Without therefore detracting anything from the active moment in the perception and interpretation of hypoicons, we must admit that there are semiosic phenomena in which, even if we know that we are dealing with a sign, before perceiving it as a sign of something else we must first perceive it as a set of stimuli that creates the effect of our being in the presence of the object. In other words, we must to accept the idea that there is a perceptual basis even in the interpretation of the hypoicon (Sonesson 1989: 327) or that the visual image is first and foremost something that offers itself to the perception (Saint-Martin 1990).
If we get back to the waxwork figure, and admit that a good photograph poses the same problem, even though the stimuli it brings into play are «more» surrogate and vicarious, it must be admitted that the majority of attempts to analyze so-called iconic signs in morphological and grammatical terms—as if they could be subdivided in a way typical of other sign systems, starting from the principle that a photo, for example, can be broken down into the smallest elements of the screen upon which it is founded—have been largely frustrated. These minimum elements become grammatical entities when they are intentionally magnified as such, that is, when the screen does not tend to disappear to give the effect of a perceptual surrogate but instead is enlarged and highlighted with a view to constructing (if nothing else, an objet trouvé, in terms of aesthetic interpretation) abstract symmetries and oppositions.
In this case, in a picture, all we are doing is distinguishing the figurative elements from the plastic ones. Whereas a hypoicon refers (however it refers, and whatever the form of the expression) to a content (whether it be an element of the natural world or the cultural world, as in the case of the unicorn), in the perception of plastic elements one is essentially interested in the form of the expression. Therefore an enlargement of a photo that magnifies its screen would be a way of pertinentizing the plastic elements of the form of the expression, almost always at the expense of the figurative elements.31 As has already been said, as long as the image is still perceivable, the fact that its digital nature has been made clear is no argument against its iconism. It is as if on the television screen we were to go to individuate, from close up, the lines traced by the electron beam. It would be an interesting plastic experiment, but usually the effect of those lines is comparable to the effect of a mirror that has been painted with opaque strips at regular intervals. If the strips are not too many, so that recognition of the image is rendered impossible (just as if on the television screen the lines are not too few), we treat the surface of the mirror as if it were misted over or marked (with reduced definition, as if the water in Narcissus’s pool became cloudy, but not too much), and we do our best to integrate the stimuli and perceive a satisfactory image.
Yet the screen test is not a useless one. The fact is that, by working on enlarged screens, we measure the threshold beyond which the image is no longer perceivable and a purely plastic construction appears. What counts (see Maldonado 1974, plate 182) is the last stage of rarefaction at which the figure is still perceived: that stage represents the minimum of definition necessary for any stimulus to function as a surrogate stimulus (and not to function as a purely plastic stimulus). Naturally this threshold varies according to how well the object represented is already known. No matter how grainy the screen is, the faces of Napoleon or Marilyn Monroe will always be more recognizable than those of unknown persons: the lower the definition and the less known the object, the greater the inferential process required. But I think we can say that beyond this threshold we leave the territory of surrogate stimuli to enter that of the sign.
There is a passage in Ockham that has always perplexed and disquieted me, in which the philosopher states not only that if on coming across the statue of Hercules, I do not compare the statue to the original, I cannot say if it resembles him or not (an observation born of pure good sense) but also that the statue does not allow me to know what Hercules looks like if I have not met him before (i.e., if I do not have notitia mentalis regarding him). Yet, as the police forces of the world have shown us, on the basis of a passport photo one can (or one can try to) identify a wanted person.
One possible interpretation of this curious opinion is that Ockham was familiar with the Gothic and Romanesque statuary of the previous centuries, which portrayed human types, through highly regulated iconic schemata, rather than individuals, as was the case with Roman statuary and with the statuary of the centuries to come. Therefore he wanted to tell us that, in conditions of low definition, the hypoicon allows us to perceive generic but not individual features.
Let us think of a normal passport photograph, one of those taken in a hurry and badly in a photo booth. On the basis of such a document it would be very difficult for a policeman to identify the right person in a crowd without making a major blunder. The same thing happens with police sketches, on the basis of which many of us could be held responsible for horrendous crimes, because it frequently happens that the sketch does not resemble the wanted person and that many of us resemble the sketch.
Passport-size photos are imprecise, because both the pose and the lighting leave much to be desired. The police sketch is imprecise, because it represents an artist’s interpretation of the verbal expressions a witness uses to make a schematic reconstruction of the features of an individual who in many cases was seen for only a few moments. In both instances, hypoicons refer to generic and not individual features. This does nothing to alter the fact that, in the presence of both the photo and the sketch, each of us is able to recognize these generic features (the person is a male, has a mustache and a low hairline, or the person is a woman, not that young, blond, with full lips). All the rest is inferred in order to pass from the generic to the individual. But that modicum of the generic that is grasped depends on the fact that a very poor portrait has provided us in an even poorer fashion with surrogates for perceptual stimuli, otherwise the photo on my driving license would be indistinguishable from that of a penguin.
6.14 Recognition
Let us imagine that in a family the mother keeps on her desk a stack of rectangular filing cards, of various colors. She uses them for different kinds of notes: red cards for kitchen expenses, blue for travel and holidays, green for clothing, yellow for medical expenses, white for her work appointments, sky blue for noting down the passages that strike her most when she reads a book, et cetera. Every so often she adds these cards to the others in the filing cabinet, divided up by color, so she always knows where to find a certain piece of information. For her those rectangles are signs: not in the sense that they are the physical support for the graphic signs she has made on them but in the sense that, even before they are filed away, they already refer to their own particular topic—in accordance with the various colors; they are expressions of an elementary semiotic system, within which every color is correlated to a content.
But her little boy always tries to get hold of them to play—to build houses of cards, let’s say. Naturally he distinguishes between their shapes and colors very well, but for him they are not expressions, they are objects and that’s that.
We may say that the cognitive type that allows the mother to identify the cards is more complex than that of the child. She might feel a sense of disquiet on picking up a yellow card, blank or filled in as the case may be, because this would mean that she would have to deal with questions of health; whereas the child might be indifferent to the color and more interested in the consistency of the filing cards (or he might simply prefer his houses of cards to be red). But if mummy tells the little boy to go to fetch her a red card from the desk, and the act of reference is crowned with success, this means that the fundamental perceptual process for the recognition of the cards is the same for both mother and son. Before the upper levels of semiosis, where the cards become expressions, there is a level of stable perceptual semiosis for all the actors in this little domestic comedy.
We can now consider ways of recognition that concern pertinent nonvisual features, such as sound phenomena. The phenomenon of recognition is also at the root of a fundamental semiosic activity, such as verbal language.32
As is suggested in Gibson (1968: 93–93), phonemes are potential stimuli like natural sounds, but the characteristic thing about them is that for the listener they must be interpreted not just as pure stimuli but also as responses (for Gibson, in the sense that they have been deliberately
