3. THE EXPERIMENTAL ANALYSIS OF VISUAL BEHAVIOR
75
the timing of the snap, Mouthbreeders do require continuous motion
feedback while pursuing food since, under a stroboscopic light at 3 Hz,
they badly misjudge the rate of fall and snap too high. The multispotted
stimulus might also be an analog of temporally moving features within
the visual array through which the fish swims. Here, as well, visual feedback informs the fish of his own progress (in a moving stream, the visual
cue to swimming rate would be the most reliable one). Rate of temporalward motion also provides information on distances of objects or surfaces
ahead or to the side. In this context, the dominance of horizontal or
vertical axes, revealed by shape recognition experiments, makes good
sense: These particular axes of image translation are produced by fish
that themselves locomote within horizontal or sagittal planes.
The interesting ability of goldfish to notice “parallelness” might also
serve to discriminate the tilt of surfaces (such as inclining rocks). If a
retinal image displacement resulting from body movement maintained
the orientation of a contour in the retinal image (i.e., displaced, but still
parallel), the fish might infer that the contour is confined to a plane
perpendicular to the horizontal plane through which it swims. Nonparallel
displacements, on the contrary, represent contours or surfaces that
incline toward, or slope away from, the organism. The broad category
“curvature” can be seen in this light: A curving retinal displacement must
correspond to an object moving independently of the fish. Perhaps as
various shapes may have physionomic connotations to man (Werner,
1940), curved lines suggest something “animate” to the fish.
These teleological speculations cannot pass for explanations of visual
behavior, but they can direct the psychologist toward new methods of
measuring specific visual abilities. As yet, we know almost nothing about
the sensitivity of fish to the various transformations of the optic array
through which they swim. The method of “false feedback used to analyze
human sensorimotor abilities could be used with many fish (as Sperry,
1950, achieved by eye rotation). Already, some physiological evidence
indicates that concern for differential patterns of motion seen by fish is
likely to be fruitful in research. Jacobson (1968) has described directionally sensitive ganglion cells in the goldfish retina that can be suppressed
by moving a second object outside of the receptive field in a direction opposite to that of the stimulating object. Discrimination between objects
that move together and those that converge may utilize these peripheral
intraretinal inhibitions to code differential movement. It is worth adding
that the frog-who cannot obtain such parallax information-does not
show the kind of direction-specific cross inhibition that Jacobson has
revealed in the fish. Perhaps in gaining physiological hints as to the
mechanism of spatial vision in higher mammals, the active fish will prove
a better model than statuesque amphibians and reptiles,
75
the timing of the snap, Mouthbreeders do require continuous motion
feedback while pursuing food since, under a stroboscopic light at 3 Hz,
they badly misjudge the rate of fall and snap too high. The multispotted
stimulus might also be an analog of temporally moving features within
the visual array through which the fish swims. Here, as well, visual feedback informs the fish of his own progress (in a moving stream, the visual
cue to swimming rate would be the most reliable one). Rate of temporalward motion also provides information on distances of objects or surfaces
ahead or to the side. In this context, the dominance of horizontal or
vertical axes, revealed by shape recognition experiments, makes good
sense: These particular axes of image translation are produced by fish
that themselves locomote within horizontal or sagittal planes.
The interesting ability of goldfish to notice “parallelness” might also
serve to discriminate the tilt of surfaces (such as inclining rocks). If a
retinal image displacement resulting from body movement maintained
the orientation of a contour in the retinal image (i.e., displaced, but still
parallel), the fish might infer that the contour is confined to a plane
perpendicular to the horizontal plane through which it swims. Nonparallel
displacements, on the contrary, represent contours or surfaces that
incline toward, or slope away from, the organism. The broad category
“curvature” can be seen in this light: A curving retinal displacement must
correspond to an object moving independently of the fish. Perhaps as
various shapes may have physionomic connotations to man (Werner,
1940), curved lines suggest something “animate” to the fish.
These teleological speculations cannot pass for explanations of visual
behavior, but they can direct the psychologist toward new methods of
measuring specific visual abilities. As yet, we know almost nothing about
the sensitivity of fish to the various transformations of the optic array
through which they swim. The method of “false feedback used to analyze
human sensorimotor abilities could be used with many fish (as Sperry,
1950, achieved by eye rotation). Already, some physiological evidence
indicates that concern for differential patterns of motion seen by fish is
likely to be fruitful in research. Jacobson (1968) has described directionally sensitive ganglion cells in the goldfish retina that can be suppressed
by moving a second object outside of the receptive field in a direction opposite to that of the stimulating object. Discrimination between objects
that move together and those that converge may utilize these peripheral
intraretinal inhibitions to code differential movement. It is worth adding
that the frog-who cannot obtain such parallax information-does not
show the kind of direction-specific cross inhibition that Jacobson has
revealed in the fish. Perhaps in gaining physiological hints as to the
mechanism of spatial vision in higher mammals, the active fish will prove
a better model than statuesque amphibians and reptiles,