245
sentation. In the first instance T9 (iambus) elicited negligible search and ap-
proach responses from brain-damaged or intact birds. In the second instance
Iesioned groups showed less search after a filter-trill postoperatively than pre-
operatively. Nonetheless this level after operation was significantly higher
than the sham-controls with the same experience.
In summary the anterior group appear to ‘switch on’ then ‘switch off’
approach and search responses quicker than the posterior and control groups.
As such a measure is suggestive of a threshold change, I suggest the hypothesis
that the learned stimulus specifications are activated more quickly following
a stimulus and in the absence of a match (or consummation) deactivated
more quickly (early mismatch, early decision to retreat).
The posterior group did not show these characteristics. They recognised
the test for the training stimulus and approached. In the absence of consum-
mation approach and search continued longer. They made a match for the
physical characters of a test stimulus (train-trill similar to filter-trill) and for
message content (T2>6 9). The iambus did not elicit a strong approach res-
ponse. The impairment resembles a specialised type of discrimination deficit.
The hypothesis is that the rules for selecting input were damaged.
GENERAL DISCUSSION ’
There are three classes of measurement that may be indicative of changing
central processes in response to auditory stimuli. Variations of these measures
for two lesion (anterior and posterior hyperstriatum∕hippocampus) and two
control groups (preoperative and sham-operates) suggest that the central con-
trol of these processes has been differently affected. The first class of measure-
ment consists of approach and departure latencies. The second is the frequen-
cy of behavioural change (of body tension, head and body movement). The
third is the change of dominance of these tension and movement responses.
The direction of change from one phase to the other (S-HT, S-LT) can be im-
portant. The tension components and locomotion are related. The following
model is an interpretation based on observations of these measures.
A model
Previous work using visual task with chicks has shown that damage similar
to the anterior lesions of this study induced a persistence in use of a strategy
for selecting one class of input for response (Oades, 1976a, b). The current
study confirms this and that more posterior damage can induce persistence
associated with auditory and visual stimuli. This persistence, in comparison
with controls, is seen as an ‘activation’ by the anterior group toward test sti-
muli that are similar to the training stimuli and as one of continued explora-
tory responses by the posterior group. On the basis of the detailed measures
taken it is proposed that there are two types of ‘persistence’. Further testing
will be necessary to confirm this hypothesis.