The “Writing of the Birds”. Sorting Iconic Variation through Decipherment in Egyptian Hieroglyphs

The conference’s leading issue — whether the identification of allographs necessarily presupposes the linguistic analysis of a decipherment, or whether an understanding of graphic variation can in fact inform that decipherment — is, we argue, on the basis of Egyptian hieroglyphic writing, best approached by inverting it: the distinction between allograph and distinct sign is not the precondition of decipherment but a categorisation it produces. Egyptian hieroglyphs offer a revealing case, because they are at once images and signs of writing, so that a visible difference between two signs is very often an iconic difference — more or fewer details, a different posture, an added attribute — whose linguistic pertinence is rarely self-evident.
Our argument proceeds in three steps. We first survey the kinds of variation a figurative script exhibits — partly externally conditioned (a single writing system realised in coexisting norms; further modulated by medium, diachrony, diatopy), partly internal to the system (logo-phonographic economy, open specification of the signified, polygraphy, mutual interferences, and graphic dissimilation). The common thread is that a visible difference does not wear its semiotic status on its face.
We then turn to Champollion’s decipherment. Two apparently contradictory but complementary movements were at work. He demanded an ever more faithful rendering of the sources, which made the variation visible. He then identified which differences were pertinent, pre-eminently through the hieratic cursive, whose equivalences with the hieroglyphic forms he established on purely formal grounds before any reading was assigned. The hieratic case illustrates what we call de-iconization (in a strictly formal sense): figurative shape recedes while graphemic value is preserved. Form and function are independent dimensions — the principle that, through Knorozov, became the working maxim of every later decipherment.
What a decipherment produces, finally, is a categorisation imposed upon a gradient of iconic variation, and the categories it builds are of two orthogonal kinds: a graphemic inventory of classes individuated by linguistic value, identified by commutation, and a grammemic inventory of classes individuated by visuo-spatial values. Any repertoire that would fail to take those two dimensions into account would inevitably misrepresent the script. The Thot Sign List is an attempt to operationalise this distinction and is offered less as a finished product than as a model to follow for the figurative scripts that still await decipherment.