Inflection and Movement in Old English

Rich verbal inflection has been argued (a) to license pro-drop (e.g. Safir 1985); (b) to be connected to overt Verb-second (Platzack & Holmberg 1989; van Gelderen 1993; Roberts 1996); and (c) to activate the AGR(eement) and T(ense) positions (Bobaljik & Jonas 1996). The connections in (b) and (c) follow if one assumes that the relationship between morphology and syntax is a direct one: elements move because they need to `pick up' inflection in higher functional categories (cf. Baker's 1988 Mirror Principle). There are problems with each of the claims of (a) (c) and the issue is therefore relevant to the present collection. On the one hand, languages exist where inflection is rich but where the three phenomena do not (all) occur (German and French have rich inflection but no pro-drop); on the other hand, languages with weak inflection can, for instance, have Verb-second (Afrikaans and Yiddish). In this paper, I focus on Old English verbal inflection in relation to (b) and (c) in Beowulf, The Junius Manuscript and The Exeter Book. Many of the arguments hold for other Germanic languages as well. I show that, even though the inflection is rich in Old English, the language has some reduced inflection when the verb is in second position. This is unexpected but in accordance with Chomsky (1995) who does not connect morphological strength with the feature strength that triggers movement. I argue this reduction indicates that, in the `normal' Old English clause, agreement is checked inside VP but that if verbs move to second position, the checking may occur in C resulting in incomplete agreement. Following Kiparsky (1995), I claim that Old English has limited Verbsecond because the C position is not generally available. In addition, there is no evidence for AGR and T positions and therefore, Old English is a problem for Bobaljik & Jonas (1996). I suggest that positional, rather than morphological, evidence triggers Functional Categories. The outline is as follows. In 1, I provide some background on the impact of rich inflection in some languages. In 2, I show that there is no link between rich inflection on the one hand and movement and the presence of functional categories such as T and AGRs on the other in Old English. In 3, I discuss some later developments regarding the loss of inflection and the introduction of functional categories such as C, AGRs and T.