Relative priming of temporal local-global levels in auditory hierarchical stimuli
TIMOTHY JUSTUS
0
0
University of California
,
Berkeley
,
California and Medical Research Service, Department of Veterans Affairs Medical Center
, Martinez,
California
Priming is a useful tool for ascertaining the circumstances under which previous experiences influence behavior. Previously, using hierarchical stimuli, we demonstrated (Justus & List, 2005) that selectively attending to one temporal scale of an auditory stimulus improved subsequent attention to a repeated (vs. changed) temporal scale; that is, we demonstrated intertrial auditory temporal level priming. Here, we have extended those results to address whether level priming relied on absolute or relative temporal information. Both relative and absolute temporal information are important in auditory perception: Speech and music can be recognized over various temporal scales but become uninterpretable to a listener when presented too quickly or slowly. We first confirmed that temporal level priming generalized over new temporal scales. Second, in the context of multiple temporal scales, we found that temporal level priming operates predominantly on the basis of relative, rather than absolute, temporal information. These findings are discussed in the context of expectancies and relational invariance in audition.
-
Prior experiences can influence present behavior in
many ways. Here, we focus on primingthat is, changes
in behavior based on repeated relative to novel
experiences. Frequently, the repetition of an experience leads
to improvements in behavior, although in certain cases,
repetition can impair behavior (e.g., negative priming;
Tipper, 1985). Priming between the initial and repeated
experience may be observed across a broad range of
temporal delays: on the order of seconds (e.g., on a
trial-bytrial basis), hours, days, or months (e.g., within or across
testing sessions). Note that any change in behavior related
to the repetition or novelty of a situation indicates an
influence of previous events. For this reason, priming is a
powerful experimental tool for understanding what
experiences or aspects of those experiences are important in
determining behavior over time.
In the present study, we examined how the current
processing of an auditory stimulus is influenced by the
processing of the immediately preceding auditory stimulus
(i.e., auditory priming). To situate the present study, we
first describe some of the rich literature on how
preceding experiences influence current auditory processing.
The most straightforward example of prior influence
on current processing is repetition, or sensory,
priming: When perceivers are presented with one stimulus,
performance is enhanced for the same ensuing stimulus.
One illustrative example from audition comes from Ward
(1997), who showed that presenting two tones in
succession of the same (vs. different) frequency improved
performance on an orthogonal intensity judgment task (see
also, e.g., the uninformative cue conditions in Green &
McKeown, 2001).
There are also more sophisticated forms of auditory
priming beyond those found with stimulus repetition. For
example, Bharucha and Stoeckig (1986) demonstrated
priming between two harmonically related vs. unrelated
chords: Performance was better for harmonically related
chords. This priming advantage was not reliant on
repetition priming, as was demonstrated by its persistence even
when any overlap in frequency spectra was removed from
the chord pairs (Bharucha & Stoeckig, 1987). Tekman and
Bharucha (1998) demonstrated a compelling extension of
this finding: Priming was stronger between chords with
a close harmonic relation, even when the more distant
harmonic relation contained greater overlap in frequency
spectra. The last example reveals that repetition priming
can be overcome by other forms of priming.
Bigand and colleagues further elaborated on the
priming paradigm by demonstrating harmonic priming using
a temporally extended frequency context; that is, within
a trial, several events set up a tonal context that affected
the processing of the final event (e.g., Bigand & Pineau,
1997; Bigand, Poulin, Tillmann, Madurell, & DAdamo,
2003; Johnston & Jones, 2006; Marmel, Tillmann, &
Dowling, 2008; Tillmann, Bigand, & Pineau, 1998;
Tillmann & Lebrun-Guillaud, 2006). Namely, judgments
of, for example, completion, timbre, or mistuning were
improved if the final event was harmonically related to the
preceding context relative to when it was not. Together,
these examples of priming effects illustrate that lifelong
experiences with tonal schemata induce listeners to create
expectations for temporally proximal auditory events to
be harmonically related.
The creation of expectations by a temporally extended
context need not rely on tonal schemata, however. Other
demonstrations exist using, for example, good
continuation of pitch (i.e., each successive tone is higher or
lower). For example, Johnston and Jones (2006,
Experiment 1) presented listeners with a (...truncated)