Yalálag Zapotec
-Phonetics
The Phonetic Correlates
of Yalálag Zapotec.
Zapotec languages have been described as making a contrast between fortis
and lenis consonants in both obstruents and sonorants, but it seems that
there is not a single phonetic feature accounting for the phonological
contrast. Instead, the descriptions often mention notions such as articulatory
force and duration as the basis of the difference. In this paper
I present the results of an acoustic phonetic analysis of the fortis/lenis
consonants looking at several parameters such as durational properties
and acoustic amplitude in stops bursts and in the portion of the adjacent
vowel following fricatives and sonorants. The results obtained indicate
that the chief characteristic differentiating fortis and lenis consonants
is length. The results show that the obstruents in Yalálag Zapotec
can be defined by parameters as VOT and voicing rather than fortis/lenis
features.
Functional Specialization in Tone Perception:
Evidence from Dichotic Listening in Yalálag Zapotec. There is
an important tradition of clinical and experimental studies suggesting
that there is a left hemisphere specialization substrate for linguistic
functions whereas the right hemisphere is less involved in linguistic tasks.
(Fitch, Miller & Tallal, 1997, Kraemer 1992, Liberman 1973, Moffat
and Hampson 2000, among others). With respect to the processing of fundamental
frequency some studies have proposed that the left hemisphere plays a role
in the processing of of lexical tone (Gandour and Dardarananda 1983,
Van Lancker and Fromkin 1973, 1978). However, some other investigations
claim that the right hemisphere is specially involved in processing stimuli
based on fundamental frequency such as music, pure tone and intonation
(Blumstein & Cooper, 1974, Efron et al. 1977, Perkins et al. 1996,
Shipley-Brown 1988). Because the acoustic parameter underlying the
two classes of stimuli is the same (i.e. fundamental frequency) the study
of functional specialization in the perception of F0 in languages with
contrastive use of pitch represents a crucial evidence on the issue. So,
in this paper I investigate the patterns of hemispheric specialization
of tone perception in Yalálag Zapotec, a tone language, using a
dichotic listening technique. Two experiments were designed: In the first
experiment minimal pairs of words contrasting in tone were tested; in the
second, dichotic pairs of pure tones (hums) with similar frequency values
to those found in words were tested. In both cases, the condition of selective
attention was introduced (Asbjørnsen and Hugdahl 1995). The results
indicate a significant rigth ear advantage for phonemic tone while there
is no correlation between the perception of pure tones and ear advantage.
Further, selective attention increased the pattern of lateralization in
word stimuli. The pattern of asymmetries obtained supports a functional
lateralization hypothesis (Van Lancker 1980) where a cognitive linguistic
task, i.e. the perception of phonemic tone, involves a left hemispheric
dominance.
Categorical Perception of Phonemic Tone
in Yalálag Zapotec. Previous investigations on the perception
of lexical tone in human languages have reached contradictory conclusions
about its nature. Some studies suggested that the perception of tone is
categorical (Chan et al. 1975). Some other studies claimed that categorical
perception cannot be supported (Abramson 1961, 1977; Francis et al. 2000).
This research investigates the nature of the perception of tone in Yalálag
Zapotec (YZ) (Otomanguean, Mexico), a language with phonemic contrasts
between high, low and falling tones. Two experiments were designed where
subjects, native speakers of YZ, were asked to discriminate and identify
stimuli consisting of sets of resynthesized (PSOLA) variations of 32 words
(16 minimal pairs) in which the original f0 was varied in intervals of
5 Hz (661 tokens). All other spectral characteristics remained the
same. In the discrimination experiment the subjects were presented with
pairs of auditory stimuli and asked to judge whether the pairs were same
or different. In the identification experiment, the subjects were presented
with a series of a visual stimulus (a pair of pictures evoking a minimal
pair) and an auditory stimulus (a resynthesized version of one of the words
of the lexical minimal pair). The subjects were asked to select the picture
that better matched with the auditory stimulus. The results showed clear
category boundaries for the level tones, high and low, in both discrimination
and identification tasks, with respect to the accuracy and reaction time
variables. The results on accuracy and reaction time for falling tone were
overall more complex than those for level tones. For falling tone the results
varied according to whether the begining or the end of the f0 contour was
manipulated. First, when the upper end of the contour was lowered there
was a reliable categorization to the extent that the whole trajectory exceeded
a frequency range similar to that of the low tone. Second, when the lower
end of the contour was raised, there was an enhanced categorization as
falling tone when the trajectory showed a minimum slope of D15 Hz. Overall,
the results suggest that the perception of tone in YZ is categorical. The
results also indicate that the perception of the contour tone depends on
a trajectory ranging from high to low frequencies in the perceptible space
of the listener than an absolute magnitude of slope.