The retinal and cortical components of the duration of motion aftereffects


Meredith M. Peiken
University of Georgia



Abstract

The present study examined the retinal and cortical components of motion aftereffects (MAE). It was posited that MAE resulted from both retinal and cortical neuronal contributions and that adaptation time of the moving stimulus would affect MAE duration. Twenty-four undergraduate psychology students participated in this 2x3 within subjects study. Monocular and Interocular MAE conditions were adapted at 1, 3 and 6 minute intervals. MAE duration was measured after the adaptation time was completed. The results support the view that MAE results from both the monocular and interocular viewing, and MAE duration is affected by adaptation time. The findings suggest that repeated desensitization of retina and cortical neurons result in the motion aftereffect illusion, and with increased duration of adaptation, the duration of the motion aftereffect increase.

The retinal and cortical components in the duration of motion aftereffects

Visual illusions have wowed the scientific community for years. These spectacular displays of visual excitation, which occur after fixating on various stimuli, have been examined time after time in search for an explanation. One of the most common visual illusions and commonly studied is that of the motion aftereffect (MAE). The MAE occurs after adaptation to a moving object in a continuous direction is stopped, and the viewer sees a stationary stimulus, or background, as moving in the opposite direction (Keck & Pentz, 1977). A popular MAE is caused after watching a waterfall where the background foliage and land appear to move upwards.

Although this is a common MAE, an exact explanation of why this illusion occurs is inconclusive. Many researchers have concluded that the MAE is due solely by retinal neurons while others state that the motion aftereffect is explained farther back in the visual system (Antis & Gregory, 1964; Barlow & Brindley, 1963). Conversely, some researchers believe that it is not the retinal neurons or ganglion cells only causing the MAE, but rather a combination of monocular conditions and the binocularity of the cortical cells of the lateral geniculate nucleus (LGN) that cause the perception of an aftereffect (Mitchell, Reardon, & Muir, 1975). Other theorists have hypothesized that the real cause of MAE are due to orientation specific neurons becoming desensitized by repeated stimulation.

The hypothesis of MAE being monocular in nature dates far back into the mid-1900s. The historical paper by Antis and Gregory (1964) concluded that the motion aftereffects only occur from image-retina movement. Image-retina movement is when a stimulus systematically moves across the retina causing the perception of movement. MAE are restricted to the area of the visual field where the original movement occurred (Weisstein, Maguire, & Berbaum, 1977). Therefore, motion aftereffects are due specifically to retinal, or monocular, components. Concurrent with these views is the finding from Wade and Swanston (1987). They concluded that continuous adaptation of relative-motion signals, which is a fundamental aspect of motion perception, resulted in the MAE. These relative-motion signals occur early in the motion processing prior to any binocular information being integrated which implies that the MAE is monocular.

However, after replicating their results, Wade, Swanston, and de Weert (1993) suggested that the motion aftereffects might be occurring after the binocular integration. In their 1993 study, Wade, Swanston, and de Weert had participants adapt one eye to a moving stimulus and tested for a motion aftereffect with the opposite eye. They found that interocular transfer, or the transfer of information processed by one eye to the other eye, plays a major role in MAE. These aftereffects are dependent upon the stimulation of orientation-specific neurons (Movshon, Chambers, and Blakemore, 1972). This stimulation of such cells occurs at some cortical areas, supposedly the LGN or in the primary visual cortex. The stimulation can transfer from one eye to another, supporting the interocular transfer theory, but this transfer relies on orientation-specific neurons with binocular input (Movshon, Chambers, and Blakemore, 1972). The interocular transfer supports a central locus for this phenomenon (Mitchell, Reardon, and Muir, 1975). On the other hand, this research on interocular transfer is not consistent with past findings. Day and Dickinson (1977) did not find interocular transfer when they adapted a moving surround stimulus monocularly and showed a central stationary stimulus to the opposite eye; there was not a MAE. When Day and Dickinson showed both stimuli to one eye a clear MAE was seen.

From the polar theories, a combination hypothesis was made. It seems that interocular transfer of motion aftereffects is partial with the other information coming from the adapted eye. Motion signals from the adapted eye combine with the signals from the other eye in the LGN which then influences the perception of the MAE (Mitchell, Reardon, and Muir, 1975). Thus suggesting both retinal and cortical neurons are components of the motion aftereffects. Mitchell, Reardon, and Muir (1975) continue to further explain their position. They compare the motion aftereffects to an after-image, which can be induced in one eye and seen by the other. After-images can originate in the cortical areas of the visual pathway from monocular neurons or neurons that receive input from only one eye. Therefore, MAE can result from adapting both monocular and binocular cortical neurons.

Another theory of MAE exists independently from the monocular versus binocular debate. Keck and Pentz (1977) suggest that the motion aftereffect is due to desensitization of neurons. Constant movement of a stimulus desensitizes neurons with a preferred motion direction similar to the stimulus. This then leaves these neurons sensitive to the opposite direction. Barlow and Brindley (1963) stated a similar theory. The main difference between these two theories is that Barlow and Brindley believed these direction-specific neurons are the retinal ganglion cells which after repeated firing undergo a refractory period. It is during this period that the opposite direction-specific ganglion cells are sensitive to the perception of movement. If the desensitization of these neurons, whether retinal ganglion cells or cortical neurons, is determined from the amount of excitation that occurs during the adaptation period then time could be a factor. The longer these direction-specific neurons are excited by the moving stimulus, the more they are desensitized to that particular direction and more sensitive to the opposite direction.

Because of this inconsistent data, the present study will examine whether motion aftereffects have both monocular and interocular components. However, the study will consider Barlow and Brindley’s (1963) findings and those of Keck and Pentz (1977) and examine how adaptation time affects the duration of the MAE. The present study posits that the motion aftereffects will have both monocular and binocular components, but the monocular MAE will be stronger. With respect to time, the hypothesis is that the longer the adaptation to the moving stimulus, the longer the duration of the MAE. It is also posited that the longest duration of the MAE will occur from the longest adapted monocular viewing.

Method

Participants

Twenty-four Undergraduate Psychology majors from the University of Georgia participated in this study. Participants were selected on the basis of enrollment in PSYCH 4120-Sensation and Perception and was part of fulfillment of the course criteria.

Materials

The computers used were Dell Optiplex GX110 with Dell monitors in the undergraduate computer lab. Duration of adaptation and motion aftereffect was timed using either a stop-watch or a watch with a second hand. Participants sat 15 inches away from the computer screen. A 3x5-index card was used to cover the appropriate eye during adaptation and testing. The stimulus was taken from The Exploring Perception CD-ROM, unit 2.7, module 6, titled “The Waterfall Movement Aftereffect,” (Ryan, 1997). All data was collected using a score sheet included at the end of the lab description.

Procedure

First, all the participants separated themselves into groups of two. The order of adaptation is counterbalanced across partners, therefore, approximately half of the total participants (N=24) were in each condition. The counterbalancing controlled for any cumulative effects of adaptation and allowed for adaptation duration to be examined as a separate variable. Table 1 in Appendix A describes the conditions set for each participant to follow.

Using the ruler, the group members found a distance of 15 inches from the computer screen in which to conduct the trials. After this distance was established, and after the correct module was located, each group member familiarized himself or herself with the motion aftereffect phenomenon by participating in one practice trial. The interframe interval, or speed of the stimulus, was manually set at 90 by moving the tab at the bottom to the right of the screen until the counter registers 90. Each group member was to fixate on the red cross in the center of the stimulus and starts the motion of the stimulus. After 60 seconds, the motion stopped which should have produced an adequate motion aftereffect.

Once the practice trial was completed, the groups were ready to begin gathering data. The first participant began with the correct condition set as explained in Table 1. The participant covered the appropriate eye with the index card and fixated on the red cross. Once the participant was ready, the experimenter began the movement of the stimulus and started timing. When 10 seconds remained in the adaptation period, the experimenter notified the participant to prepare for the test phase. When the adaptation time was over, the experimenter stopped the motion and began timing the aftereffect period until the participant told the experimenter that s/he can no longer detect the motion aftereffect. The experimenter recorded the duration of the motion aftereffect in seconds in the correct location on the score sheet. The participant and experimenter exchanged places and prepared for the next trial.

Results

A 3x2 repeated measures ANOVA was used to analyze the results, with the significant level set at 0.05. The main findings support both the hypotheses that MAE resulted from both retinal and cortical components as explained from the monocular eye and interocular transfer, and that adaptation time affects the duration of the MAE. Significant main effects, at the a=0.05 level, was found for both the eye condition F(1,23)= 13.82, p= 0.001, and adaptation time F(2,46)= 28.34, p=0.00. The main effect for eye condition shows that differences do exists in the duration of MAE between the monocular and interocular viewing conditions. The main effects plot for monocular and interocular conditions can be seen as Figures 1 and 2, respectively in Appendix B. The main effect for adaptation time shows that differences exist between the three adaptation times and MAE duration. A pairwise comparison was conducted to examine where the differences lie. All comparisons were significant showing that the differences exist between all adaptation times. A significant interaction between adaptation time and eye was also found, F(2,46)= 3.297, p=0.046. This interaction exhibits that the main difference between the two groups occur at the 6 minute adaptation time. The monocular viewing condition with a 6 minute adaptation period (M=26.11, SD=18.43) had a significantly larger MAE duration than the interocular condition with the same adaptation time (M=14.19, SD= 14.28). The interaction plot can be found in Appendix B as Figure 3. The ANOVA summary table can be found as Table 2 in Appendix A. Descriptive statistics table can be found as Table 3 in Appendix A.

Discussion

The results from this study clearly support the hypotheses that motion aftereffects are both retinal and cortically mediated and that adaptation time affects the duration of the MAE by means of a longer adaptation period eliciting a longer MAE duration. The main effect of the eye condition, bolsters the hypothesis that both retinal and cortical components are involved in MAE. The main effect for adaptation time, 1 minute, 3 minutes, or 6 minutes, bests supports the hypothesis that the longer adaptation time results in a longer MAE duration. The significant interaction between the eye condition and time also supports the hypothesis that the longer the eye is adapted to the moving stimulus, the longer the MAE duration is by comparing mean values.

The findings from the present study support Mitchell, Reardon, and Muir’s 1975 study where they concluded that the motion aftereffects are due to both retinal and cortical components. Like Wade, Swanston, and de Weert (1993), this study concluded that interocular transfer does play a role in motion aftereffects. In order for the MAE to be viewed after adapting one eye and testing the other, binocular information must be getting into the system somewhere in the advanced stages of visual processing.

This fact, however, does not discount the importance of the retinal information. Monocular adaptation and testing produced a stronger MAE suggesting retinal influences are more prominent in the perception of MAE. The relative-motion signals which are being adapted to the moving stimulus early in visual processing is highly significant in the perception of the motion aftereffects (Wade and Swanston, 1987). Therefore, information prior to binocular integration is the initial step to MAE perception, with the interocular transfer of information preceding.

The theory that direction-specific neurons are becoming desensitized to repeated excitation leaving the opposite direction-specific neurons more sensitive can also be supported by these results. These neurons are located in the LGN and in the primary visual cortex (V1). These neurons are also eye specific meaning each eye has identical direction-specific neurons, and these neurons follow the topographic map of the retina in the cortex. If the moving stimulus is adapted and tested monocularly, the direction-specific neurons for that eye will consequently be desensitized and the MAE is perceived. If the motion is adapted and viewed interocularly, those neuron for the eye adapted are still desensitized leaving the other neurons of both eyes senstitive to the MAE.

This also suggests why the monocular MAE was stronger than the interocular MAE. When monocular, the amount of the direction-specific neurons in the eye that are being repeatedly excited and the opposite direction-specific neurons is almost equal, thus mediating a strong MAE. When tested interocularly for the MAE, both opposite-direction neurons are remaining eliciting a motion aftereffect, but it is not as strong because the ratio of desensitized neurons to sensitive neurons is greater than in the monocular condition. There are more sensitive neurons than desensitized neurons creating a weaker MAE.

Many different areas in the brain process motion. Murakami and Shimojo (1995) have suggested that relative-motion is processed by the medial temporal (MT) area. The neurons in the MT are known for their high level of binocularity suggesting binocular information is needed for the processing of motion aftereffects. It has also been suggested that V1 and other subcortical area process MAE. It is most likely that all these areas and others mediate the motion aftereffect.

Limitations

Limitations for this study include low control. Because the participants tested each other, the procedures may not have been followed exactly causing skewed data. Also, the procedures may not have been understood completely by the participants where they did not know what was expected of them. Further and more complete explanation of the procedures and expectations could mollify this limitation.

Implications for Future Research

Future projects on this topic should examine whether ocular dominance has any implications on these motion aftereffects. A person with a clear ocular dominance could in fact have a stronger interocular transfer. Mitchell, Reardon, and Muir (1975) have found that persons with a clear ocular dominance showed a greater transfer from the dominant eye to the non-dominant eye, and vice versa. Varying adaptation times and controlling for ocular dominance could reveal interesting factors of the MAE and provide a clearer picture of where the retinal and cortical components come into play.

References

Antis, S.M., & Gregory, R.L. (1964). The aftereffect of seen motion: The role of retinal stimulation and eye movements. Quarterly Journal of Experimental Psychology, 17, 173-174.
Barlow, H.B., & Brindley, G.S. (1963). Inter-ocular transfer of movement aftereffects during pressure blinding of the eye. Nature, 200, 1347.
Day, R.H., & Dickinson, R.G. (1977). Absence of color-sensitivity in Duncker-type induced visual movement. In Symons, L.A., Pearson, P.M, & Timney, B. (1996). The aftereffect to relative motion does not show interocular transfer. Perception, 25, 651-660.
Keck, M.J., & Pentz, B. (1977). Recovery from adaptation to moving gratings. Perception, 6, 719-725.
Mitchell, D.E., Reardon, J., & Muir, D.W. (1975). Interocular transfer of the motion aftereffect in normal and stereoblind observers. Experimental Brain Research, 22, 163-173.
Movshon, J.A., Chambers, B.E.I., & Blakemore, C. (1972). Interocular transfer in normal humans and those who lack stereopsis. Perception, 1, 483-490.
Murakami, L., & Shimojo, S. (1995). Modulation of motion aftereffects by surround motion and its dependence on stimulus size and eccentricity. In Symons, L.A., Pearson, P.M, & Timney, B. (1996). The aftereffect to relative motion does not show interocular transfer. Perception, 25, 651-660.
Ryan, C. (1997). Exploring Perception: A CD-ROM for Macintosh and Microsoft Windows. Pacific Grove, CA: Brooks/Cole Publishing.
Symons, L.A., Pearson, P.M, & Timney, B. (1996). The aftereffect to relative motion does not show interocular transfer. Perception, 25, 651-660.
Wade, N.J., & Swanston, M.T. (1987). The representation of nonuniform motion: induced movement. In Symons, L.A., Pearson, P.M, & Timney, B. (1996). The aftereffect to relative motion does not show interocular transfer. Perception, 25, 651-660.
Wade, N.J., Swanston, M.T., & de Weert, C.M.M. (1993). On interocular transfer of motion aftereffects. In Symons, L.A., Pearson, P.M, & Timney, B. (1996). The aftereffect to relative motion does not show interocular transfer. Perception, 25, 651-660.
Weisstein, N., Maguire, W., & Berbaum, K. (1977). A phantom-motion aftereffect. Science, 198, 955-957.