1234
V-V units
1 2345
V-V units
Figure 3: Normalized RT for click detection along positions in
stress groups of four V-to-V units. Whiskers indicate standard
error.
faster it is attended to. It means that perception facilitation is
somehow a function of time, what can be seen as a preliminary
evidence that question (2) can also have a positive answer.
The issue to be settled now is whether time is to be inter-
preted just as order (V-to-V position in stress group) or if the
perceptual entrainment is related to actual produced timing. In
order to tackle this problem, duration and RT data were corre-
lated.
Mean V-to-V duration of the four V-to-V units of the eight
sentences in group (A) were correlated to the corresponding
normalized RT means. The best correlation was achieved
through non-linear estimation using a polynomial function like
y = a + bx + cx2 yielding r = -0.764 (p < 0.009). Duration
data explain 58% of RT variance in this case.
As for group (B), means of position 1 to 3 were pooled over
and so were means of position 4 and 5 as suggested by Scheffe
homogeneous groups test. RT means were pooled over the same
way. Likewise group (A), the best correlation was the one we
got through non-linear estimation using the same function. In
this case r = -0.77 (p < 0.002) and a proportion of 59% of
RT variance can be accounted for by duration data.
4. Discussion
Correlation results seem to represent preliminary evidence that
speech perception and production patterns can in fact be closely
related, since almost 60% ofRT variance in the experiment can
be traced back to duration scaffolding in production. Future ex-
periments are to show if correlates of intonation play any role
in predicting perception. Besides that, they should also investi-
gate how semantic information helps listeners predict when and
where speakers will place stress along a sentence.
5. Acknowledgements
This work was supported by FAPESP grants 03/11619-0 and
05/02525-7. We thank Luciana Lucente and Sandra Madureira
for their help with intonational labelling.
Figure 4: Normalized RT for click detection along positions in
stress groups of five V-to-V units. Whiskers indicate standard
error.
6. References
[1] Collischonn, G., 1994. Acento secundario em portugues
brasileiro. Letras de Hoje, 29, 43-53.
[2] Said Ali, M., 1908. Difficuldades da Lingua Portugueza:
Estudos e Observacoes. Rio de Janeiro: Laemmert.
[3] Moraes, J. A., 2003. Secondary stress in Brazilian Por-
tuguese: perceptual and acoustical evidence. In Proceed-
ings of the 15th ICPhS. Barcelona, Spain, 2063-2066.
[4] Prieto, P.; van Santen, J., 1999. Secondary stress in Span-
ish: some experimental evidence. In Aspects of Romance
Linguistics. C. Parodi et al. (eds.). Washington: GUP, 337-
356.
[5] Barbosa, P. A.; Arantes, P.; Silveira, L. S., 2004. Unify-
ing stress shift and secondary stress phenomena with a
dynamical systems rhythm rule. In Proceedings Speech
Prosody 2004, Nara, Japan, 49-52.
[6] Barbosa, P. A., 2002. Explaning cross-linguistic rhytmic
variability via a coupled-oscillator model of rhythmic pro-
duction. In Proceedings of Speech Prosody 2002. Aix-en-
Provence, France, 163-166.
[7] Barbosa, P. A.; Bailly, G. 1994. Characterisation of rhyth-
mic patterns for text-to-speech synthesis. Speech Commu-
nication, 15: 127-137.
[8] Cutler, A.; Foss, D. N., 1977. The role of sentence stress in
sentence processing. Language and Speech, 20(1), 1-10.
[9] Martin, J. G., 1972. Rhythmic (hierarquical) versus ser-
ial structure in speech and other behavior. Psychological
Review, 79(6), 487-509.
[10] Quene, H.; Port, R. F., 2005. Effects of timing regular-
ity and metrical expectancy on spoken-word perception.
Phonetica, 62(1), 1-13.
[11] Large, E. W.; Jones, M. R., 1999. The Dynamics of At-
tending: How People Track Time-Varying Events. Psy-
chological Review, 106(1), 119-159.