How to connect linguistic atlases with digital lexicography ? First step for a dynamic network in the Galloromance field

The digital revolution is at the origin of an international effort that started a couple of years ago towards digitizing and linking lexicographic resources. Historical French and Romance linguistics also took part in this dynamic, through initiatives such as the creation of new digital dictionaries (Dictionnaire du Moyen Français a.k.a. DMF, Dictionnaire Étymologique Roman a.k.a. DERom) and the digitization of printed work (Trésor de la Langue Française informatisé a.k.a. TLFi). In the future, it will be possible to link most of these key resources to the Französisches Etymologisches Wörterbuch (FEW), which they complete and often correct on specific diachronic and diatopic aspects (cf. Renders/Baiwir/Dethier 2015).
Among reference works addressing diatopic variations, the French linguistic atlases are not yet available in a digital form, although the preservation and visibility of the dialectal linguistic heritage would greatly benefit from its digitization and linking. Whether the integration of an atlantographic resource in a lexicographic network can be achieved, remains to be proved.
The APPI (Atlas Pan-Picard Informatisé) project aims to prepare the future integration of Galloromance linguistic materials into the network in the making. The study focuses on the picard dialect, which is particularly interesting because it spreads over France and Belgium and developed differently on either side of the border (cf. Baiwir 2017). The Belgian method, applied in the Atlas Linguistique de la Wallonie (ALW), defines a model for an atlantico-lexicographic structure which could, after a few adjustments, provide a solution to link French atlases to the FEW. In addition to this, the APPI project offers a global view on the picard heritage that goes beyond administrative borders, and its valorization through new modes of exploitation.
This new project involves three phases. The first one consists in the development of an atlantographic pan-picard corpus, unifying for the first time dialectal materials collected via surveys that were conducted in the twentieth century in France (Atlas Linguistique et ethnographique picard a.k.a. ALPic) and Belgium (ALW). The second step aims to turn this corpus into a digital atlanticolexicographic resource, replacing the dialectal data into the historical context of their lexical family. At this stage, an online user interface will give access to the new digital data, enabling researchers to analyse them with new methods. Finally, the third phase addresses the linking of the new digital resource with the FEW and explores the feasibility of extending the model to other linguistic areas.
The project will bring an important contribution to the fields of digital lexicography and French and Romance linguistics. The numerous projects ongoing in the latter will benefit from an easier access to picard data, in particular pan-roman initiatives (DÉRom, ALiR, etc.), whereas digital lexicography will gain a laboratory to test a model that links lexical data through their etymology. This paper exposes the scientific issues that need to be solved and explains how this new digital resource will help integrate linguistic atlases into a lexicographic network.