Word formation in Latin is a fascinating subject that has been approached from different theoretical perspectives in recent decades. In this area there are some prominent topics that have captured the attention of various researchers.
Within the 20th International Colloquium on Latin Linguistics, held at the University of Las Palmas de Gran Canaria on 17-21 June 2019, we organized a workshop to congregate researchers who are working on word formation in Latin from different theoretical perspectives. Our goal was to begin a free and informal exchange of ideas in order to promote a fruitful cross-theoretical debate.
In this number of the journal we have collected a selection of the papers presented in the workshop. Each paper has been reviewed by two anonymous reviewers and we are very grateful to them for their efforts and help to improve the quality of the contributions. The papers offer different paths of argumentation and linguistic theorizing, which provide a substantial overview on the area of word formation in Latin today.

In this chapter, I concentrate on some of the most relevant theoretical and empirical issues where argument structure has been shown to play an important role. Below, I address the question of to what extent (verbal) argument structures are projected from lexical-semantic structures or are constructed outside of the lexicon. In doing so, I review the recent debate between projectionist and neo-constructionist approaches to argument structure. In this context, some special attention is also devoted to some syntactic approaches to argument structure like Hale and Keyser’s (1993, 2002) configurational theory, which has been very influential for many syntacticians interested in analyzing argument structure from a minimalist perspective.

In this paper we are interested in the relation between two facts accompanying the diachronic change from Latin to Romance within the domain of the morphological and argument-structural properties of the predicates expressing change. On the one hand, the element encoding the transition itself, which we call the Path, and the verb are realised as two distinct morphemes in Latin, but as one and the same morpheme in the daughter languages: in Talmy's (2000) terms, the former is a satellite-framed language and the latter are verb-framed languages. On the other hand, there is a whole range of argument-structural patterns which are found in Latin but not in Romance: unselected object constructions, complex directed motion constructions, productive locative alternation, etc. We show, within a syntactic view of argument structure and morphology, that both facts are intimately related. Furthermore, we provide data from Old Catalan showing an intermediate stage between the Latin satellite-framed system and the Romance verb-framed system.