Temporal expectations and their violations
Olivia Ladinig

Abstract:
Temporal expectations and their violations
Olivia Ladinig

This thesis contains a collection of papers that investigate temporal
expectations.

Responses that indicate that a stimulus is perceived as unnatural,
complex, or surprising, as opposed to a stimulus being perceived as
natural, simple, or unsurprising, are taken as indicators that
expectations were not fulfilled but rather violated. The term
~expectation~ implies an active role of the listener, who constantly
predicts what events will happen at what time in the future. The more
confident the predictions are, the more will an outcome that is
different to what was predicted lead to the violation of an
expectation. This makes the responses to violations of expectations
informative about underlying cognitive schemes that generated the
expectations.
  Special consideration is given to musical expertise (rule-based
learning through formal music training), exposure (implicit learning
of statistical regularities), as well as innate cognitive mechanisms.
  Two kinds of expectation determined by different salience of events in
rhythmic patterns were shown to be active. The first one is based on
hierarchical structuring of event salience. In this regard, it could
be shown that meter is induced in all listeners, regardless of the
level of formal musical training. Hierarchical structuring could be
found on all levels of a musical measure. Furthermore, it was shown
that the most fundamental instance of meter induction, namely the
discrimination of the downbeat from other positions in a rhythm, was
active in newborn infants. The second type of expectation based on
event salience that was considered in this thesis was the serial
position effect. It could be shown that for non-musicians, as well as
for musicians (but less so), a mechanism was active that can be seen
as complementary to hierarchical processing. Primacy and recency
effects appeared, which led to an increased salience of events located
at the beginning and at the end of a rhythmic pattern.
  On a much smaller time-scale, support was found for the hypothesis
that listeners are sensitive to deviations on a temporal micro-level,
being able to distinguish tempo-transformed from non-transformed
performances, by only focusing on expressive timing. This is
supporting previous evidence that timing does not scale proportionally
with tempo, with the new finding that also non-musicians are sensitive
to distortions. A more surprising finding was that not only the level
of formal music training was responsible for this sensitivity, but
that exposure to a certain musical genre was giving the listener an
advantage in detecting the deviations.