The fact that it is spherical and, even though we can see only one side of it, that it has another side that we do not see and have never seen, is part of a more elaborate, and historically variable, molar content: for example, both Epicurus and Lucretius were convinced that the Moon (and the Sun too) was exactly as large as it appears to us.
In short, I should like to make it clear that I believe in the existence of the Moon, at least to the same degree as I believe in the existence of everything else, my own body included. I am stressing this point because once I was accused of not believing in the Moon. It all came about in the course of what has been defined as «the debate on iconism.»
6.1 The Debate on Iconism
«In their obstinate idealism, they [the «semiolinguists»] dispute everything that, one way or another, might oblige them to admit that reality—in this case, the moon—exists.» This was how, in 1974, Tomás Maldonado, with regard to what I had written on iconic signs, reminded me of my Galilean duty to look through the telescope, thus opening the final phase of the debate on iconism that was in progress during the sixties and seventies.1 To this accusation of idealism—definitely something to worry about in those days—I replied (Eco 1975b) with an equally polemical essay, entitled, «Chi ha paura del cannochiale?» (Who’s afraid of the telescope?).
It is an essay I have never since republished, because I was beginning to realize that the debate had assumed, in public, heated tones that it did not have at all in private. Almost twenty years later, Maldonado republished his article, but minus the pages that concerned me, because, he stated, some of my criticisms of his critique «have contributed—I’ll admit this willingly—to modify in part the presuppositions of my analysis» (1992: 5911). May this example of intellectual honesty inspire me now, as I review some of the positions I held at the time.
The debate came about at the wrong time, because, while Maldonado was publishing his essay, my A Theory of Semiotics (which he could not have seen) was already in print, with a chapter on sign production that would have perhaps proved to him that we were in agreement on more points than he thought. In any case it is singular that, after the row exploded, the general discussion reached an impasse, as if it had become a dead letter. There was a hiatus, I should say, of a decade: and then it flared up again, in the hands of others, who had taken a second look at the whole business. 2
The progress of the debate between iconists and iconoclasts3 seems linked to a ten-year cycle: this is not a symptom to be taken lightly, in the sense that perhaps everything should be reconsidered by bringing the Zeitgeist onto the stage every so often. Groupe p observes (1992: 125) that two works dealing with images appeared in 1968: Languages of Art, by Nelson Goodman, and my La struttura assente, and that these two books, written at the same time by two authors from two completely different cultural areas, contain some very similar examples and observations. As if, by way of a confutation of all idealism, the moment two persons distant from each other set to «looking at the figures,» they noticed some common reactions.
When I reread the debate of 1974–75, ^ emerged clearly that the discussion hinged on three problems: (i) the iconic nature of perception, (ii) the fundamentally iconic nature of knowledge in general, and (iii) the nature of so-called iconic signs, in other words, of those signs that Peirce called (and which we shall be referring to exclusively as such, from now on) hypoicons. In my answer to Maldonado it seems that I take point (i) for granted without discussing it, I do not compromise myself over point (ii), while I discuss point (iii) at length. I made the mistake of separating the three problems, but perhaps Maldonado erred in keeping them bound together too tightly. From his convictions regarding the motivated nature of perception, Maldonado derived (on the basis of the early Wittgenstein) a definition of knowledge in terms of Abbildungsthe-orie, and consequently of the cognitive value of hypoiconic signs. From convictions regarding the highly conventional and cultural nature of hypoicons, I raised doubts regarding the motivation of cognitive processes. With hindsight it seems like a comic-strip version of the Cratylus: is it by law or by nature that the image of Mickey Mouse reminds us of a mouse?
Points (i) and (ii) have already been dealt with in 2.8. I don’t think that people (whether they stood for an epistemology of specular reflection or a constructivist epistemology) had any doubts about point (i), even back in the seventies. However, I must admit that, in order to discuss the problem of the hypoicon, I relegated the problem of perceptual iconism to an area of scant semiotic pertinence. 4 On the other hand, many philoiconists (not only Maldonado) have identified the iconism of perception with the iconism of so-called iconic signs, attributing to the second the virtues of the first.
Finally, for a series of reasons we shall be dealing with later, the debate led people to identify both icons and hypoicons with visual entities, both mental images and those signs that (to avoid using a term overburdened with meanings, such as «image») we shall call pictures. Once more this sent the discussion partly off the track, whereas it should have been clear to everybody that both the concept of icon and that of hypoicon concern nonvisual experiences too.5
6.2 Not a Debate Between Madmen
Now let us try to consider the matter calmly. On the one side were people who questioned the vagueness of a concept such as «likeness» and who wanted to demonstrate how impressions of likeness caused by hypoicons were the effect of rules for the production of likeness (see Volli 1972). Is it possible that these people were denying that most of our everyday life is based on relations that, for want of a better term, are of likeness; that it is for reasons of likeness that we recognize people; that it is on the basis of likeness between tokens that we are able to use general terms; that the constancy of perception itself is ensured by the recognition of shapes; that it is for formal reasons that we can tell a square from a triangle? And even if we move on to hypoicons, is it possible that these people were denying the evidence—for example, that a photograph by Penn or Avedon looks more like the person portrayed than does a figure by Giacometti, and that even a person from a non-Occidental culture, if shown a group of statues from Ancient Rome, ought to recognize them as human bodies?6
Evidently not, and it is almost pathetic to see how, in the second phase of the debate (from the eighties until now), many illustrious iconoclasts hastened to make professions of faith in the iconic nature of perception—like the accused in a Stalinist or McCarthyite show trial, obliged above all to reiterate their loyalty to the system. See, for example, Gombrich 1975.
On the other hand, could it be that people so profoundly convinced of the iconic motivation of perception could at the same time deny that graphic conventions, proportional rules, and techniques of projection all come into play in the production and recognition of hypoicons? It seems improbable. This was not a debate between madmen.7
6.3 The Arguments of the Sixties
As Sonesson also recalls in many of his writings, in the field of semiotics it all began when Barthes (1964), in his famous essay on pasta Panzani, stated that visual language was a language without a code. This was a way of suggesting that semiotics takes images exactly as they are and appear to us, and tries if anything to find the rhetorical rules for their concatenation, or to define their relations with the verbal information that makes up for their vagueness and polyvocity, thus contributing to establishing their sense.
In the same number of Communications 4, Metz launched what was to become the semiotics of the cinema. And he too assumed the cinematographic image as an image without a code, pure analogon, reserving semiotic studies (or, as they put it in Communications, semiological studies) for the great syntagmatics of film.
This happened at a time when semiotics was proposing itself as a clavis universalis capable of reducing all communicational phenomena to analyzable cultural conventions; at a time when people were adopting Saussure’s principle, according to which the purpose of semiotics was to study «the life of signs within the framework of social life»; at a time when semiostructuralism was in the process of deciding to tackle not so much the study of laboratory-type expressions, linguistic or otherwise, like John eats apples or The present king of France is bald, as complex texts (even before there was any talk of textual semiotics).
Most of these texts were taken from the world of mass communications (advertising announcements, photographs, images of television
