Typologists usually identify synchronic cross-linguistic patterns, and account for these patterns independently of how they actually developed in individual languages. Reference grammars are also usually synchronically oriented.
The general goal of this seminar is to show that individual patterns are often best understood in terms of their diachronic origins. This has several consequences for language description. First, reference grammars should try and provide as many clues as possible on the origins of individual phenomena, for example through internal reconstruction. Also, linguists often try to identify what grammatical categories best describe particular grammatical patterns. In many cases, however, this is unnecessary, because these patterns reflect several distinct diachronic processes, rather than some overarching grammatical category.
Lecture 1 :
Unrealized states of affairs and "irrealis"
Several languages have been argued to have a grammatical category of irrealis, manifested in phenomena such as person marking, switch-reference,
negation, or the encoding of counterfactual and non-factual events. This category, however, is rather difficult to define, both cross-linguistically and within individual languages. Based on cross-linguistic diachronic evidence, it will be shown that this is because the relevant phenomena are actually a resultof several distinct diachronic processes independent of some overarching grammatical category of irrealis.
Lecture 2 :
The origins of insubordination cross-linguistically
Insubordination is usually regarded as a phenomenon whereby a former subordinate clause comes to be used independently. Cross-linguistically, however, insubordinate clauses can develop through a wider range of mechanisms than assumed so far, and many clause types are actually compatible with different developmental mechanisms and source constructions. Also, the various mechanisms are quite different in nature, and do not exclusively apply to subordinate clauses. This suggests that insubordination might actually be a result of several different processes pertaining to clause combining in general, rather than a unified process specifically pertaining to subordination.
Lecture 3 :
The origins of nominalizations cross-linguistically
Nominalizations have mainly been investigated in a synchronic perspective. The available diachronic evidence about the development of different types of nominalizations cross-linguistically poses some major challenges for traditional assumptions about nominalization in general. Nominalization is usually regarded as special strategy whereby some usually non referring expression is treated as a referring one. In many cases, however, the distinguishing structural properties of individual nominalization types are a result of the original structure of the source construction, rather than some special treatment of particular expressions. This, however, need not be the case for all nominalization types. This suggests that traditional criteria for nominalization do not actually capture a unified phenomenon, but rather a series of constructions that originate through different mechanisms and are motivated in terms of different principles.