Abstract :
[en] Semantic maps aim at detecting cross-linguistic regularities and recurrent patterns in semantic structure (Haspelmath 2003). This method was initially applied to the grammatical domain, mostly in a synchronic perspective (with some exceptions, see, e.g., van der Auwera & Plungian 1998). Recent research, however, has drawn attention to the lexical domain, showing that the model can also include lexical semantics (see, e.g., François 2008).
Intimately related to the semantic map method has been the issue of finding appropriate ways of creating those maps and of capturing visually the semantic regularities. Up until recently, the maps were plotted and drawn manually. However, Regier et al. (2013) showed that a good approximation algorithm exists for inferring semantic maps based on polysemy data. Elaborating on their method, this paper aims to demonstrate that information about the directionality and the weight of the edges can be automatically added, thereby providing valuable information regarding the paths of semantic extensions and their frequency. In order to illustrate the method, we take the example of the semantic extension of time-related lexemes (e.g. TIME, HOUR, SEASON, DAY) in Ancient Greek (8th – 1st c. BC) and Ancient Egyptian – Coptic (26th c. BC – 10th c. AD). Both languages give access to significant diachronic material, allowing us to trace long term processes of semantic change. The results of our diachronic investigations are then checked against databases giving information on synchronic polysemies (e.g., List et al. 2014). In doing so, we also assess the adequacy of the use of polysemy as a tool to investigate semantic change.
