Phonological units underlying speech production: Acoustic or articulatory goals?

Rachid Ridouane (1)

1 LPP - LPP - Laboratoire de Phonétique et Phonologie - UMR 7018

Speech utterances are made up of units such as features, phonemes and their associations into syllables. These units, which have achieved a wide success in the domain of phonological description, may also be central to the cognitive encoding of speech, which allows the variability of the acoustic signal to be related to a small number of categories relevant for the production and perception of spoken languages. In spite of the fundamental role that these units play in current linguistics, current research continues to raise many basic questions concerning their cognitive status and their role in speech production and perception. The present talk will be focused on three types of speech units: features, phonemes and syllables, and questions the relation they have to measurable physical properties in the articulatory and acoustic/auditory domains. Where in speech chain are these units best defined? Are they articulatory or auditory/perceptual? I will first review the most well-known hypotheses, by drawing selectively on the huge body of literature accumulated over the last sixty years. Then I will present results from three studies, based on data from Tashlhiyt Berber (in addition to some other unrelated languages). The first study addresses the question of how distinctive features are phonetically implemented and proposes a language-independent definition of the feature [spread glottis]. It will be shown that the definition of this feature in terms of a single common articulatory configuration or gesture would be insufficient to account for the full range of speech sounds characterized by this feature. For this feature to be recovered, its defining acoustic attribute, or failing that its enhancing attributes, must also be present in the signal. The second study deals with the production and perception of initial voiceless geminated stops. Conceivably, these stops may only slightly differ from their singleton counterparts in word initial position since the main correlate of geminated stops is a longer closure which, when voiceless, translates acoustically into a silent gap whose duration cannot be perceived in this position. Do speakers nonetheless produce the length contrast? And do listeners perceive this distinction? It will be shown that although the contrast is systematically encoded articulatorily, available acoustic information is not sufficient to elicit the standard perception performance expected from native listeners on a native contrast. The third study deals with syllables without vowels. I will first provide some data on how these syllables are articulatorily implemented and then tackle the issue of their acoustic make-up. It will be shown that the vocalic elements that are sometimes present within such syllables are mere transitional elements with no structural relevance: they are the automatic outcome of how laryngeal and supralaryngeal specifications of consonants are coordinated. Interestingly, native speakers are most often unconscious of the presence of these elements in their speech. This shows that the knowledge of the phonological system of the language, including its phonotactic constraints, biases the listener toward disfavoring certain percepts that are inconsistent with the syllabic structure of the language. In sum, while the first study shows that both acoustic and articultaory structure of speech units should be incorporated into the definition of phonological features, the second study provides a case of articulation without acoustics, and the third study illustrates a case of acoustics without articulation. The problem then is probably not that of making a strict choice between the articulatory vs. acoustic nature of the primitives of speech. These facts indicate that: “the question of where in the speech chain units are best defined is probably a spurious one” (Lindblom 2004). Lindblom, B. (2004). The organization of speech movements: Specification of units and modes of control. In Slifka J et al (eds): From sound to sense: 50+ years of discoveries in speech communication, Research Laboratory of Electronics, MIT, Cambridge, Mass

Mots clés: Speech production/perception; Units of planning

Domaines: Linguistique

Rachid Ridouane. Phonological units underlying speech production: Acoustic or articulatory goals?. Summer School CPMSP2 - 2010 “Cognitive and Physical Models of Speech Production, Speech Perception and Production-Perception Interaction” Part III: Planning and Dynamics, Sep 2010, Berlin, Germany. ⟨halshs-00523441⟩