publicationsANALYSIS OF TEMPORAL PATTERNS IN SPEECH PRODUCTION
summary publications
In addition to my main focus -- i.e. the acoustic correlates
of quantity -- during the course of analyzing the data from Skolt
Saami I also studied certain phenomena in connection with temporal
organization in that language. While exploring the characteristics
of temporal patterns I was able to identify two issues which promise
to lead to investigations of some theoretical significance. These
are: (i) temporal patterns in relation to boundary signalling,
and (ii) temporal strategies associated with different speaking
modes.
Timing strategies apparent on the basis of paragraph studies
(see references below) show the following four tendencies: (i)
the employing or not employing of the vowel reduction rule, depending
on the distance from the boundary of the paragraph; (ii) shorter
absolute durations achieved while keeping significant segment
duration ratios constant; (iii) shorter word duration in paragraph-final
sentences; (iv) consistently shorter pause durations in paragraph-final
sentences.
In connection with examining the phenomenon of boundary signalling,
on the basis of my research so far the manifesting of temporal
patterns supports the hypothesis concerning the non-availability
of duration for boundary signalling in languages in which duration
plays a significant role in the prosodic system.
As far as the second issue is concerned, the acoustic analysis of durational patterns apparent in the spontaneous speaking mode in Skolt Saami suggests that there exists a definite hierarchy in terms of the employing of the four timing strategies listed above. It appears that (i) the constraint for maintaining characteristic durational ratios -- one of the four timing strategies -- overrides the strategy aiming at significant durational decrease close to boundaries; (ii) only three of the four strategies attested in the controlled experiment were consistently employed. The implications of these two tendencies observed in connected speech suggest that different temporal strategies may be associated with different speaking modes.
More recently, I have been investigating these issues in several
other languages, focussing on the realization of temporal stuctures
with regard to the phonological status of duration: i.e., whether
it has a contrastive or non-contrastive role in the language.
Current research indicates that languages with contrastive duration
tend not to utilize duration for additional functions in the grammar.
Accordingly, they behave differently with regard to boundary signalling
than do languages with non-contrastive duration. My most recent
experiments aim at exploring this difference in relation to temporal
patterns within the paragraph. It can be argued that there is
an effort apparent on the part of the speaker to time paragraph
constituents in a manner that evidences correlation of the temporal
organization of these constituents, and also that such temporal
strategies aimed at maintaining a target duration are realized
differently in the two language types. Boundary signalling with
durational increase co-occurs with a greater degree of variation
in duration within the paragraph in languages with non-contrastive
duration. In languages in which duration is contrastive little
durational change could be noticed in boundary positions, and
there is a lesser degree of durational variation.
Selected publications/presentations relevant to the analysis of temporal patterns in speech production include: