This issue of Signata aims to review the heuristic character of this notion of form (or formula, in Warburg’s case) and to evaluate it from the point of view of the ambitions of an art history that makes increasing use of new technologies. One will immediately notice several issues and research fronts to be considered: (i) a critical updating of the notion of form as conceived of by the contemporary debates in art history, by integrating it with the contributions of visual studies and semiotics; (ii) an operational definition of the notion of form so that it can be implemented in automatic image processing programs; (iii) an elaboration of a history of forms program able to establish an articulation, in diachronic studies, between (digital) art history and the semiotics of art.

This issue aims to reconsider some theoretical assumptions taken for granted and different categories now considered as tacit; at the same time, it wishes to offer a second chance to paths of investigation considered to be dead ends or sterile approaches. In order to promote the catalysis of this re-examination, we consider that it would be relevant to articulate what comes to us from current events—the digital turn in the humanities—with its latent tradition, that of the history of symbolic forms, which has not been able to express its potential for lack of the right means. Indeed, without the support of digital technology, it is impossible to work comparatively on large corpora of images and thus to annotate and make the inventory of more or less recurrent traits and motifs that are traversed by various evolutions. That said, the fact that the theories of the past have remained, at least in part, programmatic, or even at the level of intuitions, shouldn’t now lead to a “devolution” of the work of the historian or semiotician to machines and software. If computer-assisted analysis has been much discussed, it is because the automatic processing of data can only remain a passage within an interpretative practice that goes beyond machine horizons and that encompasses additional critical means to achieve its objectives. Basically, we could reverse the usual rhetorical expression so as to underline the complementary dimension of “interpretation-assisted calculation.”

This “chiasmatic” version of mutual assistance between machine and interpreter may be a good way to correct some of the mistakes of the past, but it may not yet be able to take into account the advances in available technologies. When we talk about deep learning, we are referring to machine learning structured on architectures with various levels of representation; these levels are hierarchical and can take over from one another and thus accompany non-linear and recursive data processing. Most of the calculations could be reduced to the extraction of information and to the recognition of models, but as soon as deep learning is embraced as a prosthesis by the researcher, it appears to construct a new perspective on the investigated phenomena and, although not neutral, it produces knowledge and thus confers a significance to the results that goes beyond simple recognition.

To get to the bottom of the matters and reflections our issue would like to investigate, it seems to us that the artistic forms we want to bring to the fore and study not only become observable through digital devices, but that they are also, in turn, enlightened by the symbolic forms that the latter express, which become another modality of comprehension or, even better, of self-understanding of our culture. But the fact that we simultaneously find symbolic forms taken as a corpus of analysis, i.e. as an object-semiotics, and a digital perspective that also brings its own symbolic form to the level of objectification—as per descriptive semiotics—does not indicate a unilateral hierarchization and construction. In the human sciences, the object cannot be constructed unilaterally by the eye, since this object exerts constraints due to its status as cultural heritage and its own epistemic aims. In short, this issue also seeks to host a critical dialogue between art and science, an encounter between symbolic forms that shows homologies but also resistances, recognitions but also blind spots, and significant results but also insignificant ones. It is from this not so innocent perspective of the machine, but which is still capable of having a “depth” of its own, that we would like to invite to this debate semioticians, art historians, computer scientists and other colleagues who work in the field of the digital humanities.

The objective is also to understand the issues pertaining to the “digitality” of Digital Art History (thus, as endowed with its own symbolic form), as well as the perspectives of an art history that considers it can take advantage of the digital while objectifying its contribution as one way of accessing heritage. Putting symbolic forms into perspective and questioning their mediation requires the intervention of a semiotic reflection. In this sense, Signata will be able to integrate, in a relevant way, contributions which will touch in an implicit and explicit way the problematic points that we tried to illustrate.

Finally, it seems to us that, in order to establish a dialog between observed images and observing instances, the notion of form can work as a fruitful mise en abyme, or even as the most appropriate system of reflections. The symbolic form is not idealistic, but anchored in technical devices, linked to a tradition of practices and evidenced by a genealogy of representations. Each artistic image responds to other images; this response can be the result of a project but it can also emerge in an unintentional way as a progressive revelation of its belonging to a family of images.

The archaeology of the artistic forms entails the instruction of a conjectural inquiry and works on clues to bring out possible genealogies, causal determinism being only the fruit of a reductionist approach. The cultural environment, the period, the techniques available, artistic meetings, collective movements, etc., remain multiple and copresent factors of which the artistic form assumed by the artwork is but a synthetic response, readable only from patterns of convergent or divergent intentionalities (cf. Baxandall 1985). But precisely, these patterns may only be described a posteriori and from a position required to explain, in turn, its own starting point, for example the web of intentionalities which constitutes its own disciplinary current.

By presenting the symbolic form as the intersection of a web of intentionalities and a genealogy of representations, we thus aim to bring together the perspectives of traditional art history and of its digital version, those of the iconologists and of the semioticians, and those of the humanists and of the computer scientists. But to the scientific figures it will be necessary to add all the taxonomies and the informal histories of the symbolic forms that one finds in our digital society. They transparently manifest a viewpoint and produce genealogical reconstructions that may indeed be unscientific and even aberrant at times, but that are capable, in any case, of exemplifying a symbolic form. The latter is not always to be patrimonialized and, essentially, to favor its analysis also means to promote its demythification, if not its critical deconstruction, if necessary. To this end, this issue also promotes comparisons between acquired and reconstructed symbolic forms, by analyzing the differences in vocation (self-assigned) and social status (artistic, scientific, mythical, etc.).