Despite being invaluable for the whole development of structural semantics, Hjelmslev’s proposal does not account for other kinds of signs in which it appears that the two functives are not analyzable further into figurae. If the cloud which announces the storm and the portrait of the Mona Lisa are to be taken as signs, there must be signs without expres-sion figurae, and perhaps without content figurae as well. Prieto (1966) has decidely widened the field of sign analysis by showing the existence of systems without articulation, and systems which have only a first ar-ticulation. The white stick of the blind —a positive presence which constitutes itself as pertinent against the absence of the stick, as a signifier without articulations — represents blindness in general, request-ing the right of way, postulating understanding on the part of bystanders. In short, it conveys a content nebula. As a system the stick is quite simple (presence vs. absence), but its communicational use is very complex. If the stick is not a sign, what is it, and what should it be called?
1.5.2. Signs vs. sentences
In the same years which saw Hjelmslev’s critique of the sign format as too broad, Buyssens maintained that the format of the sign was too minute. The semantic unit is not the sign, but something corresponding to the sentence, which Buyssens calls seme. The example given by Buys-sens concerns street signs as well as linguistic signs. He maintains that an arrow, isolated from the context of the street sign, does not allow for the concretization of a «state of consciousness.» In order to perform this function it will have to have a certain color, a certain orientation, and it will have to appear on a specific street sign, placed in a specific location. «The same thing happens with the isolated word, for instance, the word table. This word appeared as the potential member of different sentences in which different things are talked about» (Buyssens 1943:38).
Strange opposition: Hjelmslev is uninterested in the sign because he is interested in language as an abstract system; Buyssens is uninterested in the sign because he is interested in communication as a concrete act. Obviously, the opposition extension vs. intension is in the background of this debate. Unpleasant homonymy: componential semantics will call Hjelmslev’s content figurae (smaller than the sign) femes’, while-ffie tradition which developed from Buyssens (Prieto, De Mauro) will use the term ‘seme’ for utterances larger than the sign.
In any case, Buyssens’ seme is what others will call sentence or a performed speech act. What is surprising is the initial statement by Buyssens, according to which a sign does not have meaning. If it is true that nominantur singularia sed universalia significantur, one should rather say that the word table by itself does not name (it does not refer to) anything, but has a meaning, which Hjelmslev could have subdivided into figurae. Buyssens admits that this word (like the arrow) can be a potential member of different phrases. What is there, then, in the con-tent of table which allows it to enter expressions such as dinner is on the table or the table is made of wood, and not in expressions such as the table eats the fish or he washed his face with the dinner table? It must be agreed, then, that precisely because of its susceptibility to analysis by content figurae, the word table must include both atomic units of content and contextual instructions ruling over the word’s capacity to enter linguistic segments larger than the sign.
Prieto (1975:27) clarified this apparent disagreement between Hjelmslev and Buyssens by stating that the seme (for Buyssens) is a functional unit, whereas the figura is an economic unit. Hjelmslev postu-lated the sign as a functional unit and the figura as an economic unit. The problem is to identify, not two, but three or more levels, where the lower level is always constituted as the economic unit of the upper
level’s functional unit.
Buyssens’ distinction certainly anticipates, with its concreteness and complexity, all the theories opposing to the sign the speech act. How-ever, Plato and Aristotle, the Stoics and the Sophists had already talked about the differences existing between the meaning of words and the pragmatic nature of the question, the prayer, and the order. Those who oppose a pragmatics of discourse to a semantics of sign units shift the atten-tion from the systems of signification to the process of communication (Eco 1976); but the two perspectives are actually complementary. One cannot think of the sign without seeing it in some way characterized by its contextual destiny, but at the same time it is difficult to explain why a certain speech act is understood unless the nature of the signs which it contextualizes is explained.
1.5.3. The sign as difference
The elements of the signifier are set into a system of oppositions in which, as Saussure explained, there are only differences. The same thing happens with the signified. In the famous example given by Hjelmslev (1943:39), the difference in the content of two apparently synonymous terms, /Holz/ and /bois/, is given by the different segmen-tation of the continuum. The German /Holz/ encompasses everything which is not /Baum/ and is not /Wald/. The correlation between expression-plane and content-plane is also given by a difference: the sign-function exists by a dialectic of presence and absence, as a mutual exchange between two heterogeneities. Starting from this structural premise, one can dissolve the entire sign system into a net of fractures. The nature of the sign is to be found in the ‘wound’ or ‘opening’ or ‘divarication’ which constitutes it and annuls it at the same time.
This idea, although vigorously developed by poststructuralist thought, that of Derrida in particular, was actually developed much earlier. In the short text De organo sive arte cogitandi, Leibniz, searching for a restricted number of thoughts from whose combination all the others could be de-rived (as is the case with numbers), locates the essential combinational matrix in the opposition between God and nothing, presence and ab-sence. The binary system of calculation is the wondrous likeness of this dialectics.
From a metaphysical perspective, it may be fascinating to see every oppositional structure as based on a constitutive difference which dis-solves the different terms. Still, in order to conceptualize an oppositional system where something is perceived as absent, something else must be postulated as present, at least potentially. The presence of one element is necessary for the absence of the other. All observations concerning the impor-tance of the absent element hold symmetrically for the present element as well. All observations concerning the constitutive function of differ-ence hold for the poles from whose opposition the difference is gener-ated.
The argument is, therefore, an autophagous one. A phoneme is no doubt an abstract position within a system, and it acquires its value only because of the other phonemes to which it is opposed. Yet, for an ’emic’ unit to be recognized, it must be formulated somehow as ‘etic’. In other words, phonology builds up a system of oppositions in order to explain the functioning of a number of phonetic presences which, if they do not exist prior to the system, nonetheless are associated with its ghost. Without people uttering sounds, phonology could not exist, but without the system postulated by phonology, people could not distinguish be-tween sounds. Types are recognized through their realizations into con-crete tokens. One cannot speak of a form (of the expression or of the content) without presupposing a matter and linking it immediately (neither before nor after) to a substance.
1.5.4. The predominance of the signifier
The answer given to the preceding question could confirm a further critique of the notion of sign. If the sign can be known only through the signifier and if the signified emerges only through an act of perpetual substitution of the signifer, the semiotic chain appears to be just a ‘chain of signifiers’. As such, it could be manipulated even by the unconscious (if we take the unconscious as being linguistically constituted). By the ‘drift’ of signifiers, other signifiers are produced. As a more or less direct consequence of these conclusions, the universe of signs and even of sentences would dissolve into discourse as an activity. This line of thought, derived from Lacan, has generated a number of varied, but essentially related, positions.
The basis for this critique is actually a misunderstanding, a wordplay. Only by substituting ‘signified’ every time ‘signifier’ appears, does the discourse of these theoreticians become comprehensible. The misun-derstanding derives from the fact that every signifier can only be trans-lated into another signifier and that only by this process of interpretation can one grasp the ‘corresponding signified’. It must be clear, though, that in none of various displacement and condensation processes de-scribed by Freud —however multiplied and almost automatic the generative and drifting mechanisms might appear—does the interplay (even if based on assonances, alliterations, likeness of expression) fail to reverberate immediately on the aggregation of the content units, actually determining the content.
In the Freudian passage from /Herr-signore/ to /Signorelli/, a series of expression differences is at work, based on iden-tities and progressive slidings of the content. The Freudian example can, in fact, be understood only by someone who knows both German and Italian, seeing words as complete sign-functions (expression + con-tent). A person who does not know Chinese cannot produce Freudian slips interpretable in Chinese, unless a psychoanalyst who knows Chinese demonstrates that his or
