>The manuscript partially completed during these years and published posthumously as The Prose of the World (1969/1973) pursues these themes through a phenomenological investigation of literary language and its relationship with scientific language and painting. Critiquing our commonsense ideal of a pure language that would transparently encode pre-existing thoughts, Merleau-Ponty argues that instituted language—the conventional system of language as an established set of meanings and rules—is derivative from a more primordial function of language as genuinely creative, expressive, and communicative. Here he draws two insights from Saussurian linguistics: First, signs function diacritically, through their lateral relations and differentiations, rather than through a one-to-one correspondence with a conventionally established meaning. Ultimately, signification happens through the differences between terms in a referential system that lacks any fixed or positive terms. This insight into diacritical difference will later prove important to Merleau-Ponty’s understanding of perception and ontology as well (see Alloa 2013a). Second, the ultimate context for the operation of language is effective communication with others, by which new thoughts can be expressed and meanings shared. Expression accomplishes itself through a coherent reorganization of the relationships between acquired signs that must teach itself to the reader or listener, and which may afterwards again sediment into a taken-for-granted institutional structure.
>Merleau-Ponty distinguishes between primary and secondary modes of expression. This distinction appears in Phenomenology of Perception (p. 207, 2nd note Fr. ed.) and is sometimes repeated in terms of spoken and speaking language (le langage parlé et le langage parlant) (The Prose of the World, p. 10). Spoken language (le langage parlé), or secondary expression, returns to our linguistic baggage, to the cultural heritage that we have acquired, as well as the brute mass of relationships between signs and significations. Speaking language (le langage parlant), or primary expression, such as it is, is language in the production of a sense, language at the advent of a thought, at the moment where it makes itself an advent of sense.
>
> It is speaking language, that is to say, primary expression, that interests Merleau-Ponty and which keeps his attention through his treatment of the nature of production and the reception of expressions, a subject which also overlaps with an analysis of action, of intentionality, of perception, as well as the links between freedom and external conditions.
>
> The notion of style occupies an important place in "Indirect Language and the Voices of Silence". In spite of certain similarities with André Malraux, Merleau-Ponty distinguishes himself from Malraux in respect to three conceptions of style, the last of which is employed in Malraux's The Voices of Silence. Merleau-Ponty remarks that in this work "style" is sometimes used by Malraux in a highly subjective sense, understood as a projection of the artist's individuality. Sometimes it is used, on the contrary, in a very metaphysical sense (in Merleau-Ponty's opinion, a mystical sense), in which style is connected with a conception of an "über-artist" expressing "the Spirit of Painting". Finally, it sometimes is reduced to simply designating a categorization of an artistic school or movement. (However, this account of Malraux's notion of style—a key element in his thinking—is open to serious question.31)
>
> For Merleau-Ponty, it is these uses of the notion of style that lead Malraux to postulate a cleavage between the objectivity of Italian Renaissance painting and the subjectivity of painting in his own time, a conclusion that Merleau-Ponty disputes. According to Merleau-Ponty, it is important to consider the heart of this problematic, by recognizing that style is first of all a demand owed to the primacy of perception, which also implies taking into consideration the dimensions of historicity and intersubjectivity. (However, Merleau-Ponty's reading of Malraux has been questioned in a recent major study of Malraux's theory of art which argues that Merleau-Ponty seriously misunderstood Malraux.)32 For Merleau-Ponty, style is born of the interaction between two or more fields of being. Rather than being exclusive to individual human consciousness, consciousness is born of the pre-conscious style of the world, of Nature.
メルロ=ポンティは、表現の一次様式と二次様式を区別している。この区別は『知覚の現象学』(207頁、第2注[仏語版])に現れ、話し言葉と話し言葉(le langage parlé et le langage parlant)の言葉で繰り返されることもある(『世界の散文』10頁)。話す言語(le langage parlé)、つまり二次的表現は、私たちの言語的荷物、私たちが獲得した文化的遺産、そして記号と意義の間の関係のブルートマスに立ち戻るのである。話す言語(le langage parlant)、つまり一次表現は、そのようなものであり、感覚の生産における言語、思考の到来における言語、それ自体が感覚の到来となる瞬間における言語である。
>Thus expression involves nothing more than replacing a perception or an idea with a conventional sign that announces, evokes, or abridges it. Of course, language contains more than just ready-made phrases and can refer to what has never yet been seen. But how could language achieve this if what is new were not composed of old elements already experienced—that is, if new relations were not entirely definable through the vocabulary and syntactical relations of the conventional language?
>The word possesses no virtue of its own; there is no power hidden in it. It is a pure sign standing for a pure signification. The person speaking is coding his thought. He replaces his thought with a visible or sonorous pattern which is nothing but sounds in the air or ink spots on the paper. Thought understands itself and is self-sufficient. Thought signifies outside itself through a message which does not carry it and conveys it unequivocally only to another mind, which can read the message because it attaches the same signification to the same sign, whether by habit, by human conventions, or by divine institution. In any case, we never find among other people's words any that we have not put there ourselves. Communication is an appearance; it never brings us anything truly new.
>How could communication possibly carry us beyond our own powers of reflection, since the signs communication employs could never tell us anything unless we already grasped the signification?
>It is frequently observed that it is impossible at any given moment to make an inventory of a vocabulary—whether of a child, of an individual, or of a language. Should one count as distinct words those that are formed mechanically from the same original word? Does one include a word that is still understood but hardly used and marginal to everyday use? like the visual field, an individual's linguistic field ends in vagueness, because speaking is not having at one's disposal a certain number of signs. Speaking is possessing language as a principle of distinction, whatever number of signs it permits us to specify. There are languages in which one cannot say "to sit in the sun," * because they use particular words to refer to the rays of sunlight and keep the word "sun" for the star itself. In other words, the linguistic value of each word is defined only through the presence or absence of the words surrounding it. Since one can say the same thing in turn of those words, it seems that language never says anything; it invents a series of gestures, which between them present differences clear enough for the conduct of language, to the degree that it repeats itself, recovers and affirms itself, and purveys to us the palpable flow and contours of a universe of meaning. Moreover, the words and very forms for an analysis of this kind soon appear to be secondary realities, the results of a more originary differentiation. The syllables and letters, the turns of phrase, and the word endings are the sediments of a primary differentiation which, this time, precedes without any doubt the relation of sign to signification, since it is what makes the very distinction between signs possible.
>
> The phonemes, too, which are the real foundations of speech, since they are reached through the analysis of spoken language and have no official existence in grammar and dictionaries, by themselves mean nothing one can specify. But for this very reason, they represent the originary form of signifying. They bring us into the presence of that primary operation, beneath institutionalized language, that creates the simultaneous possibility of significations and discrete signs. like language itself, phonemes constitute a system, in other words, they are less a finite number of tools than a typical manner of modulation, an inexhaustible power of differentiating one linguistic gesture from another. Finally, to the extent that the differences are more precise, more systematic, phonemes appear in situations which themselves are better articulated and suggest even more that the whole process obeys an internal order, the power of revealing to the child what the adult had in mind.