clarendon

Press, 1984, p. 89.
3 Coffa, 1. A., The Semantic Tradition from Kant to Carnap. edited by L. Wessels. Cambridge University
Press. 1991.
4 One can make a similar point about the term 'linguistics' which is introduced in connection with Saussure
and which is often associated with a systematic and structuralist approach to the study of language. I will
follow the English and French tenninology, however, and use the term 'linguistics' in the broader sense
which includes even pre-Saussurean philology and studies of language.
5 See Hacking. I., Why Does LanRuage Maller to Philosophy~. Cambridge University Press, 1975, p. 166.
ON THE LINGUISTIC TURN IN PHILOSOPHY I3
tions also concerned the constitution of ideas, representations and thoughts in human
reason or the understanding. Yet in contemporary accounts of the history of philosophy,
it is not uncommon to find surveys of the history of the theory of knowledge from antiq-
uity to the present.
I do not wish to condemn such anachronistic or more liberal uses of modern aca-
demic philosophical terminology under all circumstances. On the contrary, it can often
be illuminating and clarifying to bring some aspect of an older philosophical work into
sharper relief by the use of such tools as long as one is aware that this is what is being
done. The problem is rather that such aspects can easily be transformed into incontro-
vertible historical facts within the internal philosophical historiography through the atti-
tude one has to one's own terminological apparatus. Such attitudes or predispositions are
constitutive elements of the identity of a discipline or theoretical orientation. In such an
attitude, one is inclined, for example, to treat semantic or epistemological questions as
we understand them today as eternal questions which must have been posed in the same
way and in the same sense by philosophers throughout the ages, as if the questions must
have concerned something which was essentially the same throughout the ages.
My aim here is not to provide an historical account as an answer to the question of
how it really was. My main interest in this essay is rather certain problematic features of
our notions and ideas about how it really was. 6 These ideas and notions have been
formed largely within the internal historiography that has delineated the tradition. It is
easy for us to forget that 'the philosophical tradition', as much as the boundaries be-
tween the various traditions and schools within philosophy, is not something given a pri-
ori, nor something that can be found directly in the historical material. We ought to
remind ourselves, more often than we do, that the internal historiography, the discipli-
nary historiography, the various sciences' 'own' historiography, is a specific genre the
purpose of which is largely to confirm, define and support the discipline and to satisfy
ideological needs within a certain tradition or orientation. This aim can be more or less
explicit or dominant in different internal historical works. It is perhaps most explicit in
those cases in which the point is to launch and legitimate a new orientation. In the case
of linguistics, Noam Chomsky's Cartesian Linguistics is a clear example of this. 7
There is a pervasive tendency in the internal historiography, within philosophy as
well as within other disciplines, to represent an important forerunner within a field
through a canonized version of his contributions. The origin of his discoveries and in-
sights is elucidated in terms of 'rational motives' and intellectual standpoints that are
6 This essay to a great extent builds upon certain historical facl~ (and in a number of cases upon an interpre·
tation of them) found in Hans Aarsleffs book From Locke to Saussure (London: Athlone Press Ltd.,
1982).
7 It was crucial for Chom~ky, for instance, to meet Locke's critique of the notion of 'innate ideas'. Regard-
ing Chomsky'S reply, Aarsleff writes: "Chomsky'S version of Locke's philosophy - in so far as it is possi-
ble (and fair) to judge on the slender evidence we are offered - is plainly false. The consequence of that
error is historical chaos."(p. 176) 11 is also doubtful that Chomsky's Carlesianism finds support in Des-
cartes' writing.
14 SOREN STENLUND
supposed to have driven the research forward, as if the thinker saw himself as an exem-
plary professor of the discipline in question. 8 One characteristic feature of such a canon-
ized historiography is precisely that one tries to excise or disguise all external influences,
all influences from other competing disciplines or factors of a 'non-scientific' nature that
might have played a role in the context. One might wonder, however, if it is not the case
that most great innovations and achievements, not only in philosophy but also in other
fields, arise largely as the result of external influences. Decisive influences and ideas
have often come from external sources. The historian of linguistics, Hans Aarsletf, ar-
gues convincingly, for instance, that Saussure's new ideas were decisively influenced,
not by the German philological tradition, but by intellectual ideas prevalent in Paris at
the time, especially those of the historian and literary critic Hippolyte Taine, who was
not a philologist or linguist, but a professor of art history.
Returning to the significance of the linguistic turn in philosophy, it is not inessential
to consider where one draws the line between philosophy and general linguistics or theo-
retical linguistics. One can question if there can be a history of the philosophy of lan-
guage that is not at the same time, at least in part, a history of linguistics, and vice-versa.
According to Aarsleff, the ideas of the philosopher John Locke, for instance, came to
have an enormous influence on linguistics during the eighteenth and nineteenth centu-
ries, even until Saussure's day. And the influence came via Condillac, among others. The
idea of the arbitrariness of the sign is central in Locke's thought (who, however, was not
the first or the only one to express the idea).9 Was Saussure a linguist or a philosopher?
He had a background in philology, but when one looks at the questions that he wanted to
answer with his innovative ideas Cia langue', the concept of structure, the synchronic as-
pect of language, etc.), one is inclined to say that Saussure was no less of a philosopher
of language than Frege - who, by the way, was a professor of mathematics. And both
Saussure and Frege were philosophers of language to a far greater degree than many
contemporaneous professors of philosophy were.
One need not go so far back in time to see how the linguistic and philosophical tradi-
tions merge, and it is not always evident which current leads into philosophy and which
leads into linguistics, as we define the disciplines today. The departmental boundaries of
our day did not exist in earlier centuries, even if one can see tendencies in that direction
8 Aarsleff calls this form of historical writing "the internal history of an institutionalized professorial craft."
(Aarsleff. p. 7)
9 Aarsleff writes: "Mersenne and Locke both (without any direct connection) made the arbitrariness of the
linguistic sign a cardinal principle, a principle that is best known from Locke's Essay. which was the chief
source of eighteenth-century linguistic philosophy. It is therefore not surprising that there are evident simi-
larities between Locke and the eighteenth century on the one hand and what we read in Saussure on the
other." (Aarsleff. p. 317) That Saussure should find certain features of the view of language stemming
from Locke mistaken. although sufficiently important to question and correct, ought to be included in the
influence described here by Aarsleff. One such feature is, for example, Locke's view that language ulti-
mately consist~ of the free volitional acts of individual speakers, and not in the fact that they apply or real-
ize a sort of mental system that is socially determined and which stands outside of the control of the
individual speaker.
ON THE LINGUISTIC TURN IN PHILOSOPHY 15
during the nineteenth century; German historical comparative linguistics, for example,
like many other disciplines during the same period, aspired to define itself as an inde-
pendent field of study. Perhaps it is fair to say that the theorizing about language that is
connected to and motivated by the demands of language instruction, translation and the
study of specific languages (where traditional grammar has been the most important
tool), belongs to the linguistic tradition. But there are also numerous examples of theo-
rizing of this kind in which the instruments of linguistics and the introduction of linguis-
tic norms are justified by al1uding to philosophical or metaphysical ideas - even in our
own century.
The Reaction against Psychologism
Questions having to do with the meaning of expressions or statements which we call se-
mantic, as I said earlier, were just the sorts of problems that had begun to be posed dur-
ing the period when the linguistic turn is said to have taken place. Natural1y, questions
concerning the meaning of words and sentences had been posed earlier. but hardly in the
same spirit or with the same purpose that characterizes semantic theories or theories of
meaning in our sense. The issue is complicated by the fact that there occurred a consider-
able shift in the sense of these questions within philosophy and linguistic theory during
the twentieth century. Frege, for instance, was not primarily concerned with providing
answers to general questions as to how expressions and statements in a language have
meaning in the way that contemporary linguistic theorists wish to do. In this respect,
Frege's philosophy of language was not a 'pure philosophy of language'. It can even be
misleading to speak of 'Frege's semantics'. Given the established use of the term 'se-
mantics' in analytic philosophy since Carnap and Tarski, Frege's critical study of lan-
guage is not 'semantic'. Frege explicitly distanced himself from the metamathematical
view of language that is at the core of 'semantics' in this sense.
As Talbot Taylor points out, Frege shows no great interest, at least compared to mod-
ern theory of language, in questions such as, "In what does the objective meaning of a
statement consist?" or "What does it mean to understand the sense of a statementT IO
Frege's metaphorical responses to these questions do not constitute much of a 'theory'.
In that respect, Frege was not what one today would call a 'philosopher of language'.
Yet philosophy's internal historiography (especially Michael Dummett's account) por-
trays him as a meaning theorist and philosopher of language in precisely this sense.
Frege's studies in the philosophy of language, however, were not his principal aim, but
were rather part of the process of achieving clarity in other conceptual issues, in particu-
lar, issues having to do with the nature of logic and mathematics. Frege writes:
The pictorial character of language presents a difficulty (i.e. in his logical investigations).
The sensible always breaks in and makes the expression pictorial and so improper. So one
10 Taylor. 1. T .• Mutual Misunderstanding: Scepticism and the Theorizing of Language and Interpretation.
Durham and London: Duke University Press. 1992. p. 107.
16 SOREN STENLUND
fights against language. and I am compelled to occupy myself with language although it is
not my proper concern here. I I
In one sense, one might say that the language, about which one has posed semantic
questions and created philosophical theories of meaning during the twentieth century,
did not exist as a possible object of study for European philosophy and linguistic theory
before the second half of the nineteenth century. It was first then that a concept of lan-
guage was established in which language is thought to contain propositions of the kind
that already Bernard Bolzano called Siitze-an-sich, by which he meant something other
than expressed or thought statements. The kind of statements which Frege took to be
bearers of begrijj1iche lnhafte or expressions of Gedanken are of a similar kind, as are
the statements which express 'propositions' in the analytic tradition. As I wish to show
in this paper, it is no coincidence that these ideas emerge at the same time as the notion
of intentionality in act psychology and phenomenology. The thought behind the notion
of intentionality is that mental acts are directed toward an object, in a manner similar to
the way that semantic 'meanings' have a kind of objective, 'intersubjective' character.
No equivalent to what Saussure calls fa langue is to be found in theorizing about lan-
guage from the seventeenth and eighteenth centuries either. The philosopher John Locke
has a great deal to say about how "words signify ideas", but what he is treating is not
what we call semantic questions or problems in the theory of meaning (a point to which I
shall return). From the perspective of our own epoch, one might say that Locke writes
about language as if it were merely a practical instrument for communication, without
any underlying 'langue' that is 'language itself, or any 'objective meaning' in the words
or statements that guarantees that we can all speak about the same things and mean the
same things with the same words. A number of traits in Locke's philosophy of language
are captured in the following passage (from Bk. III, Ch. II, in An Essay Concerning Hu-
man Understanding, from 1690):
The comfon and advantage of society not being to be had without communication of
thoughts. it was necessary that man should find out some external sensible signs whereby
those invisible ideas. which his thought~ are made up of, might be known to others. For this
purpose. nothing was so fit, either for plenty or quickness, as those aniculate sounds which
with so much ease and variety he found himself able to make. Thus we may conceive how
words, which were by nature so well adapted to that purpose. came to be made use of by men
as the signs of their ideas: not by any natural connection that there is between panicular ar-
ticulate sounds and cenain ideas, for then there would be but one language amongst all men;
but b~ a voluntary imposition whereby such a word is made arbitrarily the mark of such an
idea. 2
According to Locke, "words signify ideas" and the ideas signified by words are what
Frege would call subjective representations which we associate with words in our lan-
II Frege G.. "Thoughl~". in McGuinness, B. (ed.) Collected Papers on Mmhemmics. Logic and Philosophy,
Oxford: Basil Blackwell, 1984, p. 360 (footnote).
12 Locke, J.• An Essay Concerning Human Understanding, An Abridgement. edited by John W. Yolton, Lon-
don and Melbourne: Dent & Sons Ltd. Everymans Library. 1961, p. 207.
ON THE LINGUISTIC TURN IN PHILOSOPHY I7
guage, everyone in his or her way; Frege takes great care to distinguish these from the
non-subjective content of words, that is, their sense or Sinn. Frege argues forcefully and
indignantly that the meaning of words is not the subjective ideas that we associate with
them. Why? The standard explanation that Frege was an anti-psychologistic philosopher
and the lexical description of what that standpoint implies usually offered in historical
surveys clarifies little. What did this antipsychologism mean for Frege, Bolzano, Husserl
and others who display similar reactions around the same period?
For Locke, Berkeley and many other philosophers during the seventeenth and eight-
eenth centuries, this was simply not a problem. Locke does not hesitate in the face of the
consequences of his view of language: different individuals generally do not associate
the same ideas to a word. There is nothing in what he saw as 'the nature of language'
that guarantees a priori the possibility of successful linguistic communication. One could
almost say that Locke's standpoint was that actually, or rather in principle, we do not un-
derstand each other.
Locke's understanding of language was not unique. He sums up a rather widely ac-
cepted view that had been spread through the so-called Port Royal logic and grammar,
among other things. From the perspective of our own epoch, regardless of which school
of thought one adheres to, we are inclined to regard Locke's view, taken as philosophy
of language, as an almost absurd psychologism or mentalism which leaves open the
question of how it is possible that we in practice do understand one another. For Locke
never denied that there do indeed exist numerous cases of successful linguistic commu-
nication. But explaining how it could be so was not an urgent task for Locke. 13
In my view, the decisive difficulty in understanding the significance of the linguistic
turn in philosophy is in seeing why this way of looking at language was not some absurd
psychologism for Locke and his contemporaries, but came to be seen as such by Frege
and others in the nineteenth century. How and why did the philosophical language of the
mental in the works of so many seventeenth- and eighteenth-century philosophers lose
its authority as the language of important universal truths, and instead come to be seen as
psychologistic and speculative?
The Idea of Language as the Origin of Philosophical Problems
In one respect, one might wonder if all the talk about the linguistic tum in philosophy is
not considerably exaggerated. One important idea in the standard accounts of this change
13 Hacking writes: "Frege, like all his contemporaries. saw that public communication cannot be well ex-
plained by what he called private associated ideas. Locke and his contemporaries. did not see this at all
clearly. Nor did Locke and his friends care. He undoubtedly thought that whenever we do communicate
successfully. you get the same ideas in your breast as I have in mine. But in the works of Locke this notion
is not part of a philosophical theory of communication. [ ... J the theory of public communication (of com-
mon acceptation or Sinn) is of no importance to his philosophy. Locke did not have a theory of meaning.
He did not have a theory of public discourse. He had a theory of ideas. That is a theory of mental dis-
course." (Hacking. pp. 51-52).
18 SOREN SlENLUND
in orientation is that language came to be seen as a source of philosophical problems;

language inveigles us in conceptual mistakes. The forms of language lead our thoughts
astray. Bertrand Russell describes the situation thus:
The influence of language on philosophy has. I believe. been profound and almost unrecog-
nized. If we are not to be misled by this influence. it is necessary to become conscious of it.
and to ask ourselves deliberately how far it is legitimate. The subject-predicate logic. with
the substance-attribute metaphysic. are a case in point. It is doubtful whether either would
have been invented by people speaking a non-Aryan language ... 14
Today the idea of language as the source of philosophical confusion is associated primar-
ily with Wittgenstein, but it has been a central thought in different orientations within
analytic philosophy. Furthermore, this idea is anything but specific to philosophy in our
century. As Ian Hacking points out, it is odd how often philosophers in different genera-
tions have said that the influence of language on philosophy is deep and almost entirely
unnoticed. 15 According to Aarsleff, the phrase "the cheat of words" was common among
seventeenth-century philosophers. Already Francis Bacon offers the following recom-
mendation:
Although we think we govern our words .... certain it is that words. as a Tatar's bow. do
shoot back upon the understanding of the wisest. and mightily entangle and pervert the
judgement. So it is almost necessary. in all controversies and disputations. to imitate the wis-
dom of mathematicians. in setting down in the very beginning the definition of our
terms .... 16
The similarity between Bacon's advice and the admonitions of philosophers working
with formal logic in the analytic tradition is striking.
Regarding the problematic influence of language, Locke says that "the greatest part
of the questions and controversies that perplex mankind depends on the doubtful and un-
certain use of words ... ,,17 Among philosophers after Locke who warned against the in-
fluence of language one can name Berkeley, Mill, Nietzsche and, of course, Frege. In the
introduction to his Begriffsschrift, Frege expresses the matter in this way:
If it is one of the ta~ks of philosophy to break the domination of words over the human spirit
by laying bare the misconceptions that through the use of language often almost unavoidably
arise concerning the relations between concepts and by freeing thought from that with which
only the means of expression of ordinary language. constituted as they are. saddle it. then my
ideography. further developed for these purposes, can become a useful tool for the philoso-
pher: K
But this does not mean that all philosophers who expressed such concerns meant 'the
same thing'. For most of them, as for Frege, the caution was part of the effort to launch a
14 Russell. 8.. Logic and Knowledge. (ed. R. Marsh). London: Allen and Unwin. 1956. p. 330
15 Hacking. I.. p. 70.
16 From Bacon's The Advancement oILearnin/? cited in Hacking. p. 5.
17 Locke. Essay. p. xlvi.
18 Frege. G. BegrijJsschrif/. in van Heijenoort. 1. (ed.). From Frege to GOdel: A Source Book in Mathemati-
ml Logic. 1879-1931. Cambridge: Harvard University Press. 1977. p. 7.
new way of speaking and writing in new philosophical vocabulary. Wittgenstein was the
first for whom the idea had a purpose other than introducing and justifying a new philo-
sophical terminology. For Wittgenstein, the purpose was the opposite. (That is actually
the essence of the Tractatus as well, according to its concluding remarks.) A central
theme throughout Wittgenstein's work is the view that everyday language is all right as
it stands, if we could just avoid being misled by linguistic forms when we philosophize
and theorize.
Several of the philosophers I named earlier, on the contrary, describe language as
something problematic in and of itself, a thought that has gained currency during the last
few decades (within poststructuralism, for example). For Wittgenstein, what is philo-
sophically problematic is not language as it is given in ordinary and practically function-
ing kinds of linguistic communication between human beings; the source of
philosophical problems is rather our attitude to language in certain specific aspects of it,
namely, those aspects of our use of words and expressions which we also call our way of
thinking.
A basic thought in Wittgenstein's philosophy is this: if we wish to achieve clarity in
our thinking and ideas, we lose nothing by paying attention to our way of using words
and our attitude to certain linguistic forms of expression. More accurately, we have
much to gain, since the very notion that our thoughts and ideas should exist in them-
selves in some other medium and be manifested only indirectly and incompletely in our
way of using words and expressions is a notion which rests on a misleading picture of
how language works.
It is interesting to compare Wittgenstein with Descartes on the subject of the role of
language in philosophy. In certain regards, one finds distinct parallels between the two
philosophers. They share a certain attitude that is rather uncommon among philosophers.
In the Discourse on Method, Descartes says of the philosophical establishment of his day
that it "puts us in a position to speak convincingly about everything and be admired by
those who are less learned than ourselves." Here Descartes expresses an attitude toward
the technical terminology of the philosophy of the period and what it had to offer that is
similar to Wittgenstein's attitude toward the academic philosophy of his time. Descartes'
attitude toward 'everyday language' is also in some respects similar to Wittgenstein's.
The following defiant pronouncement from Descartes bears a fair resemblance to similar
remarks by Wittgenstein:
Those with the strongest reasoning and the most skill at ordering their thoughts so as to make
them clear and intelligible are always the most persuasive, even if they speak only low
Breton and have never learned rhetoric. 19
What is common to Descartes and Wittgenstein is that their critical attitude toward the
philosophical language, terminology and rhetorical methods of their respective predeces-
sors, was not part of an attempt to launch a new conceptual apparatus (even if that is how
19 Descartes, "Discourse on the Method'", in The Philosophical Wrilin/is oj Descartes. Vo/. I. transl. J. Cot-
tingham, R. Stoothoff, and D. Murdoch. Cambridge University Press. 1985. p. 114.
20 SOREN SlENLUND
it has been portrayed by their successors and commentators). It was rather a part of their
exhortation to us to think for ourselves. We can think for ourselves on the basis of the
circumstances in which we already find ourselves, and in the language we already pos-
sess. The role that Wittgenstein assigns to 'ordinary language', the language which we
already possess and master, has a clear parallel in the role that Descartes assigns to what
he terms 'common sense'. According to Descartes, "nothing is so fairly distributed
among human beings" as common sense, if we would only learn to use it correctly. Fur-
thermore, says Descartes, those who are most likely to lose their common sense are the
educated, the philosophers:
The learned have a way of being so clever as to contrive to render themselves blind to things
that are in their own nature evident. and known by the simplest peasant. 2(J
Wittgenstein would surely have agreed with Descartes' judgment here, as well as with
another of Descartes' diagnoses:
But it is a common failing of mortals to deem the more difficult the fairer; and they often
think that they have learned nothing when they see a simple cause for a fact, while at the
same time they are lost in admiration of certain sublime and profound philosophical explana-
tions, even though these for the most part are based upon foundations which no one has ade-
quately surveyed - a mental disorder which prizes the darkness higher than the light. 21
Inner Vision
There is, however, an essential difference between Descartes' attitude toward language
and Wittgenstein's. This difference, in my view, is of the greatest importance for our
question concerning the significance of the linguistic tum. At times, Descartes, like
Locke, Berkeley, Leibniz, Kant and other philosophers of the seventeenth and eighteenth
centuries, writes as if he could step outside of language; he speaks as if language were
merely something 'external' from which one can remove oneself and then retreat into the
consciousness or soul, where one can view the ideas for which words stand, in a kind of
immediate mental seeing. Descartes describes this 'mental seeing' as something far less
metaphorical than it is for us. In his early essay, "Rules for the Direction of the Mind",
Descartes writes: "Truly we shall learn how to employ our mental intuition from com-
paring it with the way in which we employ our eyes."
In our day, the ideal of a rigorous proof in mathematics contains the thought that ac-
curacy lay in the external form, that is, in the form of the propositions which make up
the proof and the formal relations between the propositions. Descartes' ideal is almost
the opposite. What is essential is the insight and view of the whole that we can achieve
20 Descartes, Rules for the Direction of the Mind, trans!. E. S. Haldane and G. R. T. Ross, in E. Chavez-
Arvizo (ed.), Descartes: Key Philosophical Writings, Wordsworth Classics of World Literature, 1997, p.
50.
21 Descartes. Rules for the Direction of the Mind. p. 32.
in mental vision. Descartes saw proof, in Hacking's words, as "a device for getting rid of
words, enabling a man to perceive the connections between ideas steadfastly.',22
Descartes is not alone in reasoning as if one could leave language behind and return
to some pure mental vision. In A Treatise Concerning the Principles of Human Knowl-
edge (1710), Berkeley writes:
So long as I confine my thoughts to my own ideas. divested (~"words. I do not see how I can
ea~ily be mistaken. The objects I consider. I clearly and adequately know. I cannot be de-
ceived in thinking I have an idea which I have not. 23
He continues:
[... J in order to see what ideas are included in any compound idea. and what not. there is
nothing more requisite than attentive perception of what passes in my own understanding.
[... J But the attainment of all these advantages doth presuppose an entire deli verance from
the deception of words. which I dare hardly promise my self; so difficult a thing it is to dis-
solve a union so early begun. and confirmed by so long a habit as that betwixt words and
ideas. 24
The passage quoted above arises in connection with Berkeley's critique of the belief in
abstract ideas, for example, the universal idea of a triangle. The immediacy of the vision
of the ideas emphasized by both Berkeley and Descartes is manifest in the fact that
Berkeley never really tries to demonstrate or argue for the 'ontological thesis' that 'ab-
stract ideas do not exist'. His exposition is rather intended to help the reader see what he
can see for himself: we do not have abstract ideas; it is nonsense that we could have ab-
stract ideas. We are misled to the notion by language. By trying to imagine what it
would mean to have abstract ideas, Berkeley's discussion aims to make clear how we
wind up in the absurdity of an abstract idea. An idea, as Berkeley uses the term, is some-
thing toward which we have or can have a kind of direct, linguistically un mediated rela-
tionship and is therefore never abstract or general.
One characteristic feature of 'introspective investigations' as we find them in the
works of Descartes as well as those of Locke, Berkeley and Hume, is the personal pres-
ence of the authors in their accounts and expositions. This is an important part of what it
means that the inner vision of which they speak is an immediate vision. The immediacy
of the vision they describe does not consist in that the ideas are immediately given, but
rather in that the immediacy can be accomplished in thinking. These philosophers often
write in the first person, but this does not mean that they take themselves to be present-
ing something subjective or private. On the contrary, it is rather as if they had at their
disposal a means of observing their own consciousness as arbitrary cases of human un-
derstanding or reason in general, as if they could see the universal in the particular in the
22 Hacking, p. 31.
23 Berkeley, G., "A Treatise Concerning the Principles of Human Knowledge", in Jessop, T. E.,(ed.), The
Works of Georges Berkeley. Bishop of C/oyne. Vol. II. London: Thomas Nelson and Sons Ltd. 1949. p. 39.
Italics mine.
24 Berkeley, p. 39.
22 SORENSTENLUND
immediacy of mental vision. Thus it is not a question of finding universal hypotheses or

making inductive generalizations from observations of particular cases; it is for this rea-
son more than any other that this sort of investigation is not psychology (in any of the
scientific senses arising in the latter half of the nineteenth century).
Often enough, the claim to be able to see the general in the particular is stated explic-
itly. Locke says this about his Essay:
All therefore that I can say of my book is that it is a copy of my own mind. in its several
ways of operation. And all that I can say for the publishing of it. is that I think the intellectual
faculties are made and operate alike in most men. 25
And Hume writes:

Mankind is so much the same. in all times and places. that history informs us of nothing new
and particular.26
This is more than anything a confession of faith, one which Hume shared with many oth-
ers in the eighteenth century, but also one which would become increasingly difficult to
share in the secularized scientific empiricism of the latter half of the nineteenth century.
The new scientific spirit had different norms for objective truth, norms which left little
room over for 'the personal'. In this new spirit, the 'immediacy' of inner vision and the
personal style of the reasoning found in Locke, Hume and Berkeley is seen as problem-
atic. Introspection appears to be something subjective and psychological, something
which even psychologists themselves dismiss as an illegitimate method when psychol-
ogy strives to become an empirical natural science. This is quite clear in early behavior-
ist criticism of older psychology.27
It is interesting to compare Berkeley's criticism of the notion of abstract ideas with
Frege's critique of the tendency to consider (mistakenly) what he calls concepts as ob-
jects. In one important respect, their point is the same. Today we would express it in
something like the following formulation: general terms such as 'triangle', 'human be-
ing' or 'quantity' do not signify in the same way as proper names or 'singular terms' sig-
nify, even though such general terms, like proper names, can function as the subject of
statements. Berkeley and Frege are also in agreement that it is this linguistic relation that
misleads us. They part company with regard to the question of how to correct the prob-
lem. Berkeley recommends that we "free ourselves from the deceptive appearances of
language" and tum inward toward the ideas in the soul to decide the matter; Frege, on
the other hand, argues that it is precisely the psychologistic notion that words signify
ideas in the mind that is the source of this as well as many other problems. This notion,
25 Cited in Aarsleff. p. 159.

26 Cited in Aarsleff, p. 159.
27 In the article, "Philosophical Progress in Language Theory" (Mell1philosophy, Vol. I, No. I, Jan. 1970, pp.
2-19), Quine describes this change as "the externalizing of empiricism" which, for him. represent~ "pro-
gress in language theory". In Quine's perspective, however, any attempt at understanding the linguistic
tum in philosophy is pie-empted. From Locke's point of view, this "externalizing of empiricism" would
hardly have seemed like progress.
Frege believes, prevents us from seeing that the logical structure and the grammatical
structure of language are different. Berkeley views language linguistically; that is, he un-
derstands language as the traditional grammar of his day represents it. Thus there is
something in language itself which is at the root of the problem. It is as if language had
an intrinsic defect that simply cannot be repaired. For Frege, however, the problems he
describes can and must be solved within language, and this is to be done by making it
more logically perfect.
An important starting point for Frege is to insist, against tendencies towards psy-
chologism, relativism and scepticism during the period, that different thinkers can con-
nect the same thought to the same statement. He stresses that we can, with the help of
language, transmit knowledge and information from one generation to another. On these
grounds, he refines the concept of 'thought' such that the nature of thought (or its 'being'
or 'essence') consists precisely in the potentiality for different thinkers to have the same
thought. He writes:
Thoughts are independent of our thinking. A thought does not belong specially to the person
who thinks it. as an idea does to the person who has it;2K
This pronouncement is directed, of course, at the view of language that takes words to
signify ideas, which we find in the works of Locke and Berkeley, and which was still
common among philosophers and scientists in the nineteenth century.
But Frege's concept of 'thought' has another important feature. A thought is the con-
tent of a statement that is relevant from the logical point of view, that is, from the view-
point that a statement expresses something that is objectively true or false. It is worth
noting that Frege unifies the logical aspect of the content of a statement with the possi-
bility of communicating that content: if a statement is to be objectively true or false, its
content must be communicable. For Frege, a statement's potential to express something
objectively true or false is inseparable from the possibility of communicating the con-
tent. The emphasis on the social aspect of language plays a similar role in Saussure's
· k'mg. 29
thIn
Frege reasons, quite simply, on the basis of another concept of language, one that
cannot be found in the works of Berkeley and his contemporaries, or their predecessors.
The language of which Frege speaks is the language that later comes to be incorporated
in the metamathematical way of looking at philosophical problems, and about which phi-
losophers started posing semantic questions and postulating theories of meaning. It is
also the concept of language which, like Saussure's notion of la langue, explains how it
is possible for us to understand one another. For both Frege and Saussure, as distinct
from Locke and Berkeley, it is imperative, in the historical situation in which they find
themselves, to secure the possibility of linguistic communication philosophically. (More
will be said about this later.) It is not nearly so obvious, during the second half of the
28 Frege. O. Posthumous Writings. Oxford: Blackwell. 1979. p. 127.

29 This relationship between the linguistic theories of Frege and Saussure is developed in the work of Talbot
Taylor cited above.
24 SOREN SlENLUND
nineteenth century, that "mankind is so much the same, in all times and places" (as
Hume had expressed it earlier). Many were inclined rather to say the opposite and, for
that reason, ask themselves: "Do we really understand each other at all?" Locke's scepti-
cism 'in principle' towards the possibility of genuine communication had become some-
thing much more real and, therewith, much more disturbing.
The Immanence of Language
It is not the thought that language leads us into errors in our thinking that is new and de-
cisive for the linguistic turn in philosophy. What I wish to suggest is that there is some-
thing else at the heart of this change of course, something which I would describe as the
increasing sense, starting around the turn of the century, that human thinking. under-
standing and reason are immanent in language. One could also describe it as the thought
that everything that we can distinctly say, think or imagine is already articulated in lan-
guage. Thinking, reasoning, and more or less all other intellectual functions are already
linguistic, i.e., they involve the use of signs. An idea or a representation, inasmuch as it
is determinate, is already articulated in some system of signs or symbols (even if it is not
yet 'spoken out loud' or written on paper in some alphabetic system of writing). Nothing
is given to us in consciousness, the understanding or thought before language. There ex-
ists no intellectual medium more fundamental than our language, or which could, as it
were, precede it. Thus we cannot 'emancipate ourselves from language' by turning in-
wards, or by observing language from outside of language, from some viewpoint which
would not itself involve language.
This thought, as a philosophical idea, is not new at the turn of the century. It can be
found already, in all essential points, as one of the main themes in Condillac's Essai sur
l'origine des connaissances humaines. from 1746. According to Condillac, in order to
use reason, that is, to have genuine knowledge and not merely the instinctual practical
abilities such as those we share with animals, "it is necessary to have a language; for the
ideas must be classified and determined, which presupposes signs employed according to
method.,,30 But as far as I can tell, this idea in the work of Condillac and his eighteenth-
century followers is just that, a philosophical notion; as an intellectual idea, it does not
make much of a difference to philosophy as long as one still believes in a uniform hu-
man nature and, therewith, believes oneself to be in possession of a language adequate to
the task of elucidating human nature. When this faith begins to waver, and the mentalist
philosophical vocabulary loses some of its immediacy and self-evidence, the phenome-
non that I have described as linguistic immanence becomes a practical reality in an en-
tirely different way.31

31 See Gundersen, O. "Condillac: Between Locke and Herder", in Eliassen and Stene·Johansen (eds.), Fri-
helens arhundre. Ulteratur. kons/ og.filosofi i Frankrike pU 1700-lalel, Spartacus forlag. Oslo, 1997.
For Frege, language (the logically complete and perfect language) is universal; it is
the universal medium for all that can be meaningfully said or thought. This feature of
Frege's philosophy is most clearly in evidence in his critique of attempts to define the
concept of truth. In particular, he is critical of definitions of truth formulated in terms of
correspondence, in which at least one side of the correspondence is thought to be some-
thing 'extra-linguistic', i.e., something that could be given in some other way than in lan-
guage, the universal medium of truth-telling. 32
The foregoing has a concrete, practical importance for Frege's way of conducting
philosophical investigations. He is aware that he cannot step outside of language when
he performs his conceptual studies and clarifies its logical structure. He has no given
'mental discourse', as did the eighteenth-century philosophers, in which he can conduct
and present his results. He recognizes that he must at times use the language that he is in
the process of improving and elucidating in his Erliiuterungen, and even use words figu-
ratively in his explanations. But the aim of the explanations justifies the means: the point
is to lead the reader to a correct understanding of the logical structure of statements.
Frege's sense of the immanence of language is perhaps expressed most explicitly when
he writes:
For our thinking is closely bound up with language and thereby with the world of the senses
[••• J We know we can have various expressions for the same thought. The connection of a
thought with one particular sentence is not a necessary one; but that a thought of which we
are conscious is connected in our mind with some sentence or other is for us men necessary.
But that does not lie in the nature of the thought but in our own nature. 33
What Frege says here is that we human beings must simply accept that 'the thoughts
themselves' are only available to consciousness through language!
These features of Frege's philosophy, which I have described as expressions of the
experience of the immanence of language, are carried out even more consistently and
radically in Wittgenstein' s Tractatus, where the main theme is that we can only ap-
proach the borders of language from inside. Logical structure is the shared and ultimate
boundary of language, thought and the world. According to the Tractatus, the world con-
sists of existing facts that are given to us only as they are depicted in language or in
thought. And the distinction between what can be said in language and what can only be
shown that is so pivotal for the Tractatus, is a consequence or manifestation of what I
have called linguistic immanence.
This experience of the immanence of language is clearly manifest in Saussure's
work. On the relationship between language and thinking, he writes:
32 J. Van Heijenoort discusses this aspect of Frege's thought in the article, "Logic as Calculus and Logic as
Language", Synthese Vol. 17 (1967), pp. 324-330. For more on this element in Frege's critique of lan-
guage, see Gundersen, 0., "Frege, semantikens umulighet og skriftens privilegium", Norsk JilosoJisk
lidskrift, 33 (1998). See also Gundersen, Tenkningen som tekn%gi, Subjeklivilelen i /ys av aspekter ved
lenkningens leckn%gier. En JilosofihislOrisk unders¢ke/se, Filosofisk institun, A VH, Universitetet i
Trondheim, 1994, which also addresses several other problems relevant to the present essay.
33 Posthumous Writings, p. 269.
26 SOREN STENLUND
In itself. thought is like a cloud. where nothing is delimited. There are no pre-established
ideas, and nothing is distinct before the appearance of language. 34
This thought appears already where Saussure takes up the question of what the "whole
and concrete object of linguistics" is. He says that linguistics distinguishes itself from
other sciences, which "work with objects given in advance, that one gradually can ob-
serve from different points of view." Saussure suggests that language as it really is, i.e.
as la langue, is only given to us inside of language; in other words, actual language is
given to us only as speakers or users of language. 35 Only then do we take the connec-
tions between the various aspects of language (concept and sound, language and thought,
the individual and the social, and so forth) for what they are: indissoluble bonds or, in
the terminology of the Tractatus, internal relations.
There is a clear similarity between Saussure's discussion of the indissoluble bonds in
language and the discussion in the Tractatus of the internal relations that can only be
shown in language. The reason is that both discussions rest on the idea of the immanence
of thought in the totality of human language. As such, language is only accessible to us
from the inside.
But it should also be noted that the 'total language' of which Saussure speaks is the
language of linguistics, that is, the language framed in traditional grammar. Similarly,
the 'total language' described by Wittgenstein in the Tractatus is the language of early
analytic philosophy, a language framed by the notions of the new mathematical logic.
This conception is one that Wittgenstein would later describe as a fundamental mistake,
In Philosophische Bemerkungen, for example, he remarks with self-irony: "How odd
that philosophers study an ideal language and not ours!"
Ideas about immanence similar to the one sketched above could also be found during
the latter half of the nineteenth century in disciplines other than language theory (which
is why I use the term, one which has a longer history which we need not concern our-
selves with here). One example is the neo-Kantian Immanenzphilosophie, which had an
anti-metaphysical program in which one claimed that 'to be real' does not transcend 'to
be conscious'. One finds similar ideas in Husserl's phenomenology. An important reason
for the formulation of this immanence idea at the time was, I believe, the secularization
of science and other institutions that had been taking place since the eighteenth century,
34 Saussure, F. de, Cours de linguislique generale, ed Tullio de Mauro, Paris: Editions Payot. 1972. p.\55.
Aarsleff argues that this idea was not novel when Saussure wrote, even if Saussure uses it in a unique way
when he applies it to other notions in linguistic theory. According to Aarsleff, Saussure was inspired by
idea~ in the intellectual discussions taking place in Paris. Aarsleff writes: "When Saussure said that ·With·
out the aid of signs, we would be unable to distinguish two ideas in a clear and certain manner,' and that
'there are no pre-established ideas, and nothing is distinct before the appearance of language: he was fol-
lowing Taine's argument - and a~ Taine acknowledged, Condillac's. For Taine. the joining of sound and
concept formed what Saussure called 'a complex unity. physiological and mental''' (Aarsleff, p. 359).
35 This less known side of Saussure's notion of la langue, which stands in stark contrast to the received view
of la langue as an abstract system, was pointed out to me by Michael Gustavsson. See also Gustavsson's
contribution to the present collection. "Language as Sign and Use: A Study of Certain AspeCl~ of Saus-
sure's View of Language".
and which had resulted in a new ideal for objective truth and stringency in demonstra-
.
tlOn, proof and conceptuaI"lzatlOn. 36
One background to these efforts. I believe. was the sense that human reason had been
abandoned to fend for itself in the harsh reality of the march of history. It is no coinci-
dence that these ideas about immanence have their roots in Kant, who probably contrib-
uted more to the detheologizing of philosophy than any other thinker. Kant's idea of
Religion innerhalb der Grenzen der blossen Vemunft expresses an attitude that was for-
eign to philosophers such as Descartes, Leibniz, Locke and Berkeley, and this attitude
carried with it important differences for what he saw as the nature and task of philosophy
as compared to how they saw it. Locke says, for example. that "Man is by nature ra-
tional", but also, "God commands what reason does." And the express purpose of
Berkeley's Discourse on the Methods of Human Knowledge, Berkeley says himself, is to
undermine what he calls "the grounds of scepticism, atheism and animosity toward relig-
ion." And this was not merely the sort of pious pronouncement expected of a bishop. It is
clear in a number of contexts that this is the driving force behind his writing.
When I describe the idea of immanence in terms of a sense or feeling that we cannot
step outside of language, it is not because I mean to belittle it as something merely psy-
chological. For many fin de siecle thinkers, quite to the contrary, it was a most definite
and pressing reality. Thus it was not a new viewpoint internal to linguistic theory created
or discovered by some intellectual hero that happened to become a success, even if there
were a few thinkers who could articulate this experience more aptly or precisely than
others. It was a sense or experience that was thrust upon people when the standard ways
of "speaking the Truth" fell into disuse. that is. when the older mentalist vocabulary sim-
ply could not be taken seriously anymore. 37 This experience gives rise to a new aware-
ness of how one uses language in philosophy and science, and this awareness brought
with it new problems and difficulties, problems that had never been posed in the same
way before (for example. the questions treated by Frege). The questions were never
asked before for the simple reason that there was no reason to ask them; in a sense. one
could say that they could not have been asked before?8
Against the background of what we can learn from Wittgenstein, it is clear that the
feeling of linguistic immanence is a metaphysical experience. The philosophical thesis
36 Frege refers to these new ideals in the introduction to his Grundlagen der Arithmetik.
37 Perhaps this experience is what the poet Stefan George is after in "The Word". It ends with the verse. "So
I renounced and sadly see: I Where the word breaks off no thing may be." (The quotation is taken from
Heidegger's essay. 'The Nature of Language". in Heidegger. M .• On Ihe Way to Language. transl. P.O.
Hertz. San Francisco: Harper and Row. 1982).
38 A more complete historical account of what I have called the experience or sense of the immanence of lan-
guage would discuss the expression of this feeling in the Romantic philosophy of language rooted in the
work of lG. Herder and W. von Humboldt. In this tradition, language is seen as a medium of thought that
is internally related to culture and human life. Such a study would begin with the German philosopher. Jo-
hann Georg Hamann (1730-1788). probably the first thinker to formulate these ideas. Hamann was a critic
of German Enlightenment philosophy, and is thought to have exened a great influence on Herder. Hegel.
and Kierkegaard. among others.
28 SOREN STENLUND
that human beings are trapped in language is as much nonsense as the opposing thesis
that we can transcend language. The borders of language are not borders in the sense that
they constrain our will. It is not as if we could only imagine or surmise something be-
yond the boundaries of language, but not get there, even in thought. The metaphor of a
border or boundary has become misleading when it brings us to the idea of 'things-in-
themselves' or 'thoughts-in-themselves'.
On this point, Frege is still unclear, as can be seen from the quotation above, in
which he brings the anthropological perspective into play. That passage continues thus:
There is no contradiction in supposing there to exist beings that can grasp the same thought
as we do without needing to clad it in a form that can be perceived by the senses. But still. for
us men there is this necessity.3Y
Perhaps there is no contradiction in supposing there to exist some being who was not
'bound' by language, but so what? Why does Frege make this point? Apparently, be-
cause he experiences the immanence of language as something disturbing and limiting,
but at the same time, unavoidable. Linguistic immanence seems to open the floodgates
for the historicism and relativism of which he saw, to his dismay, so many examples in
his day, and this disturbs him. But he cannot avoid the new historical consciousness that
seems to confirm the feeling that we human beings are bound by language. It is against
the background of this concern, I believe, that Frege inaugurates his 'semantic Plato-
nism'.
The Philosophical Discourse of the Mental
Seventeenth and eighteenth-century philosophers express themselves in a relatively com-

mon vocabulary toward which they display a kind of immediate trust, a trust that disap-
pears during the nineteenth century. It is a 'mental language' in which the central word is
'idea' or, in German philosophy, Vorstellung. Book II of Locke's Essay has the title "Of
Ideas", but no precise definition of an idea is given there. The concept is central in the
sense that it is assumed that its meaning is self-evident to anyone competent in everyday
language: an idea is what words (in the grammatical sense) mean or signify. Ideas are
simple or complex, and they constitute parts of thoughts in the same way as words con-
stitute parts of sentences. And the philosophers speak of this mental world of ideas and
thoughts, the human mind, as something independent of, and more fundamental than,
language itself. Hobbes, for example, says "the general use of speech, is to transfer our
mental discourse, into verbal; or the train of thoughts into a train of words: AO
The common use of this mental philosophical language does not mean that there pre-
vailed unanimity among philosophers as to how one should view this mental sphere; on
the contrary, it was precisely such questions that were the topic of the philosophical dis-
cussion of the period. One argued about whether or not there exist 'innate ideas', or if
39 Posthumous Writings, p. 269.

40 Cited in Hacking, p. 15.
there are 'abstract ideas', for example. The unanimity of which I speak consists simply
in that everyone nonetheless expresses his views, executes his analyses and presents his
answers within the mental vocabulary, that is, within a language in which one speaks of
the human mind and its constitution, the human understanding or human reason and its
functions. One takes the vocabulary of the mental, correctly used and properly under-
stood, to be dependable for achieving its purpose, namely, for arriving at the truth of
how things really are, and for expressing that truth adequately. One relies on the philo-
sophical vocabulary of the mental as if there must be some way to use and understand it
so that it is, as it were, transparent.
Descartes, Locke and Berkeley all speak as if one could step outside of language and
turn inward in an immediate mental vision of the ideas, but they never actually leave the
mental vocabulary. They think in it, and they conduct their thought experiments in it.
They describe their studies and the consequences of them in the philosophical vocabu-
lary of the mental, and they criticize their antagonists in it. The clear and distinct ideas of
which Descartes speaks are clear and distinct when they are formulated in the language
that he himself finds immediately apposite. Thus it is not as if Descartes first arrives at
his complete, clear and distinct ideas, and then takes up the problem of how to express
them in language (even if the question of how they are ultimately to be formulated for
publication comes into the picture somewhat later, but the considerations involved there
are not just those of faithfulness to the Truth).
One important aspect of what it meant for Descartes to turn inwards, I believe, is that
he turns away from the traditional way of using the philosophical vocabulary of the men-
tal that was expected of philosophers at the time. He turns away from the usual way of
relating to it in different ways that he found obsolete, opaque, empty, pernicious: " .. .I
pay no attention to the way in which particular terms have of late been used in the
schools ... ," writes Descartes when describing mental vision as a faculty of knowledge in
Rules for the Direction of the Mind. In this respect, one can indeed say that Descartes
'steps outside of language', at least the philosophical language that he was expected to
use. But his reason for doing so was in order to find another way of saying what had to
be said.
Regarding the search for an immediate relation to the ideas, as I see it, the search is
largely for an apposite and authentic philosophical language in which to formulate im-
portant insights. The immediacy sought consists in a relationship to a language and way
of speaking, one that feels apposite, true and transparent. With respect to Descartes early
talk of leaving external appearances and turning inward in mental vision, what it means
concretely, in practice, as it were, is what he describes in the first chapter of the Dis-
course on Method. There he describes how he has become the thinker he is, for example,
how he views himself in relation to the philosophical, theological and scientific estab-
lishment, societal institutions and so forth.
From our point of view, seventeenth and eighteenth-century philosophers' talk about
leaving external appearances and turning inward in an act of mental vision is, one could
say, a metaphor or allegory. But it was a metaphor that they did not regard as a meta-
30 SOREN STENLUND
phor, just as they did not read biblical stories as allegories as we do today. In our era, it
is natural to regard the bible as a great work of literature; this was certainly not the case
in the seventeenth century. Did they take the notion of 'mental vision' literally? In one
sense, yes, but hardly as what it would mean/or us, for most philosophers today, to take
it literally. We simply cannot use their mental discourse with the same gravity as they
did.
Our problem with this mental discourse is that 'literal' discussion of the mind and its
inner workings is something of a mystery. It is something about which one speculates
wildly in modern philosophy and science (for example, within so-called 'cognitive sci-
ence'), which suggests that there prevails an enormous confusion about these matters.
Problems involved in 'the philosophy of mind' are perhaps the primary and most vital
philosophical questions of our day. One thing that makes it difficult for us to understand
philosophers in the past is that our philosophical vocabulary, the system of classification
that we take seriously as professional philosophers, has been formed in a context in
which all talk of the 'literally mental' is felt to be 'mystical' or at least mysterious. This
same context is one in which we do not find it mystical or mysterious to speak literally
of 'the middle-sized physical objects' that we see around us (but to which we do not
have an especially intimate relationship). It is against this background that we employ
terms such as 'subjectivism' or 'idealism', for instance, and find them at times apposite
and clarifying. But it is not at all obvious that these terms are of any help in trying to un-
derstand the thinking of philosophers in the seventeenth and eighteenth centuries.
Ian Hacking argues that Descartes and his contemporaries did not have the same vis-
ual picture of reality as we do. For this reason, says Hacking, the assumption that the
sense of sight remains constant over time, while concepts fluctuate, is problematic. He
writes:
There is little reason to suppose that Descartes thought of eyesight as we do. The world has
been turned inside out. Ideas, in the time of Descartes, were il~ centre. Objects have been re-
versed, so that what we call subjective was then called objective. We can hardly expect 'vi-
sion' to stay politely in il~ place. 41
That there occurred such a change in the employment of the terms 'subjective' and 'ob-
jective' is clearly correct, while the rest is, of course, speculation. But there is a point to
such speculation if it encourages us to greater circumspection and self-consciousness
with respect to the conceptual apparatus of our own epoch.
It is because seventeenth-century philosophers think in the language of the mental
that they take for granted that their explanations are not merely subjective, psychological
reports about something private, but are, on the contrary, universal and objective, as well
as important and fundamental truths about humanity. Their confidence in the philosophi-
cal vocabulary of the mental that they use is expressed in the grand claims they make by
means of it; it is also manifested, naturally, in the serious spirit in which such claims are
taken, expressed in this way, by the intended audience. The claims are taken seriously by
41 Hacking, p. 32.
many at the time largely because they are expressed as they are, for it is the vocabulary
in which one was expected to discuss important and fundamental philosophical and theo-
logical matters.
The linguistic tum in philosophy begins as the philosophical vocabulary of the men-
tal gradually comes to be understood in a more profane, empirical psychological sense
during the nineteenth century. In other words, it begins when the psychologistic under-
standing of that language is considered the most natural and clearly correct one; it begins
when this language becomes a source of inspiration for empirical psychological theoriz-
ing and the attempt to make psychology a natural science. In the new understanding, the
mental is no longer something to which we can have an immediate relation, as we find,
for example, in the writings of Berkeley; now it is rather something 'mediated', some-
thing about which one makes hypotheses and for which one formulates laws. There
arises then a new consciousness of the vocabulary of the mental as a language, of the
psychological concepts as concepts, etc., for which one can formulate scientific require-
ments. There is a new awareness that one never actually 'leaves language' when one.
like Berkeley, turns inward 'in order to free himself from the deceptive appearances of
language.' The methods of Berkeley and Hume seem like introspective psychologizing.
And the step from this awareness to the feeling that we can never step outside of lan-
guage. what I have called the experience of the immanence of language, is not a great
one. In those cases where the psychological vocabulary is still employed for philosophi-
cal purposes (as, for example, in the early Husserl's Philosophie der Arithmetik). it is re-
garded as a threat (by Frege, for instance) to the truth and objectivity at which
philosophy and the sciences in general aim.
A Common Human Nature
How could the philosophical vocabulary of the mental have lost the status it had among
seventeenth and eighteenth-century philosophers? The answer to that question, I believe.
requires nothing less than a description of the shifts in European intellectual. spiritual
and cultural life that took place between the seventeenth century and the end of the nine-
teenth century. And these in turn are intimately bound up with other changes in forms of
life, for example, with the effects of the industrial revolution. Any description of these
changes is bound to be a simplification and schematization in which one chooses to em-
phasize certain specific aspects. On the other hand. most of us (among those who care
one way or the other) will always have some simplified and schematic notion of what
happened.
Having said that, there is one aspect that I would like to stress here. A suitable start-
ing point is the following remark by Aarsleff:
The nineteenth century had lost the 18th century's fundamental conviction, its first theory we
might say, that of the uniformity of human nature.42
42 Aarsleff, p. 110.
32 SOREN STENLUND
I think that Aarsleff has noted something essential. That there is such a thing as the hu-
man nature, the human understanding, or the human faculties of knowledge was a self-
evident starting point for almost all theorizing in the eighteenth century. That conviction
leaves its mark on the formulation of the theories and on the architecture of the inteJlec-
tual constructions. It was not merely a philosophical stance or ideological position about
which everyone was in agreement; it was rather the framework in which problems,
standpoints and positions were formulated.
One example of this is Condillac' s doctrine of 'natural signs', which is intended to
answer the question left unanswered by Locke. In practice, linguistic communication
works rather well; we often do understand each other. How is that possible, if it is in
principle impossible? Locke leaves that problem unsolved, as if it were not so terribly
problematic. Condillac responds to the problem via a kind of genetic explanation of the
origin and development of language. He says that artificial signs, that is, the words of
our languages, which by themselves are arbitrary and unmotivated, must have arisen as a
result of an unbroken and non-arbitrary process from natural signs (that is, gestures, re-
actions, facial expressions, etc.). Artificial and conventional signs have arisen through
analogy with natural signs, which have the same meaning for all people since they have
their origin in a common human nature. We understand each other in practice because
our language has its origin in one nature shared by all human beings. 43
This explanation becomes troublesome for many in the nineteenth century, due to an
increased historical awareness. What we call historical research becomes more system-
atic and empirical, especially in anthropological studies. Anthropology as a scientific
discipline develops out of studies of the classification of human races (with respect to
anatomical features, for instance) and out of comparative studies of languages, societies
and their economic developments, etc. 44
What becomes striking then is everything that speaks against 'the uniformity of hu-
man nature', namely, the differences and variations between languages and linguistic
communities, between different peoples' ways of being and ways of life, in different
parts of the world and in different historical epochs. These observations support rather a
thesis of linguistic relativity, which actually was not a new idea in the philosophy of lan-
guage: it is stated more or less explicitly in the works of Rousseau, Herder, and von
Humboldt. But it is only toward the second half of the nineteenth century that this idea
comes to be seen by many as a concrete reality and an historical fact.
A kind of communicative scepticism arises, not so much as a theoretical standpoint in
philosophy as a feeling, an unsettling possibility, a discomfort that could no longer be
quieted by the idea of a common human nature. 45 Perhaps we do not really understand
43 This aspect of theories of language, in Condillac and others, is developed in an interesting direction by
Talbot Taylor in Mutual Misunderstanding, cited above.
44 Some historians count the breakthrough of evolutionary theory (around 1840) as the birth of scientific an·
thropology.
45 Another expression of the lost faith in a common human nature was the emergence of the kind of existen·
tial philosophy found in the works of Nietzsche, Kierkegaard and Dostoyevsky. This kind of philosophy
each other at all. Perhaps we all speak different languages. According to Frege, that
would be the case if the psychologistic notion that words signify ideas were correct,
since, says Frege, "Another man's idea is, ex vi termini, another idea." If psychologism
were correct, Frege states, it would be impossible for the claims of two individuals to
contradict one another: a contradiction can only occur when "it is the same thought that
one person is asserting to be true and another to be false.,,46
Against the background of the experience of scepticism, it becomes pressing for
many language theorists to emphasize the social nature of language. In the internal histo-
riography of linguistics, this thought is often associated with Saussure, who is supposed
to have made the important 'discovery' that language is a social institution. Certainly, he
does incorporate that idea into his theory of language in his own way, but the idea ex-
isted earlier. As Aarsleff shows, even other linguistic theorists of the day (for example,
Madvig, Breal, and Jespersen) were eager to emphasize the social and communicative
aspects of language. 47
Aarsleff provides a convincing account of Breal's influence on Saussure, especially
with regard to his view that linguistics is not a natural science. Breal claims that "Lan-
guage is a human activity: it has no reality apart from the human mind." (Aarsleff, p.
382). Breal also writes in his Semantique: "Speech is first of all a means of communica-
tion: it would lose the most essential of its functions if it ceased to serve for the exchange
of ideas." (Aarsleff, p. 383). The communicative aspect, as Taylor points out as well, is
also Saussure's point of departure when he paves the way for the introduction of his con-
cept of la langue in the Cours.
The social nature of language becomes a key concept in the effort to appease the
anxiety aroused by the thought that we perhaps do not really understand one another. It
serves to explain and establish the possibility of communication. It works as a kind of
demonstration that we are often right when we take ourselves to understand each other in
our everyday conversations!
Saussure and Frege accomplish this through a new conception of language in which
they build upon its social or 'intersubjective' character as an essential or necessary attrib-
ute. Frege does this by focussing on the objectivity that the meaning of statements that
can be true or false must have. Saussure does it by introducing the notion of la langue,
the social linguistic system internalized in the mind of speakers of the language. (An-
other, more modern, variation on this theme in linguistics, one based in modern natural
was largely a response to, and rejection of, the idea of a common human nature containing faculties of 'the
will' or 'practical reason', such as constitutes the ground of Kant's moral philosophy.
46 Frege, Posthumous Writings, p. 133.
47 Aarsleff writes: ..... during the 1890's it became evident on all sides that the study of language was being
transformed. Saussure, Jespersen and Meillet stated the principles which they continue to argue and de-
velop; Saussure in the Cours, Jespersen in subsequent writings [ ... J and Meillet by his well-known com-
mitment to the sociological aspects of language and its study. Bn:al's Essai semanlique was often
re-issued. The reaction against the German tradition was triggered by Schleicher and continued to be
aimed at his French devotees."
34 SORENSTENLUND
science, is Chomsky's theory of language, in which the idea of a biologically grounded

universal grammar is central.)
Talbot Taylor applies the apposite term code theory (of language) to these concep-
tions. A language is seen as a code; the words and sentences codify a content or a mean-
ing such that it can be deciphered by a grammatical or semantic analysis. Since a
language is the same code for all users of language within a linguistic community, the
social nature and communicative potential of language are ensured: speakers of the same
language use the same code, and can mean the same thing by the same words.
It is characteristic for this code conception of language, which has dominated linguis-
tics and philosophy of language throughout this century, that one distinguishes between
language itself, or the linguistic system that constitutes the code (and which, as a rule,
has taken the form of a calculus, a formal system), and the linguistic practices, the uses
of language, in which the code is employed. 48 Distinctions such as langue and parole
within structuralism, or competence and peiformance in Chomskian linguistics, have
played this role within their respective theories. Within analytic philosophy of language,
the linguistic system itself is determined by the syntax and semantics of language, while
pragmatics has described language use on the basis of syntactical and semantic descrip-
tions of 'language itself .49
Within the hermeneutic tradition (taken broadly) this code conception has been criti-
cized for having lost sight of the literary, poetical, figurative, aesthetic, evaluative and
historical dimensions of language, and numerous attempts have been made to correct and
supplement it in these respects; but at bottom, the conception has hardly been questioned
at all. Radical doubters such as Derrida in the structuralist tradition and Davidson in ana-
lytic philosophy have admitted that language as described by linguists and philosophers
does not exist 'in reality', but they yet argue that we must continue to act as if it did. In
the philosophy of the later Wittgenstein, this code conception is radically called into
question; it is shown there that many of the philosophical problems of our day have their
source in this conception, and not only those directly related to the nature of language. 50
Returning to the eighteenth-century belief in the uniformity of human nature, I said
earlier that this belief was an important aspect of the faith in the philosophical discourse
of the mental as the discourse in which all truths and matters of importance were to be
expressed. The conviction that there exists a single common human nature could be
48 The code metaphor is sometimes used in other, less illuminating ways. According to Dummett, for exam-
ple, to understand language as a code is to understand it as a technique for the transmission of (non-lin-
guistic) mental content; perhaps this is a more natural way of understanding the notion from the viewpoint
of modem information technology and its approach to language. From this viewpoint. one is inclined to re-
gard languages in general as artificial languages. (Dummett, M., The Seas of Language, Oxford: Claren-
don Press, 1993. p. 97.).
49 This use of the term 'semantic' in philosophy and modem linguistic theory is fundamentally distinct from
its use in the work of Breal, who is probably the first to have employed the term. Bn!al's semantics be-
longs to an historical study of language.
50 This theme is discussed in detail in Stenlund, S .. Language and Philosophical Problems, London: Rout-
ledge, 1990.
found earlier, of course, but it appears to have had a more religious or theological sig-
nificance then. The belief in a common human nature was part of the belief in the Crea-
tor, who was, in the words of Descartes, the source of truth. The nature of a human
being was, as even Descartes often uses the term, the totality of all that God has given
man. The soul of which Descartes speaks, of which he claims it is easier to gain knowl-
edge than of the body, is also (for Descartes) the soul described by Christian theology.
When Descartes conducts his examination of the passions of the soul, he speaks of "the
passions in general and in connection with human nature as a whole." These are not pri-
vate psychological studies, or what we would normally mean by 'psychological studies'
at all. If one classifies them in this way (or, as I have done here, with the term 'mental
vocabulary'), the classification can be nothing more than a kind of 'surface grammar':
one seizes upon the words or terms Descartes uses and not the purpose they were in-
tended to serve in his philosophy.
I believe that secularization is an important element in the changes that lead to the
decline of the philosophical vocabulary of the mental. We tend to underestimate this as-
pect because it is so difficult for us to imagine what it would mean to accept this psycho-
logical-theological talk with the gravity discernible in the works of seventeenth-century
philosophers. It strikes us rather as a kind of myth or fairy tale that people 'believed in'
back then, since they did not yet know any better. The dominant scientific spirit of our
own era is rather characterized by a faith that secularization clears the path for a broader
and deeper and more enlightened understanding of our own epoch as well as of others. I
will not deny that there is ample evidence in support of that view. But I would like to
point out that the opposite is also true; secularization has blocked our view of ways of
thinking of other epochs. It is but one example of the more general condition that, as
Aarsleff puts it, "views that became accepted much later can completely block our un-
derstanding,,5 I. Descartes speaks of God as the source of truth - this does not mean that
he was foolish.
Natural philosophy at the time, even in the form of 'the new science', was not secu-
larized. Science in our sense, however, is necessarily secular. God's authority is still in
force in the seventeenth century, even if He has absconded or become more removed
from his creation than he was for certain medieval philosophers. He created the world
and set in motion the causal chains, but then left it on its own. For Newton, the ultimate
cause, the first link in the causal chain, is the Creator himself. Galileo, and many others
with him, thought that it is only first when man begins to investigate nature that he
comes to appreciate fully God's true greatness. Bacon assures his readers that natural
philosophy does not undermine the authority of religion, but is in fact the best antidote to
the kind of superstition that hinders a correct understanding of God's will.
John Locke held a similar position, which can help to explain, I believe, why he
leaves open the question of how we can in practice understand one another in our every-
day dealings, even though, according to Locke, the nature of language is such that it is in
5I Aarsleff. p. 10.
36 SORENSTENLUND
principle impossible. "Locke only examines defects inherent in language," writes Talbot
Taylor, "he does not explain how, in spite of these defects, language can function suffi-
ciently well for the 'vulgar' purposes of ordinary communicational interaction.,,52
Locke's lofty aims gave him no reason to bother with the sort of linguistic communica-
tion that only served 'the vulgar purposes of everyday discourse." Locke writes in his ca-
pacity as member of the Royal Society of London for the Promotion of Natural
Knowledge, the primary purpose of which was to promote the new natural philosophy,
but which also had moral and religious aims. Aarsleff describes John Wilkins, who was
one of the first to introduce the new natural philosophy and its ideas in England, and
who helped formulate the objectives of the Royal Society:
Whether we call some of his writings scientific and others religious is a matter of emphasis;
they all have the same aim: to guide man's conduct toward moral virtue, religious devotion,
and ultimate hope of salvation. The pursuit of happiness, even comfort, in this world is man's
legitimate interest. 53
Locke's Essay Concerning Human Understanding was ultimately motivated by such re-
ligious and moral aims. Locke appeals to the man who is in search of Truth, and who
wants to express the truths in a way that harmonizes with these moral and religious
goals.
According to Locke's philosophy of language, how a word is associated with an idea
is an act of the will on the part of the individual. The individual language user has the
freedom to associate any idea he wishes with the words that he uses. Some of this free-
dom remains even after he has adapted himself to common and standard usage. 54
Locke's philosophy of language is thus formulated such that it leaves room for moral
recommendations of sorts for how language ought to be used for philosophical and sci-
entific purposes.
Later generations, for whom religious and moral aims became less important as 'pro-
fane life' took on greater significance, could not, like Locke, simply disregard 'the vul-
gar uses of language'. Explaining how it is possible for us to understand each other even
in 'vulgar' everyday situations comes to seem like an inevitable and urgent problem for
philosophy.
If the aim of Locke's Essay was ultimately moral and religious, one might wonder
why he needs to devote so much of his energies to questions concerning the nature of
language that he develops a detailed theory of language. Part of the reason, I would sug-
52 Taylor, p. 44.
54 Locke is aware of the social nature of language, but does not regard it as a sufficiently secure ground for
successful communication. He writes: "Indeed, the necessity of communication by language, brings men
to an agreement in the signification of common words, within some tolerable latitude, that may serve for
ordinary conversation: and so a man cannot be supposed wholly ignorant of ideas, which are annexed to
words by common use, in a language familiar to him. But common use, being but a very uncertain rule,
which reduces itself at last to the ideas of particular men, proves often but a very variable standard."
(Locke's Essay, Book III, ch. xi).
gest, is that he wants to collect, summarize and justify the view of language that had
been developed to harmonize with the new natural sciences (i.e., the ideas that had come
from Copernicus, Kepler, Galileo, Gassendi, Mersenne, and others), and to adapt them to
his moral and religious purposes.
Given Locke's attitude toward these lofty goals, this 'adaptation' coincided with an
explanation of how language must be. It was a matter of Locke's attitude, and not his
opinion or a political strategy. One could not so easily distinguish at the time, as we do
today, between political motivations and religious fervour. The inability to distinguish
between these two is something we associate with religious fundamentalism. But Locke
was certainly no fundamentalist. On the contrary, one could say that Locke (like
Mersenne, Descartes and others) paved the way' for the secularization that took place in
the eighteenth and nineteenth centuries; they separated the Creator from his creation, and
made the understanding of God's creation the business of human reason.
Locke formulates a philosophy of language, a theory of knowledge, and a political
philosophy in which it is easy to disregard the religious aspect. Because his philosophy
of language is formulated to harmonize with the new natural sciences, it lends itself eas-
ily to a 'profane reading'; this is one reason, I believe, why it lives on, primarily through
the mediation of Condillac, well into the second half of the nineteenth century. The
thought that words stand for 'ideas in the mind of the speaker' and not for things is a
generally accepted view among nineteenth-century philosophers of language, even
among those who are otherwise known for having different viewpoints with regard to the
question of the nature of language. 55 The idea can also be found in linguistics and philol-
ogy in the twentieth century. Other important features of the Lockean view of language
that live on are: (i) language is a social and conventional artifact, something created by
human beings; (ii) words are arbitrary signs; (iii) the purpose of language is communica-
tion.
The Adamic Doctrine of Language
The background to Locke's philosophy of language is still inadequately described if one

fails to take into account what it was that this philosophy is challenging. Locke desig-
nates the view of language that he wishes to combat 'double conformity', the common-
sense thought that there exists a kind of correspondence, both between words and the
things for which they stand, and between the ideas that different speakers associate with
the same words. In this view, words are not arbitrarily formed, but were originally cre-
ated to conform to the nature of things.
Ever since the Renaissance, there existed a philosophico-theological doctrine of lan-
guage in which this idea of a double conformity was central. The formulation of this
55 It is this expression of psychologism in the philosophy of language that becomes problematic for Frege,
especially as it is linked up with historico-anthropological and subjectivist theories in the Gennan Roman-
tic tradition stemming from Herder and von Humboldt.
38 SOREN STENLUND
doctrine, which Aarsleff calls the Adamic doctrine of language, was prevalent well into
the seventeenth century. It is based on the story of creation in the Old Testament. One
imagined Adam as the archetype for the human being, in every respect, but especially
with regard to language. When Adam gave names to the animals and plants in the Gar-
den of Eden, it was in a state of perfect knowledge and insight. He was created by God
and received his knowledge directly from God. The names he gave to the animals and
plants were therefore in accordance with their nature or essence. A perfect harmony be-
tween words and things prevailed. Since all human beings are descendants of Adam and
Eve, all languages have a single common original language and therefore exhibit com-
mon features. The original language was believed by many to have been Hebrew, but
there was no unanimity on this point. Disagreement about this led to exhaustive com-
parative language studies which, according to Aarsleff, were "much more fruitful and
competent than the nineteenth century's inordinate pride in its own comparative, histori-
cal philology, has allowed us to know. Leibniz, for example, had great respect for this
work, which he knew intimately.,,56
Whatever the case may be on that score, these language studies show how seriously
one regarded the Adamic doctrine, while we are inclined to grin at this fanciful story.
With what could we compare this 'linguistic theory' today? Chomsky'S ideas about in-
nate universal grammar? Fodor's ideas of what he calls 'mentalese', our ostensible men-
tal language? One difference is that many of us can already grin at these speculations,
which are built upon an adoration of modern science that is not so much religious wor-
ship as a kind of blind faith.
Among those who embraced the Adamic doctrine in the seventeenth century, it was
also common to think that the original language deteriorated and was split up into sev-
eral different tongues at Babylon, and that these languages were those actually spoken at
the time. 57 But it was thought that they all still contained elements from the original pre-
Babylonian language. The Creator was, in this sense, still present in his creation. And it
was hoped that it should be possible to reveal the forms of the original language through
exhaustive comparative studies. Insofar as the original language was a direct reflection
of the nature of things, these hopes also contained the philosophical aspiration of uncov-
ering essences.
The Adamic doctrine took different shapes. Some were inspired by neo-Platonic
ideas and the Cabalah during the Renaissance, and had a mystical and esoteric character.
One representative of this variant was Jacob Bohme who, according to Aarsleff, was ex-
tremely influential. Bohme claimed that, in a moment of divine inspiration, he had been
given insight into the nature of things and could understand the Adamic language. 58 At
57 The eleventh chapter of the first Book of Moses opens with the words: "Everyone on earth had the same
language and the same words." In verses 8 and 9, one then reads the following: "Thus the Lord scattered
them from there over the face of the whole earth; and they stopped building the city. That is why it was
called Babel, because there the Lord confounded the speech of the whole earth."
58 Bohme, who lived until 1624, was probably an important source of inspiration for Hegel and Schelling.
such moments, words became direct visions of essences, even in a quasi-scientific sense.
There were those who thought that one could arrive at insight into the nature of the met-
al, gold, through an authentic and primary understanding of the word, 'gold'.
This way of thinking was perceived, understandably, as a threat to the new science of
nature, and it was more than anything else these sorts of ideas (which were also to be
found in England) that worried Locke and motivated his extensive studies in the philoso-
phy of language. It was against this gold example that Locke claims that a study of the
word cannot provide us with anything more than 'the nominal essence'; it cannot give us
'the real essence'. And it was against this particular feature of the Adamic doctrine that
Locke stresses that words stand for ideas and not for things, that the connection between
word and idea is arbitrary, volitional and individual, and that language is a conventional
human artifact.
Leibniz also distanced himself from the mystical esoterical varieties of the Adamic
doctrine, but he exhibited a related view of language in his thinking, one which is quite
distinct from Locke's. In Leibniz' philosophy, one could say, nothing is arbitrary or con-
tingent, and everything hangs together. His main object of contention in the Nouveaus
essais sur ['entendement humain, which contains a critique of Locke's philosophy of
language, is the principle of arbitrariness.
One could say, even if it is an anachronism, that Leibniz emphasizes what Saussure
would call the 'diachronic perspective', while Locke's perspective is 'synchronic'. Leib-
niz conducted what we would describe as extensive 'empirical' studies of different lan-
guages and various dialects. His linguistic studies are thought to have been important for
later historical comparative linguistics, but it would be mistaken to assess his research as
if it were motivated by historical or scientific aims in our sense (these aims first emerged
in the nineteenth century). For Leibniz, language is something natural, but not in the
sense of 'nature' that belongs to our secularized natural science. Leibniz' language stud-
ies were rather a part of his endeavour to affirm the divine, pre-established harmony. His
philological and etymological studies confirmed his view that there exists a harmony be-
tween the different languages in the world and that, as Aarsleff puts it, "there ultimate~
is a single radical and primitive language that underlies all the languages of mankind."
Leibniz, like the proponents of the Adamic doctrine, assumed that there is something
natural in the origin of words, that there is a connection between words and things, how-
ever difficult it may be to find traces of it. The purpose of many etymological studies in
the seventeenth century was to uncover these traces. 60
59 Aarsleff, p. 59.
60 It is not far-fetched to see Saussure's philosophy of language as an attempt at unifying the two apparently
conflicting views of language found in Locke and Leibniz (the synchronic and the diachronic). More pre-
cisely, one can see it as an effort to merge later, more secularized versions of them. But one may ask if this
was not already accomplished by Condillac. From that perspective, the essential difference between Con·
dillac and Saussure is that the former works within a kind of self-evident metaphysical faith in the uni·
formity of human nature, while in the work of the latter, this faith is replaced by a more scientific,
socio-psychological framework. But given Saussure's thesis of la langue, one can hardly say his theory is
40 SOREN STENLUND
Naturally, one could say that seventeenth-century philosophers used etymological

studies in order to confirm and reinforce their prejudices and preconceived notions. One
may say so provided that one does not pretend that we no longer engage in theoretical
constructions in order to confirm our prejudices. One could argue to the contrary that our
century takes first prize in this respect. In that case, one difference would be that the jus-
tifying and confirming theoretical constructions of our day leave much less room for us
as human beings than the intellectual edifices of the past did.
There is another reason for taking up the Adamic doctrine in this context. It is an ex-
ample of what I described earlier as the view of language as a code. Words as such, as
linguistic expressions, are thought to codify a (hidden) content; the etymological studies
are aimed at deciphering the code so as to reveal the content. In the seventeenth century,
there were many who took seriously plans and projects to create a universal language for
philosophy in which the hidden content is made explicit in the forms of the language; in
which the forms of language correspond directly to the nature of things; in which the
structure of language conforms to the structure of reality (in today's philosophical jar-
gon).61 In my view, it is not unreasonable to compare these projects with certain ideas
about logically perfect artificial languages in early twentieth-century analytic philoso-
phy, where a modem scientism takes over the role previously enjoyed by theology. The
search for the correct method for logical analysis or conceptual analysis or semantic
analysis in early analytic philosophy would then be the counterpart to seventeenth-cen-
tury etymological studies. In both cases, there is a kind of metaphysical essentialism at
work which justifies the manner of studying language, such as the metaphysics of logical
atomism or the very nature of 'being true' (in contrast to 'taking to be true') in the work
of Frege.
In the philosophy of Leibniz, one finds the idea of a lingua characteristica and a cal-
culus ratiocinator, something which Frege takes up in connection with his elucidation of
the aim and purpose of his Begriffsschrift. Frege expresses sympathy with Leibniz' goal
of finding "a system of notation directly appropriate to objects themselves.,,62 Frege
adds, "it is possible to view the signs of arithmetic, geometry, and chemistry as realiza-
tions, for specific fields, of Leibniz' idea. The ideography proposed here adds a new one
to those fields, indeed the central one, which borders on all others.,,63
Aarsleff argues that the form of essentialism and the idea of natural signs that one
finds in the Adamic doctrine has a counterpart in the nineteenth century, but in the sec-
ond case in alliance, not with theology, but with secularized science. Frege's 'semantic
metaphysics' (which Aarsleff does not mention) is an example of this. Frege's aversion
less metaphysical. It would be more accurate to say that it exhibits a metaphysics that is more in harmony
with its times.
61 The Englishman named earlier, John Wilkins, presents plans for such a project in his major work, An Es·
say Towards a Real Character and a Philosophical Language, the purpose of which was "to repair the ru·
ins of Babel."
62 Begriffsschrift, p. 6.
63 BegriJfsschrifi, p. 7.
to the idea of the arbitrariness of the sign, even when applied to mathematical signs, is
apparent in his criticism of the formalists. His idea of a system of notation that would be
'directly appropriate to the objects themselves' is the first half of 'the double conform-
ity'; the other half is the communicability that characterizes the thoughts expressed by
meaningful statements. 64
I said earlier that the code conception of language, which formed the background
against which one began posing semantic questions and postulating theories of meaning,
did not exist until the end of the nineteenth century. Locke's view of language, in which
words are tied to ideas according to the individual speaker's free choice, is most defi-
nitely not a code conception. But if one were to compare twentieth-century semantic dis-
cussions or theories of meaning within analytic philosophy with anything found in the
seventeenth or eighteenth century, it would be with just this form of 'genetic naturalism'
found in the Adamic doctrine and in the (more secular) ideas about the origin of lan-
guage as conceived by Condillac and others well into the nineteenth century. All of these
are based on a code conception of language that is not to be found in Locke's writings. 65
The decisive difference is that the twentieth century has interpreted the code as a
structure; a structure to be represented in a theory, with the 'ahistorical' notion of a the-
ory stemming from modern mathematics and the natural sciences, while seventeenth and
eighteenth-century philosophers treated similar questions as questions having to do with
the origin of language. It was this difference that I had in mind when I said that the lan-
guage about which one posed semantic questions and constructed theories of meaning
did not exist before the turn of the century. It is about the code conception of language
created towards the end of the nineteenth century that one begins posing semantic ques-
.
tlOns and postuIatmg
' theones . 0 f meamng.
. 66
64 Ian Hacking questions Aarsleff s reading of Leibniz in the article "Locke, Leibniz, Language and Hans
Aarsleff' (Synthese 75, 1988, pp. 135-153). Hacking is right to suggest that Aarsleffs critical attitude to-
ward German linguistics and philosophy of language stemming from Leibniz is a recurring strain in the
book, and at times it disturbs the presentation. On the other hand, Hacking seems unwilling to see the
wider implications of 'Adamism' that Aarsleff, as an historian, uses in order to show related features
among different views of language at different times. In this wider use of the term 'Adamism', Locke's
'double conformity' is the essential feature. This use of the term is illuminating, for it helps us to see the
'scientific Adamism' of our own century in analogy with the theological Adamism of the seventeenth cen-
tury.
65 The more expressivist ideas about language that became 'popular' in early nineteenth century German Ro-
manticism in the aftermath of Herder and von Humboldt are perhaps exceptions to the code conception
(the characteristic trait of which is that one isolates 'language il~elf. the linguistic system, the code, from
its uses and users). Romanticism is fixed on the notion that a language expresses the character of a lan-
guage community, that someone's way of speaking expresses his personality. and so forth. Language theo-
risl~ such as Madvig, Jespersen, Breal, Whitney, Saussure and even Frege reacted to this historicist and
subjectivist feature of the Romantic conception of language by emphasizing the social and 'intersubjec-
tive' character of language.
66 There are interesting differences between Saussure's structuralism and the more naturalistic structuralism
in linguistic theory that dominates in the Anglo-Saxon and Nordic countries. Saussure' s structuralism is
not naturalistic; it does not contain any of the scientific materialism or reductionism that is characteristic
of psychology, linguistics and analytic philosophy of language in the American tradition. Nonetheless,
42 SOREN STENLUND
The Question of the Origin of Language
How ought one to understand these questions concerning the origin of language? If we
could answer this question, it might provide us with perhaps the most striking example
of how the new scientific and philosophical ways of thinking have obstructed our view
of past ways of thinking. Due to this obstruction, it is difficult to answer the question,
but it is all the more interesting and crucial for that very reason. It is crucial because it
can help us to become more aware of our own prejudices and, therewith, freer in our
own thinking.
On this point, Aarsleff is harsh in his critique of the nineteenth-century linguistic es-
tablishment:
It is one of the curious quirks of history that the nineteenth century became so compulsively
historical in the factual sense that it misinterpreted the eighteenth century on this important
point, with the significant consequence that linguistic philosophy came to be ridiculed as use-
less and jejune, 'unempirical' speculation quite unworthy of serious attention by right-
minded scholars; that is, by university professors who, from the limited but dominant
perspective of their chairs in comparative-historical philology, collectively saw themselves as
the sole legitimate authority on all matters relating to the study of language. 67
I am not in a position to judge whether or not this accusation is entirely just, but it would
not surprise me to find that Aarsleff is correct, given similar neo-positivist and anti-
metaphysical attitudes within the tradition of analytic philosophy.
The general lack of appreciation for the seventeenth- and eighteenth-century concern
with origin in the latter half of the nineteenth century is perhaps the most explicit expres-
sion of the decline of the older philosophical language, 'the mental discourse', and the
impossibility of employing that language for philosophical purposes. The explanations
of the origin of language were formulated in a mental vocabulary that had lost its authen-
ticity. It was not uncommon to ridicule the eighteenth-century philosophers' talk of 'fac-
ulties'. One preferred to speak instead of 'methods', in a more technical and formal
sense.
There has been a tendency (which one finds already in the writings of Frege) to dis-
miss earlier philosophical explanations of origins for two reasons, one philosophical and
the other scientific, both of which rest on a misunderstanding. To begin with, one dis-
missed them in the manner described by Aarsleff, that is, as speCUlative and deficient
historiography. Second, one objected that it was a matter of misplaced historical expla-
nation, that is, the problems were not of such a kind as to be accounted for by means of
Saussure's concept of a structure is influenced by the viewpoint of mathematics and natural science of his
day. How are we to explain the difference? Saussure's use of the viewpoint of the natural sciences is more
akin to what one finds in Taine, which is not reductionisl. Taine sometimes intentionally employs a zoo-
logical and botanical point of view to structure an exhaustive historical documentation of art. The purpose,
however, is not to claim that art is ultimately biological in nature. One does not find in the worK of Taine
the dogmatic attitude toward the scientific schemes of which he makes use that is so common in 'non-con-
tinental' twentieth-century psychology, linguistics and philosophy of language.
historical examination (this attitude is typical for Frege's logicism). Both of these objec-
tions take for granted the scientific notion of explanation that was guiding for the nine-
teenth century; the supposition seems to be that when eighteenth-century philosophers
wrote so many books and essays on the 'origin of language', they would have provided
such scientific explanations if they only could have.
One can illustrate the attitude obstructing our understanding of earlier philosophical
questions concerning origins with the following thought: "If the seventeenth-century
philosophers had only known about evolutionary biology, they would have naturally
used its methods and concepts rather than the Book of Genesis to explain the origin of
language."
But would it not have been a grammatical misunderstanding (in the Wittgensteinian
sense) to have asked an advocate of the Adamic teaching, for example, "When did the
Creation take place? In what year did God create Adam and Eve?" The story of Genesis
was not to be taken in that way. And if someone had answered, "At the beginning of
time", it would signify an end to all questioning, for there is the absolute origin, the be-
ginning. But this is difficult to digest in an era when the attitude is that everything can be
explained, because everything is regarded externally, in a spatio-termporal universe
wIt. hout b"
egmmng or end.68
The seventeenth- and eighteenth-century discussions of the origin of language in-
volved notions such as 'the original language' and 'the state of nature', but these were
not intended to be understood as factual historical phenomena localized in space and
time, just as little as today's physicists intend references to the 'organization of matter'
to be understood as if someone at some point in time went out and 'organized matter.'
One must not be mislead by the fact that they used the term 'history' for their explana-
tions. The term 'history' was used in a manner somewhat alien to our usage even among
natural philosophers (Faraday for example, wrote on "The Chemical History of a Can-
dIe"). It is difficult for us to understand what it would mean to take these explanations of
origin seriously, except in the sense of historical fact or evolutionary theory.
The original language and the state of nature were rather basic principles in what we
would describe today as philosophical explanatory models. In these models, one as-
sumed a genetic understanding of concepts in which all questions were subordinate to
the issue of a common human nature. Condillac presents his philosophy of language in a
treatise with the title, "An Essay on the Origin of Human Knowledge". His philosophy of
language is part of his explanation of the human faculties of knowledge. What is original
in Condillac's reasoning is his emphasis on the linguistic character of the facuIties, that
is, on the idea that thinking and reflection are essentially the use of artificial, arbitrary
signs.
68 This is presumably what Wingenstein has in mind when he says. U(] people today stop at the laws of na·
ture, treating them as something inviolable. just as God and Fate were treated in past ages. And in fact
both are right and both wrong: though the view of the ancients is clearer in so far as they have a clear and
acknowledged terminus, while the modem system tries to make it look as if everything were explained."
(Tractatus, 6.372).
44 SOREN SlENLUND
The term 'human' is important in the title of Condillac's essay. Aarsleff is correct to
point out that "this approach in terms of origins was man-centered" (p. 158). Perhaps one
can best understand this approach by contrasting it with the most natural way of viewing
knowledge in our own day. The eighteenth-century philosophers did not speak of knowl-
edge, as we do, as something 'produced' in the sciences as institutionalized practices,
which are also societal functions in a secularized society and which are activities per-
formed by experts or specialists, the scientists (or rather, researchers). We hardly associ-
ate their role with one of realizing a universal human nature in their scientific work; on
the contrary, we expect them to raise themselves above it in order to achieve scientific
objectivity. With this twentieth-century perspective, it is natural to replace questions
concerning the origin of knowledge with questions about the means and procedures by
which the researcher arrives at his knowledge as a product, a 'result'; these latter sorts of
questions are to be answered by a 'theory of scientific method' or 'studies of the knowl-
edge process', two genres in modern philosophy of science.
Philosophers in the eighteenth century were, of course, aware that nobody can pos-
sess the entirety of human knowledge, and that most of us will come to possess very lit-
tle. But, on the other hand, all human beings, qua human, possess the capacity to gain
knowledge: the faculty of reason or of the understanding, the capacity to think and to re-
flect. They have a part in the uniform human nature that the philosophers undertook to
describe and explain from the inside. What I mean by this is that the philosopher saw as
his task to discern this nature in himself through reflection, and to formulate it in the
mental vocabulary and by means of the genetic explanatory model which were seen at
the time to be authentic and adequate to the task. In other words, the concern with ori-
gins and the faith in the uniformity of human nature were intimately conjoined. As the
latter of these declined during the nineteenth century, the former also lost its weight. 69
As I said earlier, this faith becomes increasingly secularized or detheologized during
the eighteenth century (even if there certainly are exceptions). Allusions to a divine ori-
gin are no longer suitable for philosophy. Condillac says that it is not sufficient for a phi-
losopher to say that something has been achieved by extraordinary means; it is the
philosopher's business to explain how it has been achieved by natural means. The bibli-
cal story of Adam is irrelevant. In his discussion of the origin of language, Rousseau
says that religion offers a different explanation, but it does not forbid us from seeking
explanations based solely on human nature and the creatures around us, when the ques-
69 There are exceptions. Husserl's On the Origin of Geometry, for example, is not a work in the history of
mathematics. In this respect, one can regard Husserl's phenomenology as an attempt to revitalize and reo
establish the philosophical vocabulary of the mental. In certain works of Heidegger, one also finds a con-
cern with origins and a kind of 'etymological meditation' that has more in common with the seventeenth-
and eighteenth-century idea of science than with the view of science common among his contemporaries.
The founder of mathematical intuitionism, L.E.1. Brouwer, also displays a kind of 'originary thinking' in
his explanation of certain basic mathematical concept~. It is no accident, however, that these forms of 'ide-
alist philosophy' have been met with scorn in our day and judged (in my view, somewhat justifiably) as at-
tempt~ to breathe life into something that is dead.
tion concerns what would have become of humanity if we had been left to our own de-
vices. Herder too was of the opinion that alluding to a divine origin explains nothing.
For several decades, starting in the middle of the eighteenth century, a discussion
concerning the origin of language was conducted at the Royal Academy of Berlin. Ac-
cording to Aarsleff, the discussion culminated in Herder's "Uber den Ursprung der Spra-
che", a prize-winning essay in a subject that had been announced by the Berlin
Academy. The subject matter had the form of two questions, cited by Aarsleff:
Supposing that human beings were left to their natural faculties. are they in a position to in-
vent language? And by which means will they achieve this invention on their own?7U
The impression one gets from Aarsleffs survey is that these questions summarize basic
problems of the discussion. Condillac poses the question in a similar manner:
It seems that we would not know how to use signs of institution if we were not already capa-
ble of sufficient reflection to choose them and to attach ideas to them. How then, someone
will perhaps object, can the exercise of reflection be acquired only by the use ofsigns?71
In other words, if language as a social institution is something created by human beings,

it would seem as though the first men who created language must have already had a lan-
guage in order to be able to create language. It is this problem that Condillac intends to
solve in his 'history of language'.
Echoes of this sort of difficulty can be heard in many comers of contemporary phi-
losophy of language, such as in the discussion of 'solitary language'. Chomsky's idea of
an innate linguistic capacity and Fodor's 'mentalese' are two ways of responding to a re-
lated problem; Michael Dummett's notion of a 'full-blooded theory of meaning' through
which one could learn a language from scratch gives rise to similar difficulties. And if
one takes what is given in Quine's naturalistic perspective as 'the natural state', it is dif-
ficult to understand how one can gain access into a 'radically foreign language'. All lan-
guages tend to be 'radically foreign' in 'the natural state'.
The experience of the immanence of language, the sense that, in some respect, we are
shut up in language, is, I think, a kind of shared metaphysical feeling behind various
philosophical ways of reacting to these problems. One finds a very different attitude in
the work of the later Wittgenstein: we can only approach the bounds of language from
inside, and when you feel that to be restrictive and disturbing, it is because your concep-
tion of language is far too limited.
An increased interest in the findings of anthropological studies complicated the for-
mulation of the problem in the Berlin Academy discussion. There were those who called
into question the idea of a state of nature as a far too illusory and unrealistic basis for ex-
planation. Rousseau asks what sort of experiment would be necessary in order to gain
knowledge of 'natural man', and how such an experiment could be performed in society.
The secretary of the Berlin Academy, Samuel Formey, questioned the idea of a natural
70 Aarsleff, pp. 194-195.

46 SOREN STENLUND
state by means of a thought experiment of the kind often discussed in connection with
the problem of the origin of language. Formey imagined an experiment that would have
to be carried out over two generations. A dozen children of the same age would be pro-
vided everything they need in terms of food, clothing, care and all other necessities, but
nobody would be permitted to speak to them or to utter a word in their presence. They
are not allowed to come into contact with any achievements in art, technology or sci-
ence. They are not even allowed to see how their food is prepared. What will happen to
them when they grow up? They will presumably have children, but will the mothers
know how to care for their children and feed them, when they themselves have never
witnessed anything of the kind? If these natural families do not even survive over two
generations, what is the point, Formey objects, of asking when they create language?
The first human beings, furthermore, did not even have the benefit of being provided for
as did the children in the thought experiment. Formey's conclusion is that the idea of a
pure state of nature is an illusion, an absurdity, which strengthens his faith in a Divine
Creator. Formey sees a return to the theological perspective of the past as the only way
out.
Today, more than two hundred years after that discussion, one can wonder about the
'origin' of the idea of the origin of language. What kind of notions of language led to the
idea of a pre-linguistic natural state? What was it that provoked the idea that there 'must
have' existed such a state, in spite of the difficulties involved in making sense of that
idea? There were surely many factors, such as the older biologistic philosophical view
stemming from Aristotle, but I would claim that the most decisive factor was the linguis-
tic conception of language, i.e., the concept of language found in traditional grammar.
Book III of Locke's Essay, in which he presents his philosophy of language, is enti-
tled, "Of Words"; it is evident that he takes for granted the traditional grammatical no-
tions of words and language. In the first paragraph, he writes: "Man [] had by nature his
organs so fashioned as to be fit to frame articulate sounds, which we call words.,,72
What is primary in Locke's view of language are words, understood as articulate sounds.
Language as speaking and saying is secondary. In this regard, Condillac and the tradition
of linguistics and the philosophy of language follow in Locke's footsteps, well into our
century. In this respect, Leibniz and the advocates of the Adamic teaching are no differ-
ent.
The idea of a pre-linguistic state of nature was, I believe, a consequence of seeing the
concepts of grammar, not merely as a set of methods, terms, didactic devices, and instru-
ments for philological studies, textual interpretation, translation, etc., but as norms for
what human language as a whole really is, how it is constituted, organized, etc. This
norm included even the language that human beings already have when they learn gram-
mar, when they learn about 'words', 'sentences', and 'sentence structure'. One may sur-
mise, however, that most of Locke's contemporaries never learned grammar. They were
unfamiliar with the concept of 'word' as Locke uses it, although they could speak and
72 The emphasis is not mine.

participate in 'vulgar discourse'. The linguistic conception of language is something that

most people acquire when they learn to read and write and when they study foreign lan-
guages, and in so doing, they also learn to regard their mother tongue from a grammati-
cal perspective.
This conception of language is clearly 'man-made'. It is a cultural product, and if one
takes it to be language in its entirety, the problem arises as to how language could in
principle have begun. If language is created, by what means was it created? Must one not
already have a language in order to be able to create language? CondiIlac's 'natural
signs' were intended to serve that purpose.
If one takes the grammatical conception of language, not as a doctrine of the actual
constitution of language in general, but as a set of techniques and aids for the teaching
and study of language, one could say that the natural state is the one in which most chil-
dren find themselves when they first learn to read and write or study foreign languages:
they already have a language, they speak, they say things, and so forth. The language
that they already possess is then built up, broadened, and refined, but never replaced by
language education.
In the way in which we usually talk about human beings, one could say that human
beings already have a language in 'the natural state'. A human being is a human being,
so to speak, already 'from the beginning'. A newborn infant is a being who, among other
things, starts to speak a couple of years after his or her birth (notice that we do not say
"learns to speak English" of a child born in an English-speaking environment). A text-
book on 'child development' takes it to be self-evident that learning language is an es-
sential part of human development; one does not say, "here we will limit ourselves to
that period in human history in which human beings possess language." The biological
sciences perhaps have another concept of 'human being', but what is important in this
context is that the philosophical idea of a pre-linguistic natural man is an idea at which
one arrives when one sees grammar as the norm for what language is at bottom.
Conclusion
Finally, I would like to return to what I said at the beginning about certain features in the
internal historiography of philosophy. Reading the historical summaries written within
the analytic tradition, one can almost get the impression that the philosophy of language,
philosophical thinking about the nature of language, is an area of inquiry that arises
around the tum of the century with the work of Frege, Russell's "On Denoting", etc.
What had been said earlier with regard to philosophical questions concerning language is
presented as anecdotes and historical curiosities, of interest only insofar as they antici-
pate, however primitively, some discussion in contemporary philosophy of language.
One explanation for this feature of philosophical historiography, I think, is the domi-
nance of questions related to the philosophy of science within analytic philosophy up un-
til the 1970's. I count those questions concerned with 'the foundations of mathematics',
which were of such great importance for early analytic philosophy, to the set of ques-
48 SOREN STENLUND
tions that belong here. These questions gave rise to the entire technical conceptual arse-
nal of mathematical logic, the theory of formal languages, logical syntax and semantics.
generative grammar, etc., which have been normative for the philosophy of language and
the canonical view of language within the analytic tradition. Questions in the philosophy
of language have been interesting only insofar as they have in some way been relevant
from the perspective of the philosophy of science, and the sciences of our day are the
special sciences, the university sciences.
It is first in the nineteenth century that the natural sciences come to be firmly organ-
ized as professionalized university sciences, and begin to develop into the institutions
that, in our time, have had the last word on Truth. 73 And for this reason, questions in the
philosophy of science come to dominate within philosophy. There certainly were excep-
tions to this picture (G.E. Moore comes to mind), but on the whole, I would say that
these features have dominated, although on an unconscious level in younger generations
of analytic philosophers. Unfortunately, this attitude has hindered our understanding of
earlier discussions in the philosophy of language.
I would like to propose the possibility of writing an entirely new history of philoso-
phy, one that takes philosophical thinking on the nature of language during different ep-
ochs as its unifying perspective. It would distinguish itself from standard surveys in the
history of philosophy with respect to which names would be included as having made
the most significant contributions. Kant would receive relatively little attention, while
Locke and Condillac, for example, as well as thinkers whose works follow in their tracks
(such as Saussure) would require close study. The presentation would not retain the tra-
ditional classificatory apparatus of the history of philosophy. It is doubtful, for example,
that there would be much point in dividing philosophers up into idealists, rationalists and
empiricists in the historiography I am imagining here. (Was Locke not as much a ration-
alist as Descartes?) It would not limit itself exclusively to questions in what we now call
'themes in the philosophy of language', since thinking about the nature of language in
other epochs was always part of the treatment of other philosophical problems (which
we now classify as metaphysical, epistemological, or moral). This was still the case, as I
mentioned earlier, for Frege, whose philosophy of language was part of his treatment of
problems about the foundations of mathematics.
The idea of philosophical issues belonging to pure philosophy of language and being
disengaged from other philosophical problems, I would suggest, is a relatively recent
73 Here is a case where historical and sociological fact~ are directly relevant to a conceptual question;
namely, for the concept of science that has been taken for granted for the greater part of the twentieth cen-
tury in the philosophy of science and philosophical historiography. There is a clear tendency to depict sev-
enteenth- and eighteenth-century science, if not as though it actually was university science in our sense,
at least as if it tried to be, or ought to have been. One writes as if this were the only way in which it would
be possible to take it seriously. The spirit and philosophical aims of Newton's Principia, for instance, are
hardly consistent with the spirit animating university science in our day. In a sense, it seems fair to say that
seventeenth- and eighteenth-century science was more humanist, and far less technologically and eco-
nomically oriented, than the natural sciences today, where it is difficult to delimit clearly where science
ends and high technology begins.
twentieth-century phenomenon. I would be inclined to classify that genre as belonging to

the philosophy of the science of modern theory of language and general linguistics. The
subject matter of 'pure philosophy of language' is different forms of linguistic theory
(such as the construction of theories of meaning). The subject matter is not primarily the
language actually spoken by people in general. What is discussed and studied is different
theoretical representations of language, although these are discussed as if they were lan-
guage itself.
An interesting point in this vision of a new history of philosophy would be that
Wiugenstein's philosophy (and presumably Heidegger's) would appear far less original
and unique than it does when seen in the light of the standard internal philosophical his-
toriography, at least when it comes to the kind of problem that he takes up. Some of
Wingenstein's ideas would no longer seem as if they simply came unbidden from a soli-
tary, spontaneous act of creation on the part of a dedicated genius.
I have already pointed out that Wittgenstein is far from the first to have emphasized
how we are led into philosophical confusion by language (even if Wittgenstein gives this
idea a new and radical content). I have noted a similarity between Wittgenstein and Des-
cartes, and I have indicated one parallel between Wittgenstein's Tractatus and Saus-
sure's Cours, and could name more. The concept of a sign and the theory of the symbol
found in the Tractatus are clearly more closely related to what we find in the Cours
(which, in turn, is in all likelihood inspired on this point by Taine, who in tum alludes to
Condillac). In the analytic secondary literature on the Tractatus, one hardly notices this
theory of the symbol at all, since it does not fit with the concept of a sign and the notion
of syntax canonized by Carnap and his followers. The Tractatus' use of the term 'syn-
tax' has more in common with its use in the works of the linguist 0tto Jespersen, for in-
stance. It is also worth noting that both Saussure and Wittgenstein made use of the
analogy of a game of chess to make essentially the same point.
The extremely persistent nomenclaturist view of language, which is still to be found
in the writings of the early Russell, and which Wittgenstein portrays as a source of nu-
merous problems 74 is essentially also what Locke identifies as a source of problems in
the Adamic teaching. Furthermore, one can add that the later Wittgenstein, like Locke,
turns away from the respective forms of the code conception of language of their con-
temporaries. But Wittgenstein was not satisfied with the solution offered by Locke, of
course, and especially not with the mentalism and subjectivism it entails. But one thing
that can be read out of Wittgenstein's so-called private language argument is that he ap-
preciates the reactions to the difficulties posed by Locke's philosophy of language. Re-
garding this feature of Locke's philosophy, Aarsleff writes:
Owing to the impenetrable subjectivity of ideas to which words are tied, each individual has
a radically private language that virtually precludes all hope of perfect communication, a
principle that was to reverberate through Condillac, Herder, Turgot, Destutt de Tracy and
Humboldt and that also became Wordsworth's problem. Goethe made the same observation
74 See, for example, the quotation from Augustine in the first paragraph of the PhilosophicailnveSlillalions.
50 SOREN STENLUND
in DichlUng und Wahrheit, in a passage wrillen in 1813: "For that no man understands an-
other, that a dialogue or a text causes different trains of thought in different individuals, was
something I had long realized all too clearly."
If Wittgenstein was acquainted with this problem in seventeenth- and eighteenth-century

thought is uncertain (even if we know that he had read Goethe). It does seem probable,
however, that he had at least come in contact with the discussion, if only indirectly.
Whatever the case may be, it seems reasonable to connect that discussion with Wittgen-
stein's private language argument.
A further example is Wittgenstein's notion of 'primitive reactions' and the idea that
certain forms of language are elaborations upon or additions to such primitive reactions.
These thoughts are related to Condillac's ideas about natural signs and the role that these
play in his explanation of the development of language. One could find several parallels
between philosophical questions discussed during the second half of the eighteenth cen-
tury as questions concerning the origin of language and the development of our faculties
for knowledge, and those questions which Wittgenstein answers by posing the question
of how we learn certain forms of language.
I do not wish to claim that these examples constitute indications of direct influence
(even if that is perhaps not out of the question, in some cases). What is interesting for me
is the striking similarity in the kinds of problem posed (if not in the way of tackling
them). There is a kinship in the posing of the questions that is clarifying on both sides,
and which makes the problems more vivid in the manner one expects from comparisons
in good philosophical historiography.75
75 Earlier versions (in Swedish) of this paper have been read by Olav Gundersen. Martin Gustafsson, Mi-
chael Gustavsson. Saara Haapamaki, Mats Persson. Par Segerdahl and Sven Ohman. They have given me
valuable comment~ that have effected the final form of the essay. The article is based on research made
possible by a grant from The National Bank of Sweden's Centenary Fund in support of the project Lan-
guage and Human Action. The translation into English was done by Sharon P. Rider, with financial sup-
port from the Anders Karitz Foundation.

On The Linguistic Turn in Philosophy - Stenlund2002 PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

On The Linguistic Turn in Philosophy - Stenlund2002 PDF

Uploaded by

Copyright:

Available Formats

SOREN STENLUND

2 Wedberg. A .• A History of Philosophy. Volume 3: From Bolzam) to Wittgenstein, Oxford: Clarendon

The Reaction against Psychologism

The Idea of Language as the Origin of Philosophical Problems

in orientation is that language came to be seen as a source of philosophical problems;

immediacy of mental vision. Thus it is not a question of finding universal hypotheses or

And Hume writes:

25 Cited in Aarsleff. p. 159.

28 Frege. O. Posthumous Writings. Oxford: Blackwell. 1979. p. 127.

The Immanence of Language

30 Cited in Aarsleff, p. 154.

The Philosophical Discourse of the Mental

Seventeenth and eighteenth-century philosophers express themselves in a relatively com-

39 Posthumous Writings, p. 269.

A Common Human Nature

science, is Chomsky's theory of language, in which the idea of a biologically grounded

The Adamic Doctrine of Language

The background to Locke's philosophy of language is still inadequately described if one

Naturally, one could say that seventeenth-century philosophers used etymological

The Question of the Origin of Language

In other words, if language as a social institution is something created by human beings,

70 Aarsleff, pp. 194-195.

72 The emphasis is not mine.

participate in 'vulgar discourse'. The linguistic conception of language is something that

twentieth-century phenomenon. I would be inclined to classify that genre as belonging to

If Wittgenstein was acquainted with this problem in seventeenth- and eighteenth-century

You might also like