A quantitative analysis of e/i variation in Old Irish etar and ceta

Abstract

Thurneysen (1946: 510, GOI §835) discusses the prefix/preposition/preverb <etar> (eDIL s.v. eter, etir), which has the following variants : etar, eter, etir, itar, itir, and itar. For simplicity, I will refer to this element as <etar>. I will label the variants as either (-)e-variants (etar, eter, etir), or (-)i-variants (itar, itir, itar). Thurneysen is uncharacteristically imprecise in discussing the variation. His commentary on the variation in the initial syllable can be summarized in three points, listed in ( 1): 1. a. (-)e-variants are common in the Würzburg and Turin Glosses, although (-)i-variants are also found. b. Before infixed pronouns in Milan one finds both etar and itir. c. In 'later' Old Irish (the St. Gall and the Karlsruhe Augustine glosses), etar and itar are more common.This essentially amounts to a list of random features amongst which no similarities can be observed. In a forthcoming paper, Griffith (2016a) clarifies the distribution of some of these variants, concentrating largely on the phonological history of the second vowel. He also notes that (-)i-variants of <etar> are uncommon and that the variant is limited to pretonic position (i.e., preverbs in deuterotonic compound verbs, and non-conjugated prepositions). In the Würzburg Glosses, he notes 1 out of 32 pretonic examples. In the Milan Glosses, he notes 5 out of 64 pretonic examples. 2 In the St. Gall Priscian Glosses, he notes 10 out of 36 pretonic examples. He interprets the rarity of the (-)i-variant and the 1 I gratefully acknowledge the help I received on various points in this paper from David Stifter, Aaron Griffith, and Bernhard Bauer. 2 I have a slightly different count for the Würzburg and Milan Glosses (for full details see the Appendix). I count 33 pretonic examples in Würzburg, with 1 (-)i-variant. I count 69 pretonic examples in Milan with 4 examples having an (-)i-variant.

The rise of gemination in Celtic

“…etar (cf. Lash, 2017 for other Old Irish allomorphs of it). This is expected for the preposition in pretonic position, but the short vowel is not expected in inflected forms of the preposition and when the preverb occurs in nominal compounds.…”

When did nd become nn in which Early Irish environments?

“…Concerning the fuller variant appearing in Sg., however, the anonymous reader suggests that 'given that Sg. is a layered text, with later and earlier strata (soStrachan 1903, Lambert 1996, this could be argued to belong to an earlier stratum. See perhaps alsoGriffith (2022) in the same vein' .10 For further examples, and counter-examples, seeThurneysen (1905, 6, 8, 10-12);Feuth (1982, 89-90, 93-4).…”