The comparison with Vološinov demonstrates a difficulty of accounting for linguistic difference and strangeness from a Gramscian point of view. But Gramsci's anti-Nietzscheanism is ambiguous enough to avoid oversimplification. We can now try to supplement the struggle for unification with a struggle for dispersion. This raises the question of how to construct a methodological strategy in linguistics (and elsewhere) that incorporates the efforts made in language towards establishing 'coherence/dispersion', conceived as a unitary concept (comparable to Saussure's langage). Is there still something to be learned from Gramsci at this stage?
   Gramsci sometimes hints that he is not so sure of the necessity of a will to coherence as he appears to be. Such a will presupposes the possibility of translations between languages, considered as the necessary constituent elements of the same world conception. The unity of the social milieu ('contact') is a condition for such a translation to be possible (cf. Vološinov (1973: 46-47)). It is a premise; therefore, it is not necessarily fulfilled! That is to say: the constituent (and still dispersed) elements of a possible unity must necessarily be reconstitutable from each of the partial activities that contribute to this possible unity to the others if and only if we take as our starting-point that there are reasons to strive for such a unity. But this 'unifying' principle is "to be gone into in depth and stated in more exact terms", it is "still in need of elaboration" ((PN 403); cf. (CW 136-140)). [12] [p. 563:]
   To illustrate the 'principle of translatability', a concrete illustration (due to Gramsci himself) may be helpful. The points of the compass: North, South, East, and West, are real and yet would not exist outside civilization (since outside real history every point on the earth is North and South, East and West at the same time). "And yet these references are real; they correspond to real facts, they allow one to travel by land and by sea, to arrive where one has decided to arrive, to 'foresee' the future, to objectivise reality, to understand the objectivity of the external world. Rational and real become one" (PN 447-448). The epistemological concepts of rationality and reality thus become historical concepts. Subjective and objective knowledge become one in 'organic knowledge'; they depend on the practical relations in which there is agreement in forms of life (in how the points of the compass have to be followed, in what it is to 'arrive where one has decided to arrive', etc.). There is no problem of knowledge, reference, or identity, if and only if such an agreement can be safely presupposed.
   But there is not necessarily a historical need to meet this premise. It is not self-evident that agreement can be reached in some, or even most, cases. There is no a priori reason why terms such as East and West could not become part of an antagonism. "Japan is the Far East not only for Europe but also perhaps for the American from California and even for the Japanese himself, who, through English political culture, may call Egypt the Near East" (PN 447). In order to escape from the cultural hegemony of the 'Western' world, it may well be worth while to put such terms out of bounds, and to raise the problem of knowledge here. One may even feel an urge to dismiss the whole idea of a 'standard language', be it one that is as wide-spread as English, or the formal languages of logic and computers, and prefer to speak a dialect, a national or ideological language of a smaller group (cf. Steinberg (1987)), or even Esperanto! This would be a revival of charlatanism, but is it really excluded by Gramsci? I shall not deny that Gramsci rejects such a deviation from the main roads of history. "But", as he says, "this is a question of style, not 'theory'." (NP 369).
   Still, the above does not change the fact that "great importance is assumed by the general question of language", as I quoted as the motto of this article. Only, it has to be added that the question of language should also be faced in order to allow for differentiation between the answers to questions of social identity and meaning. The 'linguistic' dilemma of unity versus dispersion can only be faced by making reference to very different historical directions of the 'will' ('style', 'ideology', 'form of life'). This makes room for a question that [p. 564:] Gramsci, writing before the identity crises of (post-) modern thought, did not need to ask. However, not asking it at the present time, precisely from a Gramscian point of view, would be to restrict our field of interest severely. The importance of the various historical struggles for coherence (on which Gramsci provides us with useful insights) should not blind us to the struggles for dispersion, the breaking away from oppressors, or the new starts that do not possess the perspective of a totalizing and all-absorbing unity. If (and only if) it is possible to develop different styles, both of unity and of dispersion, and if these can (but need not) be considered as dealing with the 'same' social world, the question can be faced of how to establish a reciprocal translation between the specific languages of those styles. How can one convert the theoretical principles of one style to those of another one? This question remains, for the moment, an academic one; it could only be answered if there were, at this moment in history, a sufficiently unifying will in the field of the social (a presupposition that seems highly questionable).
   This means that interpretative analyses, for everyone working within the theory and practice of signs (i.e. linguists and philosophers, in a broad sense), should incorporate a practical point of view that firstly, makes interpretation an ongoing and never-completed activity, and secondly, does not presuppose a will for agreement, communication, or co-operation. One has to find out how (theoretical and ordinary) language distinguishes and unites social groups, often in one and the same move. Linguists study, often on too abstract a level, the way social groups are 'articulated' by language practices. Alternatively, they could take up organizing/dispersing practices, struggles for schism and alliance, as part of the way in which human individuals form their lives (their identities/differences) through language. In this way, linguistics could become (more fundamentally than is sometimes supposed) a social, rather than a 'natural' science, which would be a definite asset in a world where nothing is 'natural' anymore.