All this amounts to saying that the principle of interpretation (in its Peircean sense) has not to be identified with the farfetched assumption that— as Valery said—il n’y a pas de vrai sens dun texte.
When considering contemporary theories of interpretation (especially in the literary domain), we can conceive of a range with two extremes x and y. (I refuse to represent it spatially as a line going from left to right, so as not to suggest unfair and misleading ideological connotations.) Let us say that at the extreme x stand those who assume that every text (be it a conversational utterance or a poem) can be interpreted in one, and only one, way, according to the intention of its author. At the extreme у stand those who assume that a text supports every interpretation—albeit I suppose that nobody would literally endorse such a claim, except perhaps a visionary devotee of the Kabalistic temura.
I do not think that the Peircean notion of semiosis should privilege one of these extremes. At most, it provides a theoretical tool for identify-ing, according to different semiosic processes, a continuum of inter-mediate positions. If I ask someone what time it is and if he answers /6:15/, my interpretation of this expression can conclude that (provided there are no other co-textual clues and provided the speaker is not a notorious liar or a psychotic subject) the speaker positively said that it is forty-five minutes to seven and that he intended to say so.
On the other hand, the notion of interpretation can explain both in which sense a given text displays two and no more possibilities of dis-ambiguation and why an instance of the symbolic mode requests an indefinite series of alternative or complementary interpretations. In any case, between x and у stands a recorded thesaurus of encyclopedic competence, a social storage of world knowledge, and on these grounds, and only on these grounds, any interpretation can be both implemented and legitimated — even in the case of the most ‘open’ instances of the option y.
0.3.
In order to discuss these points, all the chapters of this book, while examining a series of fundamental concepts traditionally related to the one of sign, revisit each of them from a historical point of view, looking backward at the moment they were posited for the first time and were endowed with a theoretical fecundity that sometimes they have lost in the course of a millenary debate.
It is clear from the index that most of my authors are not linguists or full-time semioticians, but philosophers who have speculated about signs. This is not solely due to the fact that I started my academic career as a philosopher, particularly interested in the Middle Ages, and that since the Second Congress of the IASS (Vienna, 1979) I have advocated a revisitation of the whole history of philosophy (as well as of other dis-ciplines) to take back the origins of semiotic concepts. This is not (or not only) a book in which a semiotician pays a visit, extra moenia, to the alien territory of philosophy. This is a book on philosophy of language for the very simple reason that a general semiotics is nothing else but a philoso-phy of language and that the ‘good’ philosophies of language, from Craty/us to Philosophical Investigations у are concerned with all the semiotic questions.
It is rather difficult to provide a ‘catholic’ definition of philosophy of language. In a nondogmatic overview, one should list under this heading Plato’s discussions on nomos and phusis, Aristotle’s assumption that /Be-ing/ is used in various senses, Russell’s theory of denotation, as well as Heidegger, Cassirer, and Merleau-Ponty. I am not sure that a general semiotics can answer all the questions raised during the last two thousand years by the various philosophies of language; but I am sure that all the questions a general semiotics deals with have been posited in the framework of some philosophy of language.
0.4.
In order to make this point clear, one must distinguish between specific semiotics and general semiotics. I understand that this is a very crude dis-tinction as compared with more subtle classifications. I am thinking of Hjelmslev’s proposal according to which there are a scientific semiotic and a nonscientific semiotic, both studied by a metasemiotic; a semiology as a metasemiotic studying a nonscientific semiotic, whose terminology is studied by a metasemiology. Since semiotics can be either denotative or connota-tive, there is also a meta (connotative) semiotic. Pelc (1981) has outlined a far more analytical classification of the many levels of a semiotic study. At the present state of the art, I am inclined to take these and other distinctions as fruitfully descriptive, while I am not sure that they can be taken as normative. In any case, for the purposes of the present dis-course, I think it will be sufficient to work upon the distinction between general and specific.
A specific semiotics is, or aims at being, the ‘grammar’ of a particular sign system, and proves to be successful insofar as it describes a given field of communicative phenomena as ruled by a system of signification. Thus there are ‘grammars’ of the American Sign Language, of traffic signals, of a playing-card ‘matrix’ for different games or of a particular game (for instance, poker). These systems can be studied from a syntac-tic, a semantic, or a pragmatic point of view. Sometimes a specific semiotics only focuses on a particular subsystem (or s-code, as defined in Eco 1976) that works within a more complex system of systems: such is the case of the theory of phonemic distinctive features or of the descrip-tion of the phonemic oppositions holding for a given verbal language.
Every specific semiotics (as every science) is concerned with general epistemological problems. It has to posit its own theoretical object, ac-cording to criteria of pertinence, in order to account for an otherwise disordered field of empirical data; and the researcher must be aware of the underlying philosophical assumptions that influence its choice and its criteria for relevance. Like every science, even a specific semiotics ought to take into account a sort of ‘uncertainty principle’ (as an-thropologists must be aware of the fact that their presence as observers can disturb the normal course of the behavioral phenomena they ob-serve).
Notwithstanding, a specific semiotics can aspire to a ‘scientific’ status. Specific semiotics study phenomena that are reasonably indepen-dent of their observations. Their objects are usually ‘stable’ —even though the duration of a code for traffic signals has a shorter range than the duration of a phonological system, whereas lexical systems are in a continuous process of transformation. Being scientific, a specific semio-tics can have a predictive power: it can tell which expressions, produced according to the rules of a given system of signification, are acceptable or ‘grammatical’ and which ones a user of the system would presumably produce in a given situation.
Obviously, there are different degrees of scientificity, according to the rigidity or the flexibility of the sign system in question. The ‘grammar’ of traffic lights and the structure of a phonological system seem to be more ‘objective’ (more ‘scientific’) than the description of the narrative function in Russian fairy tales; and the narrative function of the Russian fairy tales seems to be less questionable than, let us say, a possible sys-tem of narrative function in the novels of French Romanticism. Not every specific semiotics can claim to be like a natural science. In fact, every specific semiotics is at most a human science, and everybody knows how controversial such a notion still is. However, when cultural anthropology studies the kinship system in a certain society, it works upon a rather stable field of phenomena, can produce a theoretical ob-ject, and can make some prediction about the behavior of the members of this society. The same happens with a lexical analysis of the system of terms expressing kinship in the same society.
In this sense, a specific semiotics (as any other science) can also have effects in terms of social engineering. When the anthropologist increases our knowledge of a given society, his or her descriptions can be used for ‘missionary’ purposes in order to improve, to preserve, or to destroy a given culture, or to exploit its members. It goes without saying that the natural sciences have engineering purposes, not only in the strict technological sense; a good knowledge of human anatomy also can help one to improve one’s physical fitness. In the same way, the description of the internal logic of road signals can suggest to some public agency how to improve the practice of road signaling. Such an engineering power is the result of a free decision, not an automatic side effect of the scientific research.
All around this area of more or less established and rigorous ‘grammatical’ knowledge is a hardly definable ‘twilight zone’ of semioti-cally oriented practices, such as the application of semiotic notions to literary criticism, the analysis of political discourses, perhaps a great part of the so-called linguistic philosophy when it attempts «to solve philo-sophical problems by analyzing the meanings of words, and by analyzing logical relations between words
