This article aims at discussing the premises for a unified account of auxiliarisation, here understood as aspecific subcase of grammaticalisation. The passage of Latin HABEO from a lexical verb to a tense-formingauxiliary is certainly one of the most well-studied innovations of Romance languages. Equally familiar are thecases of auxiliarisation of Latin TENEO, Germanic *habhen and *getan, as well as Scandinavian derivativesof Old Norse fà. Such processes follow a similar path, in the sense that they originally select a secondarypredication in the passive voice, which over time is reanalysed as active. At the same time, the governing verbis void of lexical content and turns into an auxiliary, while the implicit agent of the secondary predicate isreinterpreted as the surface subject of the construction. If a unified theory is to be attempted, such an approachshould capture why such a path of change is consistently observed and, moreover, why it seems to be a definingproperty of such auxiliarisation that the verbs involved originally describe possession, reception, or control.Furthermore, ideally the unified theory should account for why the semantic output of these processes variesover time: the earliest cases of auxiliarisation are precisely those involving HABEO/habhen, which give riseto the compound tenses in modern languages. Subsequent cases, however, such as the auxiliarisation ofTENEO/getan etc. do not lead to the formation of compound tenses, but rather to what could be defined ascompound aspects or, sometimes, compound causative constructions. This circumstance, too, requires aprincipled account.
Read full abstract