Adverbial Phrases

1. Semantic roles
The section contains a discussion of what S(emantic) R(oles) relevant (a) for general discussion, (b) for Latin.
SR’s are introduced by most authors without further discussion (e.g. Croft, 1991). As soon as one asks oneself how many semantic roles should be distinguished, one realizes that there are virtually as many roles as there are possible different participants in a state of affairs. For example, we speak of a SR Accompaniement in cases as “I go to school with Mary”. But what about “I go to school with my books”? On the one hand, we tend to conceive the second sentence as based on an extension of the first: in a translated way, the books ‘accompany’ the agent. On the other hand, while in the first sentence we have two agents (“I go to school” + “Mary goes to school”), in the second we have an agent and a patient (“I go to school” + “I carry my books”). Distinctions must stop at a certain point: but exactly what point is a very hard question to answer. If we turn to studies where this limit is explicitly stated, the answer is unsatisfactory. I start with concrete relations (first of all Space), because I follow an approach based on meaning and meaning extension; consequently, I assume that forms, too, are relevant (not only functions): this assumption of course is based on a more fundamental assumption, i.e. that forms (including grammatical forms) have a meaning. I think that the strictly functional assumption, inherited by classical structuralism, that forms have at the best a distinctive function blurs a number of interesting facts about SR and their mutual relations, which, if noted, lead to less complicated, more ‘natural’, explanations of linguistic phenomena ( I assume a Cognitive Grammar framework. Note that, although similar to the traditional approach, the CG approach does not try to establish a Grundbedeutung for each form, that can explain all its uses. Rather, each form is conceived as constituting an example of ‘structured polysemy’, whereby different meanings can co-exist, in a radial category structure).
2. Space
The spatial relations discussed include Location, Direction, Source, and Path (as Path I understand both examples as “I go through the door” and “I walk along the sea shore”, which is commonly referred to as ‘extension’ in Latin handbooks). Note that, syntactically speaking, these SR are not all, and not always, on the same plan. In particular, Location is certainly argumental with verbs like ‘be’, ‘stand’, ‘abide’, etc. However, most frequently it is adverbial: Direction, on the other hand, is virtually always argumental, since it typically occurs with bivalent motion verbs. Source and Path are virtually aways adverbial (but what about verbs such as ‘pass by’?). Since, as far as I can judge, there is no marking difference for these SR in either syntactic function, I will try to keep the syntactic discussion as limited as possible. One of the most interesting issues raised by Latin in the field of spatial relations is the use of plain cases vs. P(repositional) P(hrase)s. It is well known that plain cases only occur with toponyms and a restricted number of other nouns with spatial reference, and that their use involves retention of an otherwise disappeared case, the locative. The problem is further complicated by the fact that the locative only survives limited to the first and second declension, nouns of the third declension using the ablative in its place. So for nouns of the third declension, the ablative can express Location, Source, and even Path. With this situation one would expect an increase in the use of prepositions; however, apparently this is not the case until very late: in som