--------------------------------

Attila Starčević
Eötvös Loránd University (Elte), English Linguistics Department, Budapest, Hungary
Email: mazanydella at hotmail dot com

The Germanic Foot Reinterpreted

The aim of the talk is to discuss Middle English open syllable lengthening (MEOSL), as well as to address the issue of whether there was MEOSL at all, as understood traditionally, and how it relates to the Germanic foot (Dresher & Lahiri 1991). A discussion of some overlooked facts on MEOSL will be offered (Dutch and German show a similar process: cf. Prokosch 1933, Lahiri & Dresher 1999). MEOSL has been treated from various perspectives over the last few decades (see e.g. Minkova 1982, 1985, 1991, Ritt 1994). The issue itself is not recent, of course (cf. Luick 1914 for one of the first attempts to relate all of the ME quantity changes to a single cause, and Ritt 1994 for references).

OE short vowel underwent a series of changes in open syllables that resulted in a set of new long vowels in ME (e.g. clĭpian > clēpen, mĕte > mēte, ălu > āle, bŏta > bōte). Generally, only those OE words underwent MEOSL unfailingly that had the C 0VCV ‘template’: tălu > tāle. Here, ‘template’ is not used in the sense of a morphological template known from the Semitic languages but rather as a skeleton relating to alternating vocalic and consonantal positions. However, words of the C 0VC ‘template’ show either a long or a short vowelled reflex (e.g. hwæ ( l > whāle vs. hlŏt > lŏt), as do words with the C 0VCVC ‘template’ (sădol > săddle vs. crădol > crādle). This was traditionally attributed to analogy originating in the declinational characteristics of the period: e.g. ME sādel < sădol (with the usual ‘weakening’ of the OE unstressed vowels in schwa and MEOSL) was influenced by trisyllabic laxness from one of its oblique forms which was later lexicalised: sădeles ‘saddle GEN’ (obviously, this re-lexicalisation did not take place in the case of ME crādel ~ crădeles > crādle).

The overlooked issues of MEOSL include the following: (i) why was there no MEOSL in words of C 0VCV ‘template’ whose second vowel was not / « / (the only two eligible vowels are ME / i / and / u /: e.g. belly/narrow (< OE belig/n(e)aru) and (ii) why where the original OE long vowels shortened before the very same vowels in mono-morphemic words: e.g. OE sārig > sŏrry, mæ # dwe /mæ # du > mĕadow. It seems that MEOSL is not solely about having the original OE short vowels lengthened in open syllables. Murray & Vennemann (1983) appeal to the “Stressed Syllable Law” which states that the preferred Germanic stressed syllable contains exactly two morae and that this drive is found in the West Germanic languages. This is problematic because both early Germanic (* fatam) and ME had stressed syllables that weighed only one mora ( belly), underlined here. Dresher & Lahiri (1991) claim that there was indeed such an ideal but this preferred weight could be drawn from both the syllable and the foot. This leads to the postulation of a resolved Germanic foot and the equivalence of H = L X (X for either H(eavy) or L(ight)). Here, H is (or can as well be) drawn on the syllable level and L X on the foot level. This resolved foot is claimed to explain a number of Germanic processes like Siever’s Law in Gothic or High Vowel Deletion in OE.

It will be suggested that MEOSL was in fact a templatic change. The analysis will highlight some of the issues connected with a CV/VC approach to MEOSL. This framework which has grown out of Government Phonology does not recognise the syllable as a theoretical construct (cf. Scheer 2004), yet it offers some new insight into ME.

References