| The movable s defies distributional predictability, within Indo-European and later daughter-languages. Word-initial s-/zero alternations occur across every IE dialect-group, apparently without conditioning or living morphological function, nonetheless in certain languages--chiefly Germanic and Balto-Slavic--exhibiting unexpected, regenerative durability. Enigmas surrounding its morphophonemic origins have eluded satisfactory resolution. Phonologically-conditioned sentential sandhi, involving *s-degemination across word-boundaries by sound-rule, alongside metanalysis (n-orange), does not adequately account for the data's nature and spread, even with the help of analogical extension and hypercorrection (accretion/loss).;Chapters 1-2 deal with phonological-distribution issues; Chapter 3 explores origin, character and function. IE s-movable's original and subsequent nature is investigated for possible morphological explanations, particularly as applied to root-structure constraints, to motivate and classify the doublets' structurally anomalous behavior. Initial ;S-loss primacy is borne out by the secondary, successive continuation of s-movable into Germanic. The chronological layering of Germanic evidence relative to the Sound Shifts supplies an unusual vantage-point for observing the process over time. Chapter 4 views s-mobile's later development against the extraordinary preponderance of material specifically from the Germanic dialects, morphologically analyzed case-by-case. To questions of word-boundaries' non-salience, OHG evidence (Celtic, Indic) for Auslaut-conditioned word-initial voicing alternants (Notkers Anlautgesetz) suggests that the IE sentence-string was a linguistically significant unit surprisingly late. Stabreim, binomial pairs, phonetic and prosodic attributes, and expressive (affectively-flavored) features contribute to the composite Germanic picture. S-movable's survival and successive renewal are considered in the light of evidence for Germanic's dialectal archaism within IE.;Chapters 5-6 discuss the detailed Germanic, Baltic, and pan-IE evidence against relative chronology and phrasal-domain hierarchies (word-boundaries' porosity), and against the cross-linguistically marked, unstable phonetic character of sibilants, particularly their clusters. Beside the IE's sibilant's phonological isolation are set S-clusters' phonemic extraneousness, late acquisition, and ready simplifiability, in child-language and aphasia. |
