250
D.L. Herzing
regulate the cadence of a message that may signal an increase or decrease
in vigilance or aggression, as in prairie dog vocalizations or in human narrative (Hauser 1996). Considering the above, and the individual, gender,
and relational information overlaid in such signal sequences, complicated
decoding such as, "My associate X is escalating his conflict with dolphin Y.
I hear no other coalition partners around synchronizing their vocalizations
with his, therefore my action could be to help," seems plausible during longdistance scenarios.
In human communication, both structural design features and their perceptual salience playa role in decoding human speech (Hauser 1996).
Human voices convey information about objects and emotional or affective states. If voices encode information about emotional states, exposure
to this information is likely used by individuals to assess social situations.
The actual form of the vocalization may contain intent in its structure, as
in human speech (Clynes 1977). Such stereotypic prosodic contours of
affective states provide consistent information for individuals in a communicative setting. Prosodic contours provide information about affective
states while referential information resides in perceptually discrete units of
sound. In the human language, such aspects as "voice onset time" are critical in separating functional units of vowels and consonants. Such features
can also be discriminated across species (Kuhl and Miller 1978). Since the
pioneering work in primate alarm calls (Seyfarth et al. 1980), such subtle
acoustic features of nonhuman animal calls have been proposed as a place
where information, salient and perceivable by the species, may reside.
However, methodological issues of sampling dolphin behavior in the contexts with potential referential information (feeding and predators: see
Hauser 1996, p. 520 ) are still difficult to obtain.
If we attempt to move beyond the analysis of single calls and into
sequential analysis (time-dependent modulation of parameters, or timeindependent aspects of the order of signals), we begin to look at the longterm ordered relations between signal use and its determinant behavior,
such as changes in arousal level or even semantic meaning. Signal relations
have been discussed on theoretical grounds (Johnson 1993). The inherent
assumption of the independence of signals in restricted time is limiting
for signal relations analysis.
In addition, the simultaneous perception of sound is also of relevance in
signal relations. Au (1993) discusses the possibility that dolphins may perceive a "time separation pitch" based on differences in highlight arrival,
thus allowing a holistic perception of a sequence of sounds. This is also
reported by Bregman (1990) in his discussion of auditory scene analysis by
both sequential and simulataneous processing by the auditory system. Dolphins have great temporal resolving abilities (Johnson et al. 1988), and
whether dolphins perceive this change as significant is unknown, but such
changes in rates or differences in onset time and cessation are viable cues
available to individuals (Barrett-Lennard et al. 1996).
D.L. Herzing
regulate the cadence of a message that may signal an increase or decrease
in vigilance or aggression, as in prairie dog vocalizations or in human narrative (Hauser 1996). Considering the above, and the individual, gender,
and relational information overlaid in such signal sequences, complicated
decoding such as, "My associate X is escalating his conflict with dolphin Y.
I hear no other coalition partners around synchronizing their vocalizations
with his, therefore my action could be to help," seems plausible during longdistance scenarios.
In human communication, both structural design features and their perceptual salience playa role in decoding human speech (Hauser 1996).
Human voices convey information about objects and emotional or affective states. If voices encode information about emotional states, exposure
to this information is likely used by individuals to assess social situations.
The actual form of the vocalization may contain intent in its structure, as
in human speech (Clynes 1977). Such stereotypic prosodic contours of
affective states provide consistent information for individuals in a communicative setting. Prosodic contours provide information about affective
states while referential information resides in perceptually discrete units of
sound. In the human language, such aspects as "voice onset time" are critical in separating functional units of vowels and consonants. Such features
can also be discriminated across species (Kuhl and Miller 1978). Since the
pioneering work in primate alarm calls (Seyfarth et al. 1980), such subtle
acoustic features of nonhuman animal calls have been proposed as a place
where information, salient and perceivable by the species, may reside.
However, methodological issues of sampling dolphin behavior in the contexts with potential referential information (feeding and predators: see
Hauser 1996, p. 520 ) are still difficult to obtain.
If we attempt to move beyond the analysis of single calls and into
sequential analysis (time-dependent modulation of parameters, or timeindependent aspects of the order of signals), we begin to look at the longterm ordered relations between signal use and its determinant behavior,
such as changes in arousal level or even semantic meaning. Signal relations
have been discussed on theoretical grounds (Johnson 1993). The inherent
assumption of the independence of signals in restricted time is limiting
for signal relations analysis.
In addition, the simultaneous perception of sound is also of relevance in
signal relations. Au (1993) discusses the possibility that dolphins may perceive a "time separation pitch" based on differences in highlight arrival,
thus allowing a holistic perception of a sequence of sounds. This is also
reported by Bregman (1990) in his discussion of auditory scene analysis by
both sequential and simulataneous processing by the auditory system. Dolphins have great temporal resolving abilities (Johnson et al. 1988), and
whether dolphins perceive this change as significant is unknown, but such
changes in rates or differences in onset time and cessation are viable cues
available to individuals (Barrett-Lennard et al. 1996).