70
DAVID INGLE
nation could be prevented if the second brain half were “occluded by
the response to an irrelevant spotted stimulus presented to the second eye
on each trial. Thus, we conclude that transfer may fail when ( a ) the
commissural system cannot resolve fine stimulus differences, or ( b ) the
second brain half is jammed by induced noise and cannot record input
from the commissures.
The fact that interocular integration is sometimes incomplete suggests
that independent visual learning might proceed simultaneously within
opposite halves of the brain. Ingle (1968a) has demonstrated that goldfish can, indeed, acquire opposing discriminations of pattern or of color
via opposite eyes; for example, avoiding red not green via the right eye
and avoiding green not red via the left. Double learning is easily achieved
when stimuli are presented either ( a ) to both eyes on each trial, or ( b )
to one eye at a time, alternating eyes on successive trials, but is more
difficult to attain when ( c ) long sequences of monocular trials are confined
to one eye at a time. Perhaps the difficulties inherent in the “alternating
sessions” method account for the inability of Schulte (1957) or Shapiro
( 1965) to demonstrate interocular double learning with carp or goldfish.
However, Schulte and Shapiro both used frontally approached (expanding) stimuli, while Ingle used size-restricted stimuli confined to the lateral
field. Perhaps interocular integration is favored in the ( overlapping)
binocular field, while dissociation is more easily obtained using lateral
stimuli.
Finally, we consider an interocular transfer problem that is rather
peculiar in its very formulation: How can stimuli appearing via opposite
eyes be judged equivalent when the discrimination is based upon differences in the directional orientation of the shapes? Since the fish‘s eyes
are set upon opposite sides of the head, there is considerable ambiguity
in predicting how one hemisphere communicates distinctions of leftright or back-front to the other. As Fig. 5 illustrates, a leftward (nasalward) arrowhead viewed by the right eye casts an image projected to
the tectum in a rostral-pointing direction just as a frontward pointing arrowhead seen in the right lateral field. If the arrowhead in the frontal
plane-seen by both eyes at once-is to produce two images that map
onto one another via the commissural system, one must conclude that
a nasalward direction for one eye is “equivalent” to a temporalward
direction as seen via the second eye. But if one coding process is applicable to all parts of the two retinas, this logic forces the absurd prediction
that an object seen in front of the fish via one eye is more similar to a
trailing than to a leading stimulus on the other side.
Ingle (1967, 1968b) has shown that both mechanisms coexist and are
called forth by different kinds of experimental stimuli. Fish trained on
DAVID INGLE
nation could be prevented if the second brain half were “occluded by
the response to an irrelevant spotted stimulus presented to the second eye
on each trial. Thus, we conclude that transfer may fail when ( a ) the
commissural system cannot resolve fine stimulus differences, or ( b ) the
second brain half is jammed by induced noise and cannot record input
from the commissures.
The fact that interocular integration is sometimes incomplete suggests
that independent visual learning might proceed simultaneously within
opposite halves of the brain. Ingle (1968a) has demonstrated that goldfish can, indeed, acquire opposing discriminations of pattern or of color
via opposite eyes; for example, avoiding red not green via the right eye
and avoiding green not red via the left. Double learning is easily achieved
when stimuli are presented either ( a ) to both eyes on each trial, or ( b )
to one eye at a time, alternating eyes on successive trials, but is more
difficult to attain when ( c ) long sequences of monocular trials are confined
to one eye at a time. Perhaps the difficulties inherent in the “alternating
sessions” method account for the inability of Schulte (1957) or Shapiro
( 1965) to demonstrate interocular double learning with carp or goldfish.
However, Schulte and Shapiro both used frontally approached (expanding) stimuli, while Ingle used size-restricted stimuli confined to the lateral
field. Perhaps interocular integration is favored in the ( overlapping)
binocular field, while dissociation is more easily obtained using lateral
stimuli.
Finally, we consider an interocular transfer problem that is rather
peculiar in its very formulation: How can stimuli appearing via opposite
eyes be judged equivalent when the discrimination is based upon differences in the directional orientation of the shapes? Since the fish‘s eyes
are set upon opposite sides of the head, there is considerable ambiguity
in predicting how one hemisphere communicates distinctions of leftright or back-front to the other. As Fig. 5 illustrates, a leftward (nasalward) arrowhead viewed by the right eye casts an image projected to
the tectum in a rostral-pointing direction just as a frontward pointing arrowhead seen in the right lateral field. If the arrowhead in the frontal
plane-seen by both eyes at once-is to produce two images that map
onto one another via the commissural system, one must conclude that
a nasalward direction for one eye is “equivalent” to a temporalward
direction as seen via the second eye. But if one coding process is applicable to all parts of the two retinas, this logic forces the absurd prediction
that an object seen in front of the fish via one eye is more similar to a
trailing than to a leading stimulus on the other side.
Ingle (1967, 1968b) has shown that both mechanisms coexist and are
called forth by different kinds of experimental stimuli. Fish trained on