Rethinking Unicode: How to Digitally Encode Non-LinearWritten Artefacts? A Tentative Encoding of Codex Mendoza, Folio 2r

Unicode encoding, even after 23 years of development since the release of version 1.0, continues to reflect the limitations of analog and alphabetic typography, emphasizing a linear and analytic structure where characters are arranged in a mechanistic sequence based on the arbitrary segmentation of speech inherent in alphabetic systems. In critically assessing this aspect, at the same time we address the main problem of avoiding the bias of homo alphabeticus, and of course typographicus. As several scholars argued, Aztec, Mixtec, and other non-Maya Mesoamerican pictographic writing systems (often considered as basically non-glottographic systems) cannot be reduced to linear sequences of isolated characters; rather, their graphic plasticity conveys semantic (and linguistic) contents through conventional patterns of units arranged in graphic space according to non-linear principles. This, in fact, explains the enduring opposition to accepting such a notation as writing “proper”; however, this account still suffers from a theoretical bias according to which only more or less systematic phoneticism defines “true writing”. In contrast, any accurate consideration of the ways in which writing works in specific visual artifacts shows that writing, unlike speech, does not need to represent phonetically an utterance in order to provide a complete encoding of linguistic content. We argue that the relevant visual values in this system are ordered in subsets (topological, spatial, and logical) according to the variables devised by Bertin. We suggest then that each unit is a “character” (in the Unicode vocabulary), but it could be coded either as a positive unit (i.e., a graphic sign) or as an “empty” spatial, topographical relation between visually expressed and coded signs. Following the digital method aimed at processing sets of basic graphic features as non-linear combinations of variables, we pointed out that Aztec pictorially agglutinated complexes (or multi-unit arrangements forming perceptually autonomous units, e.g., a toponym or a proper name) display an overall structure similar to Unicode encoding of emoji since it is a combination of different layers of variables. However, as more complex and higher-ranked coded variables are needed to account for broader textual compositions, we suggested that the coding systems available so far in the Unicode standard (including the one devised for emojis) are still not flexible enough to accommodate an appropriate encoding of Aztec textual units, above all the spatial arrangement. In this paper, we will sketch the guidelines for an exhaustive coding of the perhaps most iconic page of Codex Mendoza: the folio 2r, which relates the founding of Mexico-Tenochtitlan and has been studied extensively by Joaquín Galarza and his research group at the beginning of the 1980s.