through the intermediary of a language which furnishes not only the phonetic expression of
the characters, but also the formulation of axioms permitting the determination of the value of
these characters. It is true that at a pinch one could decipher unknown characters, but that
always supposes an acquired knowledge, a thought already formed by the usage of speech.
Therefore, in all hypothesis, mathematical symbol-ism is the fruit of a secondary elaboration,
supposing preliminarily the usage of discourse and the possibility of conceiving explicit
conventions. It is nevertheless true that mathematical algorithm will express the formal laws
of symbolization, of syntactic structures, independent of particular means of expression.” On
these problems, cf. also Gilles Gaston Granger, Pensée formelle et sciences de l’homme
(Paris, 196o), pp. 38 f. and particularly pp. 43 and 50 f. (on the “Reversal of Relationships
between the Spoken Language and Writing”) .
2.All works on the history of writing devote space to the problem of the introduction of
phonetic writing in the cultures that did not practice it previously. Cf. e.g., EP, pp. 44 f. or « La
reforme de l’écriture chinoise, » Linguistique, Recherches internationales d la lumière du
marxisme 7 (May—June 1958).
3.Here I do not merely mean those “theological prejudices” which, at an identifiable time and
place, inflected or repressed the theory of the written sign in the seventeenth and eighteenth
centuries. I shall speak of them later in connection with Madeleine V.-David’s book. These
prejudices are nothing but the most clearsighted and best circumscribed, historically
determined manifestation of a constitutive and permanent presupposition essential to the
history of the West, therefore to metaphysics in its entirety, even when it professes to be
atheist.
4.
Grammatology: “A treatise upon Letters, upon the alphabet, syllabation, reading, and
writing,” Littré. To my knowledge and in our time, this word has only been used by I. J. Gelb
to designate the project of a modern science in A Study of Writing: The Foundations of
Grammatology [Chicago], 1952 (the subtitle disappears in the 1963 edition). In spite of a
concern for systematic or simplified classification, and in spite of the controversial hypotheses
on the monogenesis or polygenesis of scripts, this book follows the classical model of
histories of writing.
Part I: Chapter 1
1. To speak of a primary writing here does not amount to affirming a chronological priority of
fact. That debate is well-known; is writing, as affirmed, for example, by
((324))
Metchaninov and Marr, then Loukotka, “anterior to phonetic language?” (A conclusion
assumed by the first edition of the Great Soviet Encyclopedia, later contradicted by Stalin. On
this debate, cf. V. Istrine, « Langue et écriture, » Linguistique, op. cit., pp. 35, 6o. This debate
also forms around the theses advanced by P. van Ginneken. On the discussion of these
propositions, cf. James Février,
Histoire de l’écriture [Payot, 1948—59], pp. 5 f.) . I shall try
to show below why the terms and premises of such a debate are suspicious.
2.I shall deal with this problem more directly in La voix et le phénomène (Paris, 1967)
[Speech and Phenomena, op. cit.].
3.Wiener, for example, while abandoning “semantics,” and the opposition, judged by him as
too crude and too general, between animate and inanimate etc., nevertheless continues to use
expressions like “organs of sense,” “motor organs,” etc. to qualify the parts of the machine.
4.Cf., e.g., EP, pp. 126, 148, 355, etc. From another point of view, cf. Roman Jakobson,
Essais de linguistique générale (tr. fr. [Nicolas Ruwet, Paris, 1963], p. 116) [Jakobson and
Morris Halle, Fundamentals of Language (the Hague, 1956), p. 16].
5.This is shown by Pierre Aubenque (Le problème de l’être chez Aristotle [Paris, 1966], pp.
106 f.) . In the course of a provocative analysis, to which I am here indebted, Aubenque
remarks: “In other texts, to be sure, Aristotle designates as symbol the relationship between
language and things: ‘It is not possible to bring the things themselves to the discussion, but,
instead of things, we can use their names as symbols.’ The intermediary constituted by the
mental experience is here suppressed or at least neglected, but this suppression is legitimate,
since, mental experiences behaving like things, things can be substitued for them immediately.
On the other hand, one cannot by any means substitute names for things” (pp. 107—08) .
6.Roman Jakobson, Essais de linguistique générale, tr. fr., p. 162 [”The Phonemic and
Grammatical Aspects of Language in their Interrelations,” Proceedings of the Sixth
International Congress of Linguistics (Paris, 1949), p. 6]. On this problem, on the tradition of
the concept of the sign, and on the originality of Saussure’s contribution within this
continuity, cf. Ortigues, op. cit., pp. 54 f.
7.Cited by Emmanuel Levinas, in Difficile liberté [Paris, 1963], p. 44.
8.I attempt to develop this theme elsewhere (Speech and Phenomena).
9.This does not, by simple inversion, mean that the signifier is fundamental or primary. The
“primacy” or “priority” of the signifier would be an expression untenable and absurd to
formulate illogically within the very logic that it would legitimately destroy. The signifier will
never by rights precede the signified, in which case it would no longer be a signifier and the
“signifying” signifier would no longer have a possible signified. The thought that is
announced in this impossible formula without being successfully contained therein should
therefore be stated in another way; it will clearly be impossible to do so without putting the
very idea of the sign into suspicion, the “sign-of” which will always remain attached to what
is here put in question. At the limit therefore, that thought would destroy the entire
conceptuality organized around the concept of the sign (signifier and signified, expression and
content, and so on) .
10.Postface to Was ist Metaphysik? [Frankfurt am Main, 1960], p. 46. The insistence of the
voice also dominates the analysis of Gewissen [conscience] in Sein und Zeit (pp. 267 f.) .[pp.
312 f.].