What status do photographic imprints take on in comparison with a possible spectacle? Are they signs? Or indices? Language is of great help in this matter as it differentiates between signs (signes) on the one hand, and indices (indices) and indexes (les index) on the other. We will take up these distinctions as they are very useful in our discussion.

Signs are intentional, conventional and systematic signals. They designate in the strong sense of the term. Paintings and sculptures are analogical signs, because they designate their designated according to a certain proportion (analogy). Words, numbers and punctuation marks are digital signs, because they designate their designated by labeling the latter in accordance with a system, and because this labeling follows a sequence of numbers (digits), which is ultimately reducible to a choice between 0 and 1. Indices are not signs; they are the physical effects of a cause they physically signalize, either through monstration "as when the imprint of a boar's paw shows this same paw" or demonstration, as when an unusual disarrangement of objects might reveal a thief's route to a detective. Indices are non-intentional signs, and are neither conventional nor systematic, but physical. Lastly, indexes indicate objects much in the same way the index finger or an arrow might point to an object. These are outright signs, as they are intentional, conventional, and systematic signs. Moreover, they are minimal signs since they designate nothing by themselves; they merely indicate.

These specifics suffice to draw attention to the fact that photography does not belong to the realm of signs, as is the case with drawings or words (even considering that one can photograph drawings or words). On the contrary, photonic imprints are precisely indices that signal their cause, i.e. the spectacle, either through monstration, as when dark and light stains might reveal a deer, or through demonstration, as when a statistical distribution of blackened points allows one, through reasoning, to discover a heavenly body or the weapon of a murderer for instance. Finally, indexes can indicate certain privileged parts of imprints, and therefore also accentuate or orient photographic indices. Such indexes are well-known. It concerns, for instance, the darkening or brightening of certain parts of imprints during development. Or the choice of film, printing, or diaphragm, showing that one attempted to draw attention to morning or evening light, or to the grades of shade of the undergrowth. Or the specific enclosing of a motive through a certain depth (superficiality) of field. It also concerns all the modalities of framing. For one must keep in mind that there are two types of framing in photography involving completely distinct effects: a) a frame-limit; this is characteristic of every photograph due to the simple fact that its borders are straight and cut at right angles; b) a frame-index or framing (centering), which possibly foregrounds, indicates or signals particular parts of the print, and therefore also specific indices.

What is exceptional in the photograph when it includes indices and indexes is that the latter maintain an extremely intimate relation. Of course, I can simply index a photograph from the outside, either roughly by writing an arrow over it, or subtly trough the inclusion of a mark on the photographic film or by maintaining its winding grooves as pointers. But true photographic indexes such as framing, brightening, darkening, depth of field and so on, signal indices from the inside, whose texture and structure they accentuate and orient.

Thus, indexed most intimately, photographic indices are all the more powerful as they are facial, that is to say, they present the spectacle from the side normally seen by the viewer, and by preserving the plane (however summarily). And this is by no means trivial. For the imprint-index in the mud of a boar shows a concave for a convex, and the imprint-index of the Turin burial shroud is reversed left to right, in the same way as the handprints on the cave walls of Pech-Merle. The imprint-index of the shade on a wall fuses front and back through its purely negative cut. By contrast, the photograph, in that it makes me see the effects of a cause according to a direction and plane by which I ordinarily perceive such causes, provokes, through these effects, my mental schemata into movements very similar to those that gave rise to the cause in the first place.

In this case, it is tempting to say that indices denounce, betray, reveal, declare, and make public their own causes. However, the slightest excess of vocabulary would be fatal here, because it threatens to obfuscate what is most specific to the photographic index, namely its terrible muteness, which one is in danger of confusing with the eloquence of signs. We must therefore content ourselves with speaking of monstrative (and demonstrative) facially accentuated and oriented indices.

In the preceding chapter, we saw that luminous imprints introduced the paradox of being simultaneously the clearest and the most blurred. We can now ascertain that its semiological, or rather indexological status is by no means more reassuring.

The photograph is made up of indices. Therefore, its unity of construction and reading is not the decision of the trait, which is characteristic of signs, even of those in China or in caverns, but of the littoral. In the photograph, the trait is always but the extreme case of rectilinear or curvilinear elongation of the littoral. And this renders its interpretation floating.

Consequently, when and at what point are indices to be distinguished from their background noise? And are they ever truly distinct? Is it not better to say that indices are in continuous overlap and in a situation of problematic emergence from their background noise? Furthermore, how can one enumerate them? Are there ten, a hundred, or a thousand on the photographic film of a celebrated journalist or even an inattentive amateur? Photographic indices are difficult to delimit, and they are always uncountable.

Of course, they are signaled, accentuated, and orientated by their indexes. But precisely what relations hold between the photographic indexes? Do they assume functions well-defined enough for us to speak in terms of a syntax or a code of indexes? Or do they rather intentionally and conventionally organize indices solely according to broad and floating aggregates, as is the case in rhetoric? Due to the floating quality we have mentioned above and to which we will return, it seems more apt to speak of a rhetoric of the index.

But there are less naive oddities. In view of that, the indices of any photograph echo their cause (their possible spectacle) through monstration and demonstration. This engenders a permanent ambiguity within the gaze, even when we do not think of it explicitly.

On the other hand, the monstration effected through the photograph is simultaneously facial and distant. And, once again, the facial and physical character of the imprint-index makes something appear, but at the same time its characteristic distance removes me from it: it is not some thing that has touched the film but only photons that have touched this thing and the film, thereby only remotely and very abstractly linking both. And to this bifurcation of space (being there, not being there) a bifurcation of time is added. Since, as the physical effect is there-now, its cause is also there-now, but nonetheless I cannot know any more than that this effect was caused by it. All photographs effectuate a terrible tension between what is near and what is distant, between the present and the past.

Concerning the notion of reference in photography, its subtlety can be summarized by pointing to three usages of the verb. Signs refer to their designated, which one ordinarily calls the referent. Indexes simply refer, since they lack the designated (referent) in themselves. At best, indices are referred to, which is the case when they are indexed by indexes, as is customary in photography. One might now therefore understand how ambiguous it is to speak of the referent of a photograph "unless we take Byzantine precaution" since indexes are the only signs and factors of reference of a photograph, and since indexes directly point to indices and point only indirectly and extremely fragilely to the signaled spectacle.

In addition, the diagnosis of photographic destination is by no means more favorable. First of all, to be veritably destined for an addressee, it helps to have rather firmly established designates (referents), which is the case with signs, but not, as we have seen, of photographic indices. Besides, as everyone knows, a vast amount of photographs is made incidentally, at random, or off chance. However, even with photographs directly intended for someone, the destination is either predominantly or wholly extrinsic to the texture and structure of the photograph itself. In brief, in order to address the status of reference or destination, it is certainly advisable never to speak of the photographic message, unless explicitly adding that delegation (mission) is extrinsic to the photographic film itself (as when I send a photograph of a citadel to an officer to tell him to besiege it and how to do so), or unless one understands the term message in the sense of an interpreted signal, which constitutes a misuse of the term now long since abandoned within the growing field of communication theory.

These are not mere quibbles. Maintaining that the photograph has no referent, or at most very indirectly, does not diminish it. The relation of reference specific to signs is preeminently exterior and conventional. Made up of physical signals that physically indicate or demonstrate their cause, the photograph has an incomparable power. Similarly, to say that the photograph has only weak destination in terms of an addresser and addressee, does not deprive it of its force either. To the contrary, it is in order to foreground the photograph's fearful sufficiency. It is the autarkic scion, always eluding our grasp.

Photography is an ambiguous word. Graphs, as in writing or drawing, are the human products par excellence; and light, as physical agent, cannot be drawn or described. A photograph is strictly an effect. Photo-effect. Effect-photo. This has to be understood in the classical sense in which the effect signals its cause, but is also self-sufficient. The photograph as new being, being sui generis; as efficient as it is indicative.