Whereas the imaginary is a simple relation between the ego and its images, the symbolic produces the subject through the language (la parole) and realizes its closed order by the Law (the Nom-du-pere). Only through the symbolic is the subject connected with the real, which is, so to speak, its umbilical cord. In Freud the symbolic is a storage of oneiric symbols endowed with a constant signification (see 4.2.4); that is, Freud attempts to set up a code of symbols. Lacan, on the contrary, flattens the relation between expression and content by considering only the internal logic of signifiers (see 1.5.4).
As happens with Levi-Strauss, Lacan is not concerned with the organization of sign-functions; he is, rather, con-cerned with the structural arrangements of signifiers. In “Seminar I” he says that thinking means to substitute elephants with the word elephant, and the sun with a circle. But the sun, insofar as it is designated by a circle, is nothing if this circle is not inserted within a system of other formalizations that, in their entirety, set up the symbolic order. A sym-bol becomes a signifying entity when it is inserted within a world of symbols (undoubtedly for Lacan the symbolic order is an s-code). In this sense Lacan speaks of symbols both for a word as elephant and for a visual sign as the sun-circle, even though the symbolic model he is more inter-ested in is undoubtedly the verbal one. Lacan is not so interested in a typology of signs as he is in the general category of symbolic.
It is, how-ever, clear that for Lacan the symbolic order is what we can call the semiotic one. It is true that, in his interpretive practice, he introduces elements of what we shall call the symbolic mode. But this happens at the level of the interpretation of oneiric-verbal texts. From the point of view of a general definition, Lacan identifies the symbolic with the semiotic in general.
Symbolic and semiotic are the same also for Cassirer (The Philosophy of Symbolic Forms, 1923). Science does not mirror the structure of being (considered as the unattainable Kantian thing-in-itself): “The funda-mental concepts of each science, the instruments with which it pro-pounds its questions and formulates its solutions, are regarded no longer as passive images of something but as symbols created by the intellect itself.” (ibid.; Eng. tr. p. 75) Cassirer mentions Hertz and Helmholtz’s theory of scientific objects as ‘inner fictions’ or symbols of outward ob-jects: “These symbols are so constituted that the necessary logical con-sequences of the image are always images of the necessary natural con-sequences of the imagined objects” (ibid.).
Cassirer does not identify symbols only with those models or diagram rules by ratio difficilis (see 4.2.3 below); his purpose is a wider one. He deals with the Kantian theory of knowledge as if it were a semiotic theory (even though Cas-sirer’s a priori is more similar to a cultural product than to a tran-scendental structure of human mind): the symbolic activity does not ‘name’ an already known world, but establishes the very conditions for knowing it. Symbols are not translations of our thought; they are its or-gans:
The logic of things, i.e., of the material concepts and relations on which the structure of a science rests, cannot be separated by the logic of signs. For the sign is no mere accidental cloak of the idea, but its necessary and essen-tial organ. It serves not merely to communicate a complete and given thought content, but is an instrument, by means of which this content develops and fully defines itself. . . . Consequently, all truly strict and exact thought is sustained by the symbolic and semiotics on which it is based. (Ibid., pp. 85-86)
4.2. Expression by ratio facilis
4.2.1. Symbols as conventional expressions
Peirce defines an icon as “a sign which refers to the Object that it de-notes merely by virtue of characters of its own,” an index as “a sign which refers to the Object that it denotes by virtue of being really af-fected by that Object,” and a symbol as “a sign which refers to the Object that it denotes by virtue of a Law, usually an association of gen-eral ideas. . . . It is thus itself a general type” (C. P. 2.249).
As such, a symbol is correlated to its Object by an arbitrary and con-ventional decision. In this sense words are symbols insofar as their lexi-cal content depends on a cultural decision. Since Peirce had decided to use the term sign for the genus generalissimum of semiotics, he had to decide whether to reserve symbol for iconic signs (as Saussure and Hjelmslev did; see 4.3.1) or for the category of arbitrary signs. He made his choice following a rather frequent scientific usage, by which symbols are conventional signs standing for chemical, physical, or mathematical entities. It is true that Peirce knew very well that these scientific sym-bols display many ‘iconic’ qualities (as we shall see when speaking in 4.3.1 of expression produced by ratio difficilis and conveying a ‘literal’ meaning), but it is also true that Peirce never identified something as a mere symbol or as a mere icon. In any case, his decision contrasts with the most common terminological usage, and he certainly never thought that symbols convey a vague meaning. On the contrary, he speaks of symbols for those expressions that mean directly and univocally what they are designed to mean.
Curiously enough, many call symbols in the Peircean sense those stylizations (such as flags, emblems, astrological and chemical symbols) that Peirce would have recognized as abundantly endowed with iconic qualities. Probably at their very beginning, the alchemical symbol for the Balneum Mariae and the astrological symbol for Lion displayed some more or less evident ‘analogy’ with their content, but nowadays they function as conventional devices. Emblems, coats of arms, and other heraldic devices do have a second sense: an image represents in first instance a tree, a hill, a city while its heraldic meaning is another one. But they are visual allegories whose meanings — even though multiple and difficult to guess —are already coded. Thus either they are symbols in the Peircean sense or they are ‘literal’ expressions ruled by ratio dif-ficilis; in both cases they must be excluded from the rank of the in-stances of a symbolic mode.
4.2.2. Symbols as expressions conveying an indirect meaning
The full content of a sign can be actualized only by progressive interpre-tations. But the notion of interpretation (rooted in the one of inference) is not sufficient to characterize the symbolic mode. It characterizes every semiotic phenomenon at large.
There are, however, many expressions (usually sentences or texts) that suggest, beyond theirprima fade interpretable ‘conventional’ or ‘lex-ical’ meaning, an additional ‘intended meaning’ (see Grice 1957). If I tell a lady that I saw her husband at a cozy restaurant with a beautiful girl, I undoubtedly try to convey along with the literal meaning the in-tended meaning that the lady’s husband is unfaithful to her. This second meaning is certainly ‘indirect’, and, as such, it must be actualized by an inferential labor on the part of the addressee; nevertheless, it is neither vague nor ambiguous.
Todorov (1978), aware of the difficulty of assigning a univocal sense to the term symbol, decides to provide a framework within which all its contrasting definitions can find a place, and wants to keep as “plural” what is in fact irreducible to a unique definition.
In doing so, however, he accepts the line of thought criticized above; he identifies symbols with the whole gamut of indirect and even of direct meanings: connota-tions, presuppositions, implications, implicatures, figures of speech, in-tended meaning, and so on. Once again the symbolic is identified with the semiotic in general, since it is impossible to think of discourses that do not elicit some inferential response.
Many of the devices people call symbols have something to do with these phenomena of ‘indirect’ meaning, but not every device conveying an indirect meaning can be called symbolic. Every semiotic device can be used, if not from the point of view of the sender, at least from the point of view of the interpreter, in order to actualize further meanings.
All these instances of indirect meaning say what they are intended to say on the basis of contextual inferences governed by semantic or prag-matic rules. What the sender intends to express, what he wishes to be understood, is so precise that the sender would be irritated if the ad-dressee did not understand it. On the contrary, the genuine instances of a symbolic mode seem to be those where neither the sender nor the addressee really wants or is able to outline a definite interpretation.
4.3. Expressions produced by ratio difficilis
4.3.1. Symbols as diagrams
Saussure called symbols what Peirce called icons, and Hjelmslev ranked diagrams and games among the ‘symbolic systems’, meaning by symbolic systems those which are interpretable but not biplanar. Thus Hjelmslev listed among symbols those signs that are isomorphic with their interpre-tation, such as in the case of pure games, in the interpretation of which there is an entity of content corresponding to each entity of expression (chesspiece or the like), so that if two planes are tentatively posited the functional net will be entirely the same in both. . . . Symbol should be used only for entities that are isomorphic with their interpretation, entities that are depictions or emblems, like Thorvaldsen’s Christ as a symbol for compassion, the ham-mer and the sickle as a symbol for Communism. . . . There seems to be an essential affinity between the interpretable pieces of a game and isomorphic
