Coordination and Agreement
In chapter 3 I argued that coordination involves a two identical projections that host the conjuncts in their respective specifiers. This analysis predicts that certain phenomena should affect only one of those segments. In this chapter I will argue that partial agreement is an operation restricted to one of the specifiers. Partial agreement displays two important characteristics. There is a typological correlation regarding word order and a division that depends on interpretive possibilities. After presenting the data, I will turn to the analysis of the different patterns in the current framework.
KeywordsSuffix Cali Percolate
Unable to display preview. Download preview PDF.
- 65.Unless otherwise stated, Arabic examples come from Aoun, Benmamoun and Sportiche 1994.Google Scholar
- 71.All of these examples are based on the ones collected by Fernández Ramírez (1995) from written sources. I have made minor modifications in some of them. Grammaticality judgements are my own.Google Scholar
- 77.(19)c is acceptable with a different structure, what Milner 1987 calls interpretive chains.Google Scholar
- 81.In fact, this is only one of the solutions ABS suggest, another one is some version of Williams’s (1978) across-the-board movement, which will extract the verb simultaneously from both clauses, although this by itself does not account for agreement, as ABS point out.Google Scholar
- 84.Munn (1999) gives a grammatical example of ser and partial agreement, but this involves a clefted structure:fui eu e as meninas que compramos us flares ‘it was I and the girls that bought the flowers.’ Agreement patterns in clefts are not necessarily the same as in non-clefts.Google Scholar
- 87.Munn (1999) claims Spanish has the same distribution as Brazilian Portuguese Native speaker judgements (including my own), suggest this is not the case.Google Scholar
- 94.Brazilian Portuguese shows a similar pattern, according to Munn (1999).Google Scholar
- 95.Munn (1999) argues that the contrast between full agreement prenominal elements and partial agreement postnominal elements reflects a difference between spec-head agreement and government agreement. His examples involve demonstratives vs. determiners in languages like English. However, in languages where determiners have overt number and gender, it is clear that there is no contrast between demonstratives (in spec, DP) and determiners (in D).Google Scholar
- 97.See Kayne (1994) for a similar analysis.Google Scholar
- 100.An anonymous reviewer points out the contrast between (115)b and (i), which in turn contrasts with (ii). Clearly, other factors are involved in this distribution that are not relevant for he discussion (see Laca 1999): (i.) No tengo lápiz. not have pencil ‘I don’t have a pencil.’ (ii.) *Tengo lápiz. have pencilGoogle Scholar
- 103.Kany (1951) observes that the Southern Cone dialects do have the pattern V-DAT-ACC-AGR. Curiously, dialect A described earlier, which does not distinguish datives and accusatives in this way, belongs to this region.Google Scholar
- 105.This configuration is only available for third person clitics. Many people have pointed out asymmetries between third person clitics and first/second person clitics, see for example Uriagereka (1995) and Torrego (1996). In this particular case, the explanation for the asymmetry could be compatible with a theory that restricts mismatches to third person features. However, other evidence may suggest that the syntactic structure of first/second and third person is quite different, as both Uriagereka and Torrego have suggested.Google Scholar