Do All Languages Have Word Accent? Or: What's So Great About Being Universal?

Abstract

UC Berkeley Phonology Lab Annual Report (2010) Do All Languages Have Word Accent? Or: What’s so great about being universal? Larry M. Hyman University of California, Berkeley Draft of Paper presented at the Conference on Word Accent: Theoretical and Typological Issues, University of Connecticut, April 30, 2010 • Comments Welcome The purpose of this paper is to address the question: Do all languages have word accent? By use of the term “word accent” I will first consider the traditional notion of “stress accent” at the word level, as so extensively studied within the metrical literature. However, I will also ask whether certain other phenomena which privilege a single syllable per word should also be identified at some higher level of abstraction as “word accent”, whether on a par with stress, or different. There of course have been claims that all languages have word accent, e.g. most recently: “A considerable number (probably the majority, and according to me: all) of the world's languages display a phenomenon known as word stress.” (van der Hulst 2009:1) On the other hand, a number of scholars have asserted that specific languages lack word stress. This includes certain tone languages in Africa, but also languages without tone. In Bella Coola there is “no phonemically significant phenomena of stress or pitch associated with syllables or words.... When two or more syllabics occur in a word or sentence, one can clearly hear different degrees of articulatory force. But these relative stresses in a sequence of acoustic syllables do not remain constant in repetitions of the utterance.” (Newman 1947:132). In fact, many languages do not provide unambiguous evidence of word accent. In many cases the interpretations have been theory-dependent and highly personal: Some people see (and even hear) stress where others don’t. Given this fact, it will be extremely difficult to demonstrate to the satisfaction of all that stress either is or isn’t universal. The approach that I advocate is what I call “properties-driven typology” (PDT). Whereas word-prosodic typology has been concerned with pidgeon-holing languages and giving them names such as “tone”, “stress” and “pitch-accent” (see Hyman 2006, 2009a for a critique of this practice), the approach of PDT is to eschew this concern and typologize on the basis of the individual properties which may or may not satisfy preconceived definitions and prototypes. In the current context focus is on the properties that one vs. another language manifests which have stress-like qualities. Thus, (1) summarizes what I earlier identified as a possible “prototype” of a stress system, namely one which, like English, piles on property after property which unambiguously point to stress (Hyman 2009a:217) : a. b. stress location is not reducible to simple first/last syllable stressed syllables show positional prominence effects i. consonant-, vowel-, and tone oppositions are greater on stressed syllables ii. segments are strengthened in stressed syllables (e.g. Cs become aspirated or geminated, Vs become lengthened, diphthongized)