Volume 50, Issue 3-4 - Speaker judgments alone cannot diagnose syllable structure

The ease and consistency with which speakers of many languages provide direct judgments about syllable structure has been taken by scholars as evidence that these judgments are accurate and sufficient argumentation for an analysis of syllable structure in descriptive works. This paper questions whether the results of direct elicitation tasks reliably indicate a prosodic domain that is meaningful in the language’s phonology and provides alternative forms of evidence for syllable structure from intonation. This is done through a case study of Budai Rukai, a Formosan language whose contact relationship with Sinitic languages affects how speakers respond to syllable judgment tasks.

This commentary offers evidence for three general language types that may lack evidence for syllables as prosodic constituents: mora-dominant languages; foot-dominant languages; and word-dominant languages. It also highlights the important role of native-speaker judgements in identifying these language types.

This target paper builds on previous literature in Budai Rukai to develop a nuanced framework for examining the syllable structure and the impact of Chinese language education on speakers’ judgments. It demonstrates new dimensions of overarching corroboration of intonational phonology and is a worthwhile endeavor. In this commentary, I address the issues of elicitation tasks and language contact by illustrating that the syllable structure of Budai Rukai is still undergoing age-related changes and remains to be discussed.

Linguistic elicitation is a kind of behavioral experiment. It remains an important source for phonological (and more generally, grammatical) argumentation even if it does not meet all the requirements for a proper behavioral experiment (in terms of the number of participants, for example). While the basic analysis proposed for Budai Rukai syllable structure is convincing, questions arise with regard to how the very basic notions syllable and stress are interpreted in the target article. Furthermore, one would wish for a more detailed argument regarding the analysis of echo vowels and phonetic glides.

This commentary paper offers a discussion on the methodological, epistemological, and theoretical issues concerning the status of the syllable and mora, as well as the problems raised by their (mis)characterization. It first addresses some methodological biases that occur in the description and analysis of the phonological units of languages, the problem of the reliability of speaker intuitions, with an example taken from French verlan, as well as the influence of writing on phonological representations. Second, building on the Budai Rukai example and on one other case of misanalyzed syllable structure, that of Tokyo Japanese, it questions the status of the syllable as an indispensable, universal unit. In order to avoid some of the problems that stem from the mischaracterization of the syllable, it is proposed that the notion of prosodeme, instead of that of syllable, should be used as a pretheoretical notion when attempting to identify the basic prosodic unit of a language.

Macaulay has performed a great service by pointing out unfounded leaps of logic in the syllable sections of descriptive grammars, but the auxiliary hypotheses that underlie such leaps require close scrutiny everywhere, including in his own argument: intuitions are more reliable than he claims, intonation is less reliable than he assumes, inconsistencies cannot be resolved by simply ignoring certain types of evidence, and innateness considerations qualify all linguistic argumentation.

Prosodic details offer valuable insight into the phonology of languages, but prosodically-grounded analysis alone does not reveal the whole picture. Benjamin Macaulay’s prosodic study into Rukai provides valuable insights into the syllable structure of the language, but phonological alternations, particularly the alternation of glides and fricatives, as well as the historical source of glides, suggests that glides still form an important part of Rukai’s phonology. Rather than doing away wtih glides, as Macaulay suggests, this study proposes a compromise position that acknowledges the utility of Macaulay’s prosodic analysis in many cases but keeps the glides [j] and [w] as part of the phonology of Rukai.

Benjamin Macaulay advises us not to diagnose syllable structure, more specifically the distinction between vowel + vowel clusters and diphthongs in the Taiwanese language Budai Rukai, on the basis of speaker judgements. I fully endorse his approach, since much of current work in phonology unwarrantedly takes syllabification for granted, without arguing for its necessity.

Using the Popperian three-world theory, we draw a general picture of linguistic variables in linguistic data elicitation, judgement, and analysis. To cope with these three worlds of variables, we have two suggestions, namely careful variable control and bigger data size, as well as two reminders with respect to indiscriminate data and proportional conclusions.

This reply explores issues with documenting prosodic structures given their unique relationship with metalinguistic knowledge. New evidence and perspectives are incorporated into a deeper investigation of Budai Rukai and the analytical decisions that underlie its reanalyzed prosodic system. This case study is then taken as a basis for discussion of a possible standardized protocol for the documentation of prosodic structures, in which diverse types of phonological evidence are integrated and contextualized.