Ese Ejja Accent

Abstract

UC Berkeley Phonology Lab Annual Report (2013) Nicholas Rolle UC Berkeley “Ese Ejja Accent” 1 n.rolle@berkeley.edu Ese Ejja is a Takanan language spoken in the Bolivian Amazon. This study presents a description and analysis of stress accent on nouns and verbs, arguing for a specific representation of these data, and positing a limited OT account. Data come from Vuillermet’s (2012) PhD dissertation - hereafter referred to as (V:page number) - who presents an extensive description and organization of these accent facts, documenting approximately 2,000 verb forms. Most words in Ese Ejja have a single primary accent (synonymous here with “stress”), which falls on one of the first three syllables of the word, whose consistent phonetic correlate is high pitch. Ese Ejja has distinct accent patterns on nouns versus verbs, showing phonological patterns differing by part of speech. Accent placement depends on a complex interaction of factors, based on (1) inherent accent of a word, (2) accent assignment from affixes/clitics, (3) accent assignment based on part of speech, (4) rules of accent clash resolution, (5) rules of (trochaic) footing, (6) rules of directionality, and (7) restrictions on the primary stress (word stress) window. [Transcription note: I use σ to indicate primary stress, and σ secondary, rather than ˈ and ˌ ] Relevant phonology background Ese Ejja has four monophthong phonemes /i e a o/, and three (rising) diphthongs /i a i o oe/, written as , , and respectively; contrastive falling diphthongs (e.g. [ai ]) do not exist. Vowel length is not contrastive, and adjacent vowel-vowel sequences are pronounced separately (V:172). The canonical syllable structure is (C)V; underlying codas do not exist. The only surface codas are glides which result from vowel-vowel sequences, e.g. /mei/ “stone” [mej]. Cliticization reveals these vowels as underlying heterosyllabic, e.g. /mei=a/ “strone=INSTR” [meia] ( * [meja]) (V:177-8). Noun Accent 2 In order to understand both noun and verb accent, we must distinguish between three types of (surface) accents. One type is an inherent accent on a syllable, specified in the lexical entry of a morpheme (an unpredictable accent). A second type is assigned accent, an accent which is assigned to a specific part of a morpheme by an accompanying affix/clitic, by construction, and/or due to its use as a particular part of speech. A third type is footing accent, an accent which falls on a syllable based on the trochaic footing algorithm, applying subsequent to the first and second types. These are shown in example (0). In (0a), bishe has inherent accent on /she/. In contrast, in (0b), the clitic =nixe assigns an accent to the final syllable of the noun inawewa, which then based on footing assigns stress two syllables away, on /na/. This manuscript owes much to discussion with Marine Vuillermet (first and foremost), Zach O’Hagan, and Sharon Inkelas. These data come from (V:200-205). I will not discuss noun compounds (e.g. noun+adj), or the e-noun class, e.g. ese “tooth”, eshaxa “ear”, emekishe “nail”, in which the e- here is best analyzed as extra-metrical.