COARTICULATION IN CATALAN VCV SEQUENCES: AN ARTICULATORY AND ACOUSTICAL STUDY (SPAIN)

In this study I investigate the existence of invariant articulatory mechanisms in the process of speech production. For this purpose, the relationship between degree of constraint on tongue dorsum to achieve dorsopalatal contact and coarticulatory effects is examined for a selected set of Catalan consonants in VCV sequences. It is hypothesized that the degree to which the consonant allows vocalic effects to occur is dependent on the degree of constraint involved in the constriction gesture. Palatal {j}, alveolo-palatals { }, {(lamda)}, dentals {t}, {(PAR-DIFF)} and alveolars {n}, {l} are investigated to see whether this relationship occurs systematically for consonants involving contrasting degrees of tongue-dorsum constraint. Moreover, coarticulatory effects for alveolo-palatals { }, {(lamda)} (produced with simultaneous alveolar closure and palatal closure) vs. clusters of alveolar + {j} (produced with alveolar closure followed by palatal closure) are also investigated to see whether contrasting timing constraints involving tongue-dorsum raising affect coarticulation.^ Electropalatographic and acoustical data have been collected. The articulatory traits and coarticulatory effects for all consonants in all VCV environments (for V = {i}, {a} and {u}) have been measured at the point of maximum linguopalatal contact and over time.^ Results show that V-to-C and V-to-V coarticulation varies systematically with spatial constraint (for palatals vs. alveolo-palatals vs. dentals and alveolars) and with temporal constraint (for alveolo-palatals vs. clusters) on tongue dorsum to achieve dorsopalatal contact. Thus, an inverse and monotonical relationship has been found to hold between degree of coarticulation and degree of dorsopalatal contact; on the other hand, an inverse relationship has been found to hold between degree of coarticulation and delay of the period of dorsopalatal closure relative to the period of alveolar closure. These findings argue for relational invariant units in the production process between context-free and context-dependent articulatory specifications. Data reported in this study argue for anticipatory effects being central and carryover effects being mechanical, for coarticulation being regulated by synergy constraints upon different regions of the tongue and, in summary, for coarticulation being a context-dependent process.