- Email: [email protected]
Dissociable effects of stimulus range on perception and action
CORTEX1906_proof ■ 5 January 2017 ■ 1/6 c o r t e x x x x ( 2 0 1 7 ) 1 e6
Available online at www.sciencedirect.com 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64
ScienceDirect Journal homepage: www.elsevier.com/locate/cortex
Special issue: Research report
Q6
Dissociable effects of stimulus range on perception and action
Q5 Q1
Gal Namdar a, Daniel Algom b and Tzvi Ganel a,* a b
Department of Psychology, Ben-Gurion University of the Negev, Beer-Sheva, Israel School of Psychological Sciences, Tel-Aviv University, Tel-Aviv, Israel
article info
abstract
Article history:
We have recently reported the discovery that the ability to detect a minimum increment to
Received 20 October 2016
a stimulus depends on the spread of the other stimuli for which this just noticeable dif-
Reviewed 9 December 2016
ference (JND) is being measured (Namdar, Ganel, & Algom, 2016). In particular, the JND
Revised 18 December 2016
around a standard stimulus was larger when the other standards tested within the same
Accepted 20 December 2016
experimental session spread a larger range. In this study we show that this range of
Published online xxx
standards effect (RSE) is limited to perceptual estimations and does not extend to action. The JND remained invariant when the participants grasped the objects rather than
Keywords:
perceptually estimated their size. This difference supports the hypothesis that visual
Perception and action
perception, on the one hand, and visually controlled action, on the other hand, are gov-
Grasping
erned by separate rules and mediated by different mechanisms.
Effects of context
© 2016 Elsevier Ltd. All rights reserved.
Relative processing Psychophysics
1.
Introduction
Perceptual resolution with respect to a given stimulus is contingent upon the context in which the stimulus is pre€ m, 2000). In sented (Namdar et al., 2016; see also, Hellstro particular, perceptual sensitivity with respect to the same stimulus is greater [i.e., the just noticeable difference (JND) or the difference threshold or limen, DL, is smaller] when the other stimuli tested are of similar magnitudes than when they are spread over a larger range of magnitudes. For example, the JND for a stimulus with a weight of 400 g was found smaller when the other standards tested within the same condition were 350 and 450 g than when they were 200 and 600 g. In
other words, the perceptual resolution for a single stimulus is contingent upon the resolution of the context, or stimulus set, in which it is embedded. Coarse resolution of the set (e.g., wider spacing between stimuli) leads to a decrease in the resolution of each stimulus within that set. Alternatively, it was also suggested that the range of standards effect (RSE) could have stemmed from adaptation effects to stimulus range, which result in an increased sensitivity to stimuli in a narrow compared to a wide range (Namdar et al., 2016). The RSE has been established across different perceptual modalities, including visual size discrimination and tactile weight perception. However, it is still unclear whether the RSE applies solely to the perceptual system. Does it also constrain performance in the domain of visuomotor control of action?
* Corresponding author. Department of Psychology, Ben-Gurion University of the Negev, Beer-Sheva 84105, Israel. E-mail address: [email protected] (T. Ganel). http://dx.doi.org/10.1016/j.cortex.2016.12.017 0010-9452/© 2016 Elsevier Ltd. All rights reserved. Please cite this article in press as: Namdar, G., et al., Dissociable effects of stimulus range on perception and action, Cortex (2017), http:// dx.doi.org/10.1016/j.cortex.2016.12.017
Q2
65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129
CORTEX1906_proof ■ 5 January 2017 ■ 2/6
2 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
Q3
c o r t e x x x x ( 2 0 1 7 ) 1 e6
Are grasping trajectories immune to the range of the stimuli being grasped? This seems plausible given evidence that grasping is less susceptible to influences of context than perception (e.g., Ganel & Goodale, 2003, 2014; Goodale & Westwood, 2004). We note though that a number of recent studies reported short-term contextual modulation of grasping, too (e.g., Dixon & Glover, 2009; LeBlanc & Westwood, 2015). However, if the RSE is found limited to visual perception so that it does not affect visually-controlled action, then a powerful diagnostic to dissociate perception and action at the level of a single stimulus becomes available. According to Goodale and Milner's influential account, vision for action and vision for perception are subserved by distinct (yet interactive) neuroanatomical streams (Goodale & Milner, 1992; Mishkin, Ungerleider, & Macko, 1983; but see; Franz, Gegenfurtner, Bu¨lthoff, & Fahle, 2000). Goodale and Milner suggested that the dorsal pathway in the primate visual system, which includes parietal regions, enables flexible control of actions directed at objects in the environment. The ventral pathway, by contrast, provides a rich and detailed representation of the environment sustaining visual perception (Goodale & Milner, 1992). Research within this framework suggests that the visuomotor system processes information in a selective manner more than does the visual perceptual system (Ganel & Goodale, 2003, 2014; Goodale & Westwood, 2004). This property often insulates grasping from the unwanted influence of task-irrelevant contextual information (Aglioti, DeSouza, & Goodale, 1995; Doherty, Campbell, Tsuji, & Phillips, 2010; Ganel, Chajut, & Algom, 2008; Ganel, Tanzer, & Goodale, 2008, but see; Franz et al., 2000; Utz, Hesse, Aschenneller, & Schenk, 2015). Several recent studies have suggested that there are some situations in which grasping trajectories can be modulated by short-term contextual information. For example, it has been reported that the trajectory of grip aperture is affected, at least to some extent, by the aperture of the immediate preceding trial (Dixon & Glover, 2009, see also; LeBlanc & Westwood, 2015). Additionally, recent evidence suggests that grasping, similarly to visual perception, can be affected by contextual information, such as that entailed in the Ebbinghaus illusion (Kopiske, Bruno, Hesse, Schenk, & Franz, 2016; but see; Ganel & Goodale, 2014; Haffenden & Goodale, 1998; Whitwell & Goodale, 2016). A different line of evidence suggests though, that unlike perceptual estimations, which are affected by irrelevant variations in stimulus shape, grasping is immune to the same effects (Ganel & Goodale, 2003, 2014). It is therefore an open question whether the RSE, an effect which has been only demonstrated in the perceptual domain, would extend to affect online control of actions.
1.1.
The present study
Our goal in this study was twofold. First, we wished to replicate the RSE so that it rests on an even firmer empirical basis. In the previous study (Namdar et al., 2016), we used the psychophysical method of Constant Stimuli. In this study, we made use of the Method of Adjustment (Baird & Noma, 1978; Gescheider, 1976; Marks & Algom, 1998). In this method, the JND is given by the standard deviation of the reproductions of the stimulus tested for resolution. An advantage of the
method is that the JND or DL derived can be easily compared between perception and grasping (Ganel, Chajut, & Algom, 2008; Ganel, Chajut, Tanzer, & Algom, 2008; Ganel, Freud, & Meiran, 2014). We expected that the JND for the same 40 mm stimulus would be larger when the other stimuli tested simultaneously are 20 and 60 mm than when the other stimuli are 35 and 45 mm. Secondly and more important, we expected that this effect would be limited to perception: grasping the same stimuli would not result in modification of the JND for the common 40 mm stimulus. Therefore, in the perceptual condition, the participants were asked to make perceptual estimations of size, whereas in the grasping condition they were asked to perform visually-guided grasping movements toward the same objects. In a control experiment we also tested whether the hypothesized insensitivity of the grasping trajectories to the measuring context can possibly be confounded by on-line corrections based on visual feedback from the fingers and the target object (Glover & Dixon, 2002). Such online corrections could mask contextual effects on grip apertures. To address this concern, grasping movements in the control experiment were performed in an open-loop design, for which visual feedback was prevented during grasp.
2.
Methods
2.1.
Participants
A group of 14 right-handed students from Ben-Gurion University of the Negev, who gave their informed consent, participated in the main experiment (3 males, mean age of 23.35, SD ¼ 1.59 years). A separate group of 14 right-handed participants (8 males, mean age of 23.57, SD ¼ 2.65 years) participated in the control, open-loop grasping experiment. All participants signed a consent form and the experimental protocol was approved by the departmental ethics committee in accordance with the Code of Ethics of the World Medical Association (Declaration of Helsinki).
2.2.
Apparatus and stimuli
The stimuli were five plastic rods of a constant width and height of 5 mm. The values of length were 20, 35, 40, 45, and 60 mm. The participants set in front of a black tabletop with the tips of the index finger and thumb (right hand) resting on a small starting button fixed to the center of the tabletop. The participants wore a set of LCD glasses (Translucent Technologies, Toronto, ON) with liquid-crystal shutter lenses used to control stimulus exposure time. The experimenter manually switched between the rods prior to each trial in a pseudorandomized order. Each rod was placed 30 cm from the participant, in the center of the table, along the midline of the participant. Rods were placed on the table perpendicularly to the participant's viewing plane (Fig. 1).
2.3.
Design
The main experiment consisted of two separate, closed-loop tasks, a grasping task and a manual estimation task. In the grasping task, the participants were asked to grasp the rod
Please cite this article in press as: Namdar, G., et al., Dissociable effects of stimulus range on perception and action, Cortex (2017), http:// dx.doi.org/10.1016/j.cortex.2016.12.017
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
CORTEX1906_proof ■ 5 January 2017 ■ 3/6 c o r t e x x x x ( 2 0 1 7 ) 1 e6
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
3
Fig. 1 e An illustration of (a) a grasping movement towards one of the objects and (b) a manual perceptual estimation of the object's size.
along its length using their thumb and index finger, and then lifted the rod a few centimeters in the air before placing it back on the table and returning their hand to the starting position. In the manual estimation task, the participants gave estimates of the rod's length without grasping it. They did so by reaching to a designated location about 20 cm to the rod's right side while opening their finger and thumb to an extent that matched the perceived length of the rod (see Fig. 1). Each task included two experimental blocks, the wide and the narrow range conditions. The wide range condition consisted of stimuli of 20 mm, 40 mm and 60 mm in length, whereas the narrow range condition consisted of stimuli of 35 mm, 40 mm and 45 mm in length. Notice that both experimental blocks included the common stimulus of 40 mm, and that they differed only by the flanking stimuli, defining the range. In both range conditions (in each task), each of the three stimuli was presented 20 times, resulting in a total of 240 trials in all. Both range conditions under both tasks began with 10 training trials which were not included in the analysis of the data. The goal of training was to familiarize the participants with the range of stimuli used in the block. Order of task and range was counterbalanced across participants. Each trial began with the opening of the goggles for a total duration of 2000 msec, providing sufficient time for completing the grasping of the object or estimating its size. In a control experiment, a separate group of participants performed grasping movements under the exact same conditions as in the grasping condition of the main experiment. However, unlike as in the closed-loop experiment, participants' vision was occluded at the moment that they initiated their movements. By doing so, we eliminated the participants' ability to perform on-line corrections based on visual feedback.
2.4.
thumb throughout the entire grasping movement trajectory (at a 200-Hz sampling rate). Movement onset was determined as the point in time at which the aperture between index finger and thumb increased by more than .1 mm for at least ten successive frames (of 50 msec). Movement offset was determined as the point in time at which the aperture between index finger and thumb changed by no more than .1 mm for at least ten successive frames, but only after reaching the MGA between fingers in the grasping condition. Trials were excluded from the analysis if the infrared diodes placed on the fingers were not visible to the camera while moving toward the object or if the participant failed to grasp the object properly and dropped it while trying to lift it off the table. This resulted in 4.7% and 2.8% of the trials being excluded from the grasping and manual estimation conditions, respectively.
2.5.
Data analysis
In order to calculate the JND for the grasping condition, we derived the MGA for each grasping movement trajectory (Foster & Franz, 2013; Ganel & Goodale, 2014; Ganel, Freud, Chajut, & Algom, 2012). In the manual estimation (perception) condition, we measured the aperture between the index finger and the thumb at the end of the estimation. The JNDs in both conditions are given by the respective standard deviations of the responses to each stimulus across trials. A correction for outliers was performed on the data by eliminating any score that was either 2.5 standard deviations below or above the mean score for each condition, resulting in the omission of less than 2% of the trials.
Kinematic recordings
3. An Optotrak Certus device (Northern Digital, Waterloo, ON) was used to track the hand and finger grasping trajectories. Two infrared light-emitting diodes were placed on the right hand thumb and index finger (located on top of the center of the distal phalanges), and another diode was placed on the wrist of the right hand. We collected data about the location of each diode and the aperture between the index finger and
Results
Consider first the result for perception, i.e., for the magnitude estimates conveyed through the fingers. On average, the JND for the common rod of 40 mm was 4.47 mm when tested along with flankers of 20 mm and 60 mm, but was 3.17 when tested along with flankers of 35 mm and 45 mm. The JND was larger by roughly 34% in the large-range condition than in the small-
Please cite this article in press as: Namdar, G., et al., Dissociable effects of stimulus range on perception and action, Cortex (2017), http:// dx.doi.org/10.1016/j.cortex.2016.12.017
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
CORTEX1906_proof ■ 5 January 2017 ■ 4/6
4
JND (mm)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
c o r t e x x x x ( 2 0 1 7 ) 1 e6
4.9
Grasping
4.4
Perceptual esƟmaƟons
3.9 3.4 2.9 2.4 Narrow
Wide
Range of standards
Fig. 2 e Effects of standards range on perceptual estimations and on maximum grip apertures during grasping of a 40 mm stimulus. Error bars represent confidence intervals of 95% (Jarmasz & Hollands, 2009). range condition. Therefore, resolution improved considerably under the small-range condition. Consider next the JND for the same rod of 40 mm determined through actual grasping. The JND for the common rod of 40 mm was 3.08 mm when tested along with flankers of 20 mm and 60 mm, and was 2.93 mm when tested with flankers of 35 mm and 45 mm. Clearly, the JND was insensitive to the spread of the flankers with grasping e in sharp contrast with the results for perception. The difference between the two visually-governed functions is supported by the interaction of task (estimation, grasping) and range (narrow, wide) in a repeated-measure ANOVA [F(1,13) ¼ 7.08, p < .05, h2p ¼ .35] (see Fig. 2). Planned comparisons revealed that range affected manual estimations [F(1,13) ¼ 9.07, p < .01], but not grasping [F(1.13) < 1, nsec]. The contingent nature of the RSE thus provides a rigorous dissociation between perception and action (see Table 1 for the data of individual participants). The range of stimuli had no effect on the average manual estimations for the 40 mm stimulus under the narrow-range condition (54.47 mm) compared with the wide-range condition (54.29 mm) [t(13) < 1, nsec]. Too, no effects of range were
found for the grasping task (65.90 and 66.51 mm for the average MGA under the narrow and wide conditions, respectively. [t(13) ¼ 1.08, nsec]). Average MGAs for the 20, 35, 45 and 60 mm objects were 51.6, 62.8, 67.9 and 78.9 mm, respectively. Average manual estimations for the 20, 35, 45 and 60 mm objects were 37.6, 48.5, 59.4 and 72.4 mm, respectively. In the control experiment, the average standard deviation during MGA was 5.11 mm in the narrow range condition and 4.73 mm in the wide range condition [t(13) ¼ .48, nsec]. To further test the null-effect of the measurement context on grip apertures under the open-loop condition, we performed a Bayesian t-test analysis to estimate the probability of the null hypothesis to be true given the aperture data (Wagenmakers, 2007). The Bayes factor was calculated using the JASP software package (Team & others, 2016). The analysis revealed a Bayes factor of 5.091 (BF0) indicating that the null hypothesis is roughly 5 times more likely compared to the alternative hypothesis. Finally, the JNDs for the grasping conditions from the main experiment and the control experiment were directly compared by a mixed-effect ANOVA analysis with range (Narrow/Wide) as a within-subject factor, and experiment (Main/Control) as a between-subjects factor. The analysis revealed a main effect of experiment [F(1,26) ¼ 23, p < .0001, h2p ¼ .46], and no main effect of range [F(1,26) < 1, nsec]. More importantly, the interaction between experiment and range [F(1,26) < 1, nsec] was nonsignificant. In order to test for possible effects for context in parts of the grasping trajectory other than the MGA (see Dixon & Glover, 2009), grasping and manual estimations trajectories for the 40 mm stimulus in the main experiment were normalized and divided to 11 equal-sized intervals. In each interval, the effect of context range was calculated as the difference in the standard deviations between the two range conditions. Post-hoc comparisons revealed no effect for range in any of the normalized time points for grasping trajectories. This provides further evidence that grasping is unaffected by the range of the contextual information throughout the entire movement. In contrast, analysis of the manual estimations trajectories revealed a
Table 1 e RSE and JND, calculated as standard deviations for action (grasping) and perception (manual estimations) of the 40 mm stimulus under both range conditions of the main experiment for each individual participant. DRSE denotes the difference in RSE between the action and the perception conditions. Participant
JND (SD)
RSE (WideeNarrow)
Action
1 2 3 4 5 6 7 8 9 10 11 12 13 14 Average
Perception
Narrow
Wide
Narrow
Wide
Action
Perception
DRSE
3.50 3.42 1.85 2.59 3.97 3.38 2.25 3.12 2.15 2.27 3.17 1.85 2.06 5.42 2.93
2.89 3.79 3.80 2.59 2.24 3.42 2.47 2.71 4.14 1.25 3.34 2.48 2.66 5.27 3.08
2.64 4.21 1.29 2.91 2.88 2.88 3.28 2.57 2.77 6.04 2.68 2.91 3.38 3.99 3.17
5.31 5.39 6.37 5.11 3.73 4.43 5.05 3.36 1.95 5.00 4.54 5.10 4.05 3.18 4.47
.60 .36 1.96 .01 1.72 .04 .22 .42 1.99 1.02 .16 .64 .60 .16 .15
2.66 1.18 5.08 2.20 .85 1.55 1.77 .79 .83 1.03 1.86 2.19 .67 .81 1.29
3.27 .82 3.12 2.20 2.57 1.51 1.55 1.21 2.82 .01 1.69 1.56 .07 .66 1.15
Please cite this article in press as: Namdar, G., et al., Dissociable effects of stimulus range on perception and action, Cortex (2017), http:// dx.doi.org/10.1016/j.cortex.2016.12.017
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
CORTEX1906_proof ■ 5 January 2017 ■ 5/6 c o r t e x x x x ( 2 0 1 7 ) 1 e6
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
strong effect for context, beginning at 50% of the movement and throughout the entire movement [F(1,13) ¼ 11.68, p < .01]. The data for the manual estimations condition largely (though not fully) obeyed Weber's Law. In the wide range condition, the JNDs for stimuli of 20, 40, and 60 mm were 3.04, 4.47, and 5.45 mm, respectively. In the narrow range condition, the JNDs for stimuli of 35, 40 and 45 mm were 4.09, 3.17, and 4.17 mm, respectively. Note that in the narrow range, the JND pattern was not fully consistent with Weber's law, probably due to low statistical power due to the relatively small differences in length between the stimuli. In contrast to the general adherence to Weber's law during manual estimations, JNDs during grasping did not obey Weber's Law. The JNDs in the wide range condition did not increase with object size and were 3.28, 3.08 and 2.83 mm for the stimuli of 20, 40 and 60 mm, respectively. In the narrow range condition, the JNDs were 3.05, 2.93 and 3.08 mm for the stimuli of 35, 40 and 45 mm, respectively. The diverging pattern of adherence to Weber's law is consistent with the results of previous studies (Ganel & Goodale, 2014; Ganel, Chajut & Algom, 2008; Smeets & Brenner, 2008). We note that a recent study (Utz et al., 2015) showed that JNDs under certain experimental conditions can be affected by ceiling effects that stem from the participants' limited range of motion (but see Manzone, Davarpanah Jazi, Whitwell, & Heath, 2016 for a different conclusion). In the Utz et al.'s (2015) study, MGAs were relatively large due to the design used (mirror apparatus in which no feedback of the hand or the object was provided). This setup is conductive to ceiling effects (Manzone et al., 2016). Average MGAs and apertures were much smaller in the present study, which minimizes the possibility to reach ceiling effects.
4.
Discussion
We provided evidence for the influence of contextual information on perceptual acuity, but for the absence of such influence on visuomotor control. Our results show that stimulus resolution is affected by the range of the other stimuli tested for resolution. A wider range is conductive to a larger JND. However, this effect, startling as it is, is limited to perception. In contrast, resolution was insensitive to context when the same stimuli were grasped rather than estimated. These results join previous indications that action is affected by taskirrelevant contextual information to a lesser degree than is perception (Ganel & Goodale, 2003; Goodale, 2011). Moreover, the current results show that grasping movements are also immune to the specific effects of contextual information (Namdar et al., 2016), as range differences between blocks had no effect on grasping trajectory deviations. Context had no effect on grasping even when the participants grasped the object with their vision occluded, suggesting that the lack of contextual effect on grasping is not the result of on-line corrections.
4.1.
Dissociation between perception and action
The results of the present study suggest that grasping is immune to the effect of context. Previous indications for shortterm effects of context on grasping trajectories (e.g., Dixon & Glover, 2009; LeBlanc & Westwood, 2015) are not necessarily
5
in odds with the result of the present study. First, those studies focused on different measures. In particular, previous studies looked at response bias in grip apertures, whereas the focus of the current study was on the resolution (JND) of the response. Furthermore, earlier studies looked at short-term effects of the previous trial, whereas here we examined the effect of context on visual resolution for perception and for action across a range of stimuli. Therefore, visual resolution during grasping escapes the effects of global, adaptation to context. In a certain sense, the RSE tallies with effects found in Garner's speeded-classification task (Garner, 1978). In the Garner task, participants classify objects on a task-relevant dimension while another, task-irrelevant dimension, is either held constant (baseline condition) or varies from-trial-to-trial (filtering condition). Worse performance in the filtering condition (called Garner interference) indicates that the task-irrelevant information could not be ignored. Notably, Garner interference was found for visual perception but was not found for grasping (Ganel & Goodale, 2003, 2014). It is possible that Garner interference and RSE are rooted in the same source with regard to the effects of contextual information on performance. In the case of Garner interference, variability along the values of the irrelevant dimension leads to a decrease in perceptual performance. The RSE documents the effect on perception of larger variability of the dimension tested for resolution.
5.
Conclusion
The present results highlight the different nature of the computations underlying perceptual estimations and visuomotor control. The former is sensitive to contextual information about the stimulus of interest, a contingency that is bound to lead to the directional effect on visual resolution documented in this study. In contrast, the visuomotor control system is less susceptible to contextual information that does not strictly pertain to the task at hand. The hallmark of the action system is its selectivity to the relevant feature of to-begrasped object. The present results provide a compelling new line of evidence for the dissociation between visually guided actions and visual perception.
Uncited references Garner, 2014; Namdar et al., 2016.
Acknowledgements This study was supported by an Israel Science Foundation (ISF) grant 274/15 to Tzvi Ganel and to Daniel Algom.
references
Aglioti, S., DeSouza, J. F., & Goodale, M. A. (1995). Size-contrast illusions deceive the eye but not the hand. Current Biology, 5(6), 679e685.
Please cite this article in press as: Namdar, G., et al., Dissociable effects of stimulus range on perception and action, Cortex (2017), http:// dx.doi.org/10.1016/j.cortex.2016.12.017
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
CORTEX1906_proof ■ 5 January 2017 ■ 6/6
6 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47
c o r t e x x x x ( 2 0 1 7 ) 1 e6
Baird, J. C., & Noma, E. J. (1978). Fundamentals of scaling and psychophysics. John Wiley & Sons. Dixon, P., & Glover, S. (2009). Perseveration and contrast effects in grasping. Neuropsychologia, 47(6), 1578e1584. Doherty, M. J., Campbell, N. M., Tsuji, H., & Phillips, W. A. (2010). The Ebbinghaus illusion deceives adults but not young children. Developmental Science, 13(5), 714e721. Foster, R. M., & Franz, V. H. (2013). Inferences about time course of Weber's Law violate statistical principles. Vision Research, 78, 56e60. Franz, V. H., Gegenfurtner, K. R., Bu¨lthoff, H. H., & Fahle, M. (2000). Grasping visual illusions: No evidence for a dissociation between perception and action. Psychological Science, 11(1), 20e25. Ganel, T., Chajut, E., & Algom, D. (2008). Visual coding for action violates fundamental psychophysical principles. Current Biology, 18(14), R599eR601. Ganel, T., Chajut, E., Tanzer, M., & Algom, D. (2008). Response: When does grasping escape Weber's law? Current Biology, 18(23), R1090eR1091. Ganel, T., Freud, E., Chajut, E., & Algom, D. (2012). Accurate visuomotor control below the perceptual threshold of size discrimination. PLoS One, 7(4), e36253. Ganel, T., Freud, E., & Meiran, N. (2014). Action is immune to the effects of Weber's law throughout the entire grasping trajectory. Journal of Vision, 14(7), 11. Ganel, T., & Goodale, M. A. (2003). Visual control of action but not perception requires analytical processing of object shape. Nature, 426(6967), 664e667. Ganel, T., & Goodale, M. A. (2014). Variability-based Garner interference for perceptual estimations but not for grasping. Experimental Brain Research, 1e8. Ganel, T., Tanzer, M., & Goodale, M. A. (2008). A double dissociation between action and perception in the context of visual illusions opposite effects of real and illusory size. Psychological Science, 19(3), 221e225. Garner, W. R. (1978). Selective attention to attributes and to stimuli. Journal of Experimental Psychology: General, 107(3), 287. Garner, W. R. (2014). The processing of information and structure. Psychology Press. Gescheider, G. A. (1976). Psychophysics: Method and theory. Glover, S., & Dixon, P. (2002). Semantics affect the planning but not control of grasping. Experimental Brain Research, 146(3), 383e387. Goodale, M. A. (2011). Transforming vision into action. Vision Research, 51(13), 1567e1587. Goodale, M. A., & Milner, A. D. (1992). Separate visual pathways for perception and action. Trends in Neurosciences, 15(1), 20e25.
Goodale, M. A., & Westwood, D. A. (2004). An evolving view of duplex vision: Separate but interacting cortical pathways for perception and action. Current Opinion in Neurobiology, 14(2), 203e211. Haffenden, A. M., & Goodale, M. A. (1998). The effect of pictorial illusion on prehension and perception. Journal of Cognitive Neuroscience, 10(1), 122e136. € m, A. A. (2000). Sensation weighting in comparison and Hellstro discrimination of heaviness. Journal of Experimental Psychology: Human Perception and Performance, 26(1), 6. Jarmasz, J., & Hollands, J. G. (2009). Confidence intervals in repeated-measures designs: The number of observations principle. Canadian Journal of Experimental Psychology/Revue Canadienne de Psychologie Experimentale, 63(2), 124. Kopiske, K. K., Bruno, N., Hesse, C., Schenk, T., & Franz, V. H. (2016). The functional subdivision of the visual brain: Is there a real illusion effect on action? A multi-lab replication study. Cortex, 79, 130e152. LeBlanc, K. A., & Westwood, D. A. (2015). Sequential actions: Effects of upcoming perceptual and motor tasks on current actions. Experimental Brain Research, 1e8. Manzone, J., Davarpanah Jazi, S., Whitwell, R. L., & Heath, M. (2016). Biomechanical constraints do not influence pantomime-grasping adherence to Weber's law: A reply to Utz et al. (2015). Vision Research, 130, 31e35. Marks, L. E., & Algom, D. (1998). Psychophysical scaling. Mishkin, M., Ungerleider, L. G., & Macko, K. A. (1983). Object vision and spatial vision: Two cortical pathways. Trends in Neurosciences, 6, 414e417. Namdar, G., Ganel, T., & Algom, D. (2016). The extreme relativity of perception: A new contextual effect modulates human resolving power. Journal of Experimental Psychology: General, 145(4), 509. Smeets, J. B. J., & Brenner, E. (2008). Grasping Weber's law. Current Biology, 18(23), R1089eR1090. Team, J., & others. (2016). JASP (Version 0.7. 5.5). Computer Software. Utz, K. S., Hesse, C., Aschenneller, N., & Schenk, T. (2015). Biomechanical factors may explain why grasping violates Weber's law. Vision Research, 111, 22e30. Wagenmakers, E.-J. (2007). A practical solution to the pervasive problems ofp values. Psychonomic Bulletin & Review, 14(5), 779e804. Whitwell, R. L., & Goodale, M. A. (2016). Real and illusory issues in the illusion debate (Why two things are sometimes better than one): Commentary on Kopiske et al. (2016). Cortex. Q4
Please cite this article in press as: Namdar, G., et al., Dissociable effects of stimulus range on perception and action, Cortex (2017), http:// dx.doi.org/10.1016/j.cortex.2016.12.017
48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94
Available online at www.sciencedirect.com 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64
ScienceDirect Journal homepage: www.elsevier.com/locate/cortex
Special issue: Research report
Q6
Dissociable effects of stimulus range on perception and action
Q5 Q1
Gal Namdar a, Daniel Algom b and Tzvi Ganel a,* a b
Department of Psychology, Ben-Gurion University of the Negev, Beer-Sheva, Israel School of Psychological Sciences, Tel-Aviv University, Tel-Aviv, Israel
article info
abstract
Article history:
We have recently reported the discovery that the ability to detect a minimum increment to
Received 20 October 2016
a stimulus depends on the spread of the other stimuli for which this just noticeable dif-
Reviewed 9 December 2016
ference (JND) is being measured (Namdar, Ganel, & Algom, 2016). In particular, the JND
Revised 18 December 2016
around a standard stimulus was larger when the other standards tested within the same
Accepted 20 December 2016
experimental session spread a larger range. In this study we show that this range of
Published online xxx
standards effect (RSE) is limited to perceptual estimations and does not extend to action. The JND remained invariant when the participants grasped the objects rather than
Keywords:
perceptually estimated their size. This difference supports the hypothesis that visual
Perception and action
perception, on the one hand, and visually controlled action, on the other hand, are gov-
Grasping
erned by separate rules and mediated by different mechanisms.
Effects of context
© 2016 Elsevier Ltd. All rights reserved.
Relative processing Psychophysics
1.
Introduction
Perceptual resolution with respect to a given stimulus is contingent upon the context in which the stimulus is pre€ m, 2000). In sented (Namdar et al., 2016; see also, Hellstro particular, perceptual sensitivity with respect to the same stimulus is greater [i.e., the just noticeable difference (JND) or the difference threshold or limen, DL, is smaller] when the other stimuli tested are of similar magnitudes than when they are spread over a larger range of magnitudes. For example, the JND for a stimulus with a weight of 400 g was found smaller when the other standards tested within the same condition were 350 and 450 g than when they were 200 and 600 g. In
other words, the perceptual resolution for a single stimulus is contingent upon the resolution of the context, or stimulus set, in which it is embedded. Coarse resolution of the set (e.g., wider spacing between stimuli) leads to a decrease in the resolution of each stimulus within that set. Alternatively, it was also suggested that the range of standards effect (RSE) could have stemmed from adaptation effects to stimulus range, which result in an increased sensitivity to stimuli in a narrow compared to a wide range (Namdar et al., 2016). The RSE has been established across different perceptual modalities, including visual size discrimination and tactile weight perception. However, it is still unclear whether the RSE applies solely to the perceptual system. Does it also constrain performance in the domain of visuomotor control of action?
* Corresponding author. Department of Psychology, Ben-Gurion University of the Negev, Beer-Sheva 84105, Israel. E-mail address: [email protected] (T. Ganel). http://dx.doi.org/10.1016/j.cortex.2016.12.017 0010-9452/© 2016 Elsevier Ltd. All rights reserved. Please cite this article in press as: Namdar, G., et al., Dissociable effects of stimulus range on perception and action, Cortex (2017), http:// dx.doi.org/10.1016/j.cortex.2016.12.017
Q2
65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129
CORTEX1906_proof ■ 5 January 2017 ■ 2/6
2 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
Q3
c o r t e x x x x ( 2 0 1 7 ) 1 e6
Are grasping trajectories immune to the range of the stimuli being grasped? This seems plausible given evidence that grasping is less susceptible to influences of context than perception (e.g., Ganel & Goodale, 2003, 2014; Goodale & Westwood, 2004). We note though that a number of recent studies reported short-term contextual modulation of grasping, too (e.g., Dixon & Glover, 2009; LeBlanc & Westwood, 2015). However, if the RSE is found limited to visual perception so that it does not affect visually-controlled action, then a powerful diagnostic to dissociate perception and action at the level of a single stimulus becomes available. According to Goodale and Milner's influential account, vision for action and vision for perception are subserved by distinct (yet interactive) neuroanatomical streams (Goodale & Milner, 1992; Mishkin, Ungerleider, & Macko, 1983; but see; Franz, Gegenfurtner, Bu¨lthoff, & Fahle, 2000). Goodale and Milner suggested that the dorsal pathway in the primate visual system, which includes parietal regions, enables flexible control of actions directed at objects in the environment. The ventral pathway, by contrast, provides a rich and detailed representation of the environment sustaining visual perception (Goodale & Milner, 1992). Research within this framework suggests that the visuomotor system processes information in a selective manner more than does the visual perceptual system (Ganel & Goodale, 2003, 2014; Goodale & Westwood, 2004). This property often insulates grasping from the unwanted influence of task-irrelevant contextual information (Aglioti, DeSouza, & Goodale, 1995; Doherty, Campbell, Tsuji, & Phillips, 2010; Ganel, Chajut, & Algom, 2008; Ganel, Tanzer, & Goodale, 2008, but see; Franz et al., 2000; Utz, Hesse, Aschenneller, & Schenk, 2015). Several recent studies have suggested that there are some situations in which grasping trajectories can be modulated by short-term contextual information. For example, it has been reported that the trajectory of grip aperture is affected, at least to some extent, by the aperture of the immediate preceding trial (Dixon & Glover, 2009, see also; LeBlanc & Westwood, 2015). Additionally, recent evidence suggests that grasping, similarly to visual perception, can be affected by contextual information, such as that entailed in the Ebbinghaus illusion (Kopiske, Bruno, Hesse, Schenk, & Franz, 2016; but see; Ganel & Goodale, 2014; Haffenden & Goodale, 1998; Whitwell & Goodale, 2016). A different line of evidence suggests though, that unlike perceptual estimations, which are affected by irrelevant variations in stimulus shape, grasping is immune to the same effects (Ganel & Goodale, 2003, 2014). It is therefore an open question whether the RSE, an effect which has been only demonstrated in the perceptual domain, would extend to affect online control of actions.
1.1.
The present study
Our goal in this study was twofold. First, we wished to replicate the RSE so that it rests on an even firmer empirical basis. In the previous study (Namdar et al., 2016), we used the psychophysical method of Constant Stimuli. In this study, we made use of the Method of Adjustment (Baird & Noma, 1978; Gescheider, 1976; Marks & Algom, 1998). In this method, the JND is given by the standard deviation of the reproductions of the stimulus tested for resolution. An advantage of the
method is that the JND or DL derived can be easily compared between perception and grasping (Ganel, Chajut, & Algom, 2008; Ganel, Chajut, Tanzer, & Algom, 2008; Ganel, Freud, & Meiran, 2014). We expected that the JND for the same 40 mm stimulus would be larger when the other stimuli tested simultaneously are 20 and 60 mm than when the other stimuli are 35 and 45 mm. Secondly and more important, we expected that this effect would be limited to perception: grasping the same stimuli would not result in modification of the JND for the common 40 mm stimulus. Therefore, in the perceptual condition, the participants were asked to make perceptual estimations of size, whereas in the grasping condition they were asked to perform visually-guided grasping movements toward the same objects. In a control experiment we also tested whether the hypothesized insensitivity of the grasping trajectories to the measuring context can possibly be confounded by on-line corrections based on visual feedback from the fingers and the target object (Glover & Dixon, 2002). Such online corrections could mask contextual effects on grip apertures. To address this concern, grasping movements in the control experiment were performed in an open-loop design, for which visual feedback was prevented during grasp.
2.
Methods
2.1.
Participants
A group of 14 right-handed students from Ben-Gurion University of the Negev, who gave their informed consent, participated in the main experiment (3 males, mean age of 23.35, SD ¼ 1.59 years). A separate group of 14 right-handed participants (8 males, mean age of 23.57, SD ¼ 2.65 years) participated in the control, open-loop grasping experiment. All participants signed a consent form and the experimental protocol was approved by the departmental ethics committee in accordance with the Code of Ethics of the World Medical Association (Declaration of Helsinki).
2.2.
Apparatus and stimuli
The stimuli were five plastic rods of a constant width and height of 5 mm. The values of length were 20, 35, 40, 45, and 60 mm. The participants set in front of a black tabletop with the tips of the index finger and thumb (right hand) resting on a small starting button fixed to the center of the tabletop. The participants wore a set of LCD glasses (Translucent Technologies, Toronto, ON) with liquid-crystal shutter lenses used to control stimulus exposure time. The experimenter manually switched between the rods prior to each trial in a pseudorandomized order. Each rod was placed 30 cm from the participant, in the center of the table, along the midline of the participant. Rods were placed on the table perpendicularly to the participant's viewing plane (Fig. 1).
2.3.
Design
The main experiment consisted of two separate, closed-loop tasks, a grasping task and a manual estimation task. In the grasping task, the participants were asked to grasp the rod
Please cite this article in press as: Namdar, G., et al., Dissociable effects of stimulus range on perception and action, Cortex (2017), http:// dx.doi.org/10.1016/j.cortex.2016.12.017
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
CORTEX1906_proof ■ 5 January 2017 ■ 3/6 c o r t e x x x x ( 2 0 1 7 ) 1 e6
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
3
Fig. 1 e An illustration of (a) a grasping movement towards one of the objects and (b) a manual perceptual estimation of the object's size.
along its length using their thumb and index finger, and then lifted the rod a few centimeters in the air before placing it back on the table and returning their hand to the starting position. In the manual estimation task, the participants gave estimates of the rod's length without grasping it. They did so by reaching to a designated location about 20 cm to the rod's right side while opening their finger and thumb to an extent that matched the perceived length of the rod (see Fig. 1). Each task included two experimental blocks, the wide and the narrow range conditions. The wide range condition consisted of stimuli of 20 mm, 40 mm and 60 mm in length, whereas the narrow range condition consisted of stimuli of 35 mm, 40 mm and 45 mm in length. Notice that both experimental blocks included the common stimulus of 40 mm, and that they differed only by the flanking stimuli, defining the range. In both range conditions (in each task), each of the three stimuli was presented 20 times, resulting in a total of 240 trials in all. Both range conditions under both tasks began with 10 training trials which were not included in the analysis of the data. The goal of training was to familiarize the participants with the range of stimuli used in the block. Order of task and range was counterbalanced across participants. Each trial began with the opening of the goggles for a total duration of 2000 msec, providing sufficient time for completing the grasping of the object or estimating its size. In a control experiment, a separate group of participants performed grasping movements under the exact same conditions as in the grasping condition of the main experiment. However, unlike as in the closed-loop experiment, participants' vision was occluded at the moment that they initiated their movements. By doing so, we eliminated the participants' ability to perform on-line corrections based on visual feedback.
2.4.
thumb throughout the entire grasping movement trajectory (at a 200-Hz sampling rate). Movement onset was determined as the point in time at which the aperture between index finger and thumb increased by more than .1 mm for at least ten successive frames (of 50 msec). Movement offset was determined as the point in time at which the aperture between index finger and thumb changed by no more than .1 mm for at least ten successive frames, but only after reaching the MGA between fingers in the grasping condition. Trials were excluded from the analysis if the infrared diodes placed on the fingers were not visible to the camera while moving toward the object or if the participant failed to grasp the object properly and dropped it while trying to lift it off the table. This resulted in 4.7% and 2.8% of the trials being excluded from the grasping and manual estimation conditions, respectively.
2.5.
Data analysis
In order to calculate the JND for the grasping condition, we derived the MGA for each grasping movement trajectory (Foster & Franz, 2013; Ganel & Goodale, 2014; Ganel, Freud, Chajut, & Algom, 2012). In the manual estimation (perception) condition, we measured the aperture between the index finger and the thumb at the end of the estimation. The JNDs in both conditions are given by the respective standard deviations of the responses to each stimulus across trials. A correction for outliers was performed on the data by eliminating any score that was either 2.5 standard deviations below or above the mean score for each condition, resulting in the omission of less than 2% of the trials.
Kinematic recordings
3. An Optotrak Certus device (Northern Digital, Waterloo, ON) was used to track the hand and finger grasping trajectories. Two infrared light-emitting diodes were placed on the right hand thumb and index finger (located on top of the center of the distal phalanges), and another diode was placed on the wrist of the right hand. We collected data about the location of each diode and the aperture between the index finger and
Results
Consider first the result for perception, i.e., for the magnitude estimates conveyed through the fingers. On average, the JND for the common rod of 40 mm was 4.47 mm when tested along with flankers of 20 mm and 60 mm, but was 3.17 when tested along with flankers of 35 mm and 45 mm. The JND was larger by roughly 34% in the large-range condition than in the small-
Please cite this article in press as: Namdar, G., et al., Dissociable effects of stimulus range on perception and action, Cortex (2017), http:// dx.doi.org/10.1016/j.cortex.2016.12.017
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
CORTEX1906_proof ■ 5 January 2017 ■ 4/6
4
JND (mm)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
c o r t e x x x x ( 2 0 1 7 ) 1 e6
4.9
Grasping
4.4
Perceptual esƟmaƟons
3.9 3.4 2.9 2.4 Narrow
Wide
Range of standards
Fig. 2 e Effects of standards range on perceptual estimations and on maximum grip apertures during grasping of a 40 mm stimulus. Error bars represent confidence intervals of 95% (Jarmasz & Hollands, 2009). range condition. Therefore, resolution improved considerably under the small-range condition. Consider next the JND for the same rod of 40 mm determined through actual grasping. The JND for the common rod of 40 mm was 3.08 mm when tested along with flankers of 20 mm and 60 mm, and was 2.93 mm when tested with flankers of 35 mm and 45 mm. Clearly, the JND was insensitive to the spread of the flankers with grasping e in sharp contrast with the results for perception. The difference between the two visually-governed functions is supported by the interaction of task (estimation, grasping) and range (narrow, wide) in a repeated-measure ANOVA [F(1,13) ¼ 7.08, p < .05, h2p ¼ .35] (see Fig. 2). Planned comparisons revealed that range affected manual estimations [F(1,13) ¼ 9.07, p < .01], but not grasping [F(1.13) < 1, nsec]. The contingent nature of the RSE thus provides a rigorous dissociation between perception and action (see Table 1 for the data of individual participants). The range of stimuli had no effect on the average manual estimations for the 40 mm stimulus under the narrow-range condition (54.47 mm) compared with the wide-range condition (54.29 mm) [t(13) < 1, nsec]. Too, no effects of range were
found for the grasping task (65.90 and 66.51 mm for the average MGA under the narrow and wide conditions, respectively. [t(13) ¼ 1.08, nsec]). Average MGAs for the 20, 35, 45 and 60 mm objects were 51.6, 62.8, 67.9 and 78.9 mm, respectively. Average manual estimations for the 20, 35, 45 and 60 mm objects were 37.6, 48.5, 59.4 and 72.4 mm, respectively. In the control experiment, the average standard deviation during MGA was 5.11 mm in the narrow range condition and 4.73 mm in the wide range condition [t(13) ¼ .48, nsec]. To further test the null-effect of the measurement context on grip apertures under the open-loop condition, we performed a Bayesian t-test analysis to estimate the probability of the null hypothesis to be true given the aperture data (Wagenmakers, 2007). The Bayes factor was calculated using the JASP software package (Team & others, 2016). The analysis revealed a Bayes factor of 5.091 (BF0) indicating that the null hypothesis is roughly 5 times more likely compared to the alternative hypothesis. Finally, the JNDs for the grasping conditions from the main experiment and the control experiment were directly compared by a mixed-effect ANOVA analysis with range (Narrow/Wide) as a within-subject factor, and experiment (Main/Control) as a between-subjects factor. The analysis revealed a main effect of experiment [F(1,26) ¼ 23, p < .0001, h2p ¼ .46], and no main effect of range [F(1,26) < 1, nsec]. More importantly, the interaction between experiment and range [F(1,26) < 1, nsec] was nonsignificant. In order to test for possible effects for context in parts of the grasping trajectory other than the MGA (see Dixon & Glover, 2009), grasping and manual estimations trajectories for the 40 mm stimulus in the main experiment were normalized and divided to 11 equal-sized intervals. In each interval, the effect of context range was calculated as the difference in the standard deviations between the two range conditions. Post-hoc comparisons revealed no effect for range in any of the normalized time points for grasping trajectories. This provides further evidence that grasping is unaffected by the range of the contextual information throughout the entire movement. In contrast, analysis of the manual estimations trajectories revealed a
Table 1 e RSE and JND, calculated as standard deviations for action (grasping) and perception (manual estimations) of the 40 mm stimulus under both range conditions of the main experiment for each individual participant. DRSE denotes the difference in RSE between the action and the perception conditions. Participant
JND (SD)
RSE (WideeNarrow)
Action
1 2 3 4 5 6 7 8 9 10 11 12 13 14 Average
Perception
Narrow
Wide
Narrow
Wide
Action
Perception
DRSE
3.50 3.42 1.85 2.59 3.97 3.38 2.25 3.12 2.15 2.27 3.17 1.85 2.06 5.42 2.93
2.89 3.79 3.80 2.59 2.24 3.42 2.47 2.71 4.14 1.25 3.34 2.48 2.66 5.27 3.08
2.64 4.21 1.29 2.91 2.88 2.88 3.28 2.57 2.77 6.04 2.68 2.91 3.38 3.99 3.17
5.31 5.39 6.37 5.11 3.73 4.43 5.05 3.36 1.95 5.00 4.54 5.10 4.05 3.18 4.47
.60 .36 1.96 .01 1.72 .04 .22 .42 1.99 1.02 .16 .64 .60 .16 .15
2.66 1.18 5.08 2.20 .85 1.55 1.77 .79 .83 1.03 1.86 2.19 .67 .81 1.29
3.27 .82 3.12 2.20 2.57 1.51 1.55 1.21 2.82 .01 1.69 1.56 .07 .66 1.15
Please cite this article in press as: Namdar, G., et al., Dissociable effects of stimulus range on perception and action, Cortex (2017), http:// dx.doi.org/10.1016/j.cortex.2016.12.017
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
CORTEX1906_proof ■ 5 January 2017 ■ 5/6 c o r t e x x x x ( 2 0 1 7 ) 1 e6
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
strong effect for context, beginning at 50% of the movement and throughout the entire movement [F(1,13) ¼ 11.68, p < .01]. The data for the manual estimations condition largely (though not fully) obeyed Weber's Law. In the wide range condition, the JNDs for stimuli of 20, 40, and 60 mm were 3.04, 4.47, and 5.45 mm, respectively. In the narrow range condition, the JNDs for stimuli of 35, 40 and 45 mm were 4.09, 3.17, and 4.17 mm, respectively. Note that in the narrow range, the JND pattern was not fully consistent with Weber's law, probably due to low statistical power due to the relatively small differences in length between the stimuli. In contrast to the general adherence to Weber's law during manual estimations, JNDs during grasping did not obey Weber's Law. The JNDs in the wide range condition did not increase with object size and were 3.28, 3.08 and 2.83 mm for the stimuli of 20, 40 and 60 mm, respectively. In the narrow range condition, the JNDs were 3.05, 2.93 and 3.08 mm for the stimuli of 35, 40 and 45 mm, respectively. The diverging pattern of adherence to Weber's law is consistent with the results of previous studies (Ganel & Goodale, 2014; Ganel, Chajut & Algom, 2008; Smeets & Brenner, 2008). We note that a recent study (Utz et al., 2015) showed that JNDs under certain experimental conditions can be affected by ceiling effects that stem from the participants' limited range of motion (but see Manzone, Davarpanah Jazi, Whitwell, & Heath, 2016 for a different conclusion). In the Utz et al.'s (2015) study, MGAs were relatively large due to the design used (mirror apparatus in which no feedback of the hand or the object was provided). This setup is conductive to ceiling effects (Manzone et al., 2016). Average MGAs and apertures were much smaller in the present study, which minimizes the possibility to reach ceiling effects.
4.
Discussion
We provided evidence for the influence of contextual information on perceptual acuity, but for the absence of such influence on visuomotor control. Our results show that stimulus resolution is affected by the range of the other stimuli tested for resolution. A wider range is conductive to a larger JND. However, this effect, startling as it is, is limited to perception. In contrast, resolution was insensitive to context when the same stimuli were grasped rather than estimated. These results join previous indications that action is affected by taskirrelevant contextual information to a lesser degree than is perception (Ganel & Goodale, 2003; Goodale, 2011). Moreover, the current results show that grasping movements are also immune to the specific effects of contextual information (Namdar et al., 2016), as range differences between blocks had no effect on grasping trajectory deviations. Context had no effect on grasping even when the participants grasped the object with their vision occluded, suggesting that the lack of contextual effect on grasping is not the result of on-line corrections.
4.1.
Dissociation between perception and action
The results of the present study suggest that grasping is immune to the effect of context. Previous indications for shortterm effects of context on grasping trajectories (e.g., Dixon & Glover, 2009; LeBlanc & Westwood, 2015) are not necessarily
5
in odds with the result of the present study. First, those studies focused on different measures. In particular, previous studies looked at response bias in grip apertures, whereas the focus of the current study was on the resolution (JND) of the response. Furthermore, earlier studies looked at short-term effects of the previous trial, whereas here we examined the effect of context on visual resolution for perception and for action across a range of stimuli. Therefore, visual resolution during grasping escapes the effects of global, adaptation to context. In a certain sense, the RSE tallies with effects found in Garner's speeded-classification task (Garner, 1978). In the Garner task, participants classify objects on a task-relevant dimension while another, task-irrelevant dimension, is either held constant (baseline condition) or varies from-trial-to-trial (filtering condition). Worse performance in the filtering condition (called Garner interference) indicates that the task-irrelevant information could not be ignored. Notably, Garner interference was found for visual perception but was not found for grasping (Ganel & Goodale, 2003, 2014). It is possible that Garner interference and RSE are rooted in the same source with regard to the effects of contextual information on performance. In the case of Garner interference, variability along the values of the irrelevant dimension leads to a decrease in perceptual performance. The RSE documents the effect on perception of larger variability of the dimension tested for resolution.
5.
Conclusion
The present results highlight the different nature of the computations underlying perceptual estimations and visuomotor control. The former is sensitive to contextual information about the stimulus of interest, a contingency that is bound to lead to the directional effect on visual resolution documented in this study. In contrast, the visuomotor control system is less susceptible to contextual information that does not strictly pertain to the task at hand. The hallmark of the action system is its selectivity to the relevant feature of to-begrasped object. The present results provide a compelling new line of evidence for the dissociation between visually guided actions and visual perception.
Uncited references Garner, 2014; Namdar et al., 2016.
Acknowledgements This study was supported by an Israel Science Foundation (ISF) grant 274/15 to Tzvi Ganel and to Daniel Algom.
references
Aglioti, S., DeSouza, J. F., & Goodale, M. A. (1995). Size-contrast illusions deceive the eye but not the hand. Current Biology, 5(6), 679e685.
Please cite this article in press as: Namdar, G., et al., Dissociable effects of stimulus range on perception and action, Cortex (2017), http:// dx.doi.org/10.1016/j.cortex.2016.12.017
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
CORTEX1906_proof ■ 5 January 2017 ■ 6/6
6 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47
c o r t e x x x x ( 2 0 1 7 ) 1 e6
Baird, J. C., & Noma, E. J. (1978). Fundamentals of scaling and psychophysics. John Wiley & Sons. Dixon, P., & Glover, S. (2009). Perseveration and contrast effects in grasping. Neuropsychologia, 47(6), 1578e1584. Doherty, M. J., Campbell, N. M., Tsuji, H., & Phillips, W. A. (2010). The Ebbinghaus illusion deceives adults but not young children. Developmental Science, 13(5), 714e721. Foster, R. M., & Franz, V. H. (2013). Inferences about time course of Weber's Law violate statistical principles. Vision Research, 78, 56e60. Franz, V. H., Gegenfurtner, K. R., Bu¨lthoff, H. H., & Fahle, M. (2000). Grasping visual illusions: No evidence for a dissociation between perception and action. Psychological Science, 11(1), 20e25. Ganel, T., Chajut, E., & Algom, D. (2008). Visual coding for action violates fundamental psychophysical principles. Current Biology, 18(14), R599eR601. Ganel, T., Chajut, E., Tanzer, M., & Algom, D. (2008). Response: When does grasping escape Weber's law? Current Biology, 18(23), R1090eR1091. Ganel, T., Freud, E., Chajut, E., & Algom, D. (2012). Accurate visuomotor control below the perceptual threshold of size discrimination. PLoS One, 7(4), e36253. Ganel, T., Freud, E., & Meiran, N. (2014). Action is immune to the effects of Weber's law throughout the entire grasping trajectory. Journal of Vision, 14(7), 11. Ganel, T., & Goodale, M. A. (2003). Visual control of action but not perception requires analytical processing of object shape. Nature, 426(6967), 664e667. Ganel, T., & Goodale, M. A. (2014). Variability-based Garner interference for perceptual estimations but not for grasping. Experimental Brain Research, 1e8. Ganel, T., Tanzer, M., & Goodale, M. A. (2008). A double dissociation between action and perception in the context of visual illusions opposite effects of real and illusory size. Psychological Science, 19(3), 221e225. Garner, W. R. (1978). Selective attention to attributes and to stimuli. Journal of Experimental Psychology: General, 107(3), 287. Garner, W. R. (2014). The processing of information and structure. Psychology Press. Gescheider, G. A. (1976). Psychophysics: Method and theory. Glover, S., & Dixon, P. (2002). Semantics affect the planning but not control of grasping. Experimental Brain Research, 146(3), 383e387. Goodale, M. A. (2011). Transforming vision into action. Vision Research, 51(13), 1567e1587. Goodale, M. A., & Milner, A. D. (1992). Separate visual pathways for perception and action. Trends in Neurosciences, 15(1), 20e25.
Goodale, M. A., & Westwood, D. A. (2004). An evolving view of duplex vision: Separate but interacting cortical pathways for perception and action. Current Opinion in Neurobiology, 14(2), 203e211. Haffenden, A. M., & Goodale, M. A. (1998). The effect of pictorial illusion on prehension and perception. Journal of Cognitive Neuroscience, 10(1), 122e136. € m, A. A. (2000). Sensation weighting in comparison and Hellstro discrimination of heaviness. Journal of Experimental Psychology: Human Perception and Performance, 26(1), 6. Jarmasz, J., & Hollands, J. G. (2009). Confidence intervals in repeated-measures designs: The number of observations principle. Canadian Journal of Experimental Psychology/Revue Canadienne de Psychologie Experimentale, 63(2), 124. Kopiske, K. K., Bruno, N., Hesse, C., Schenk, T., & Franz, V. H. (2016). The functional subdivision of the visual brain: Is there a real illusion effect on action? A multi-lab replication study. Cortex, 79, 130e152. LeBlanc, K. A., & Westwood, D. A. (2015). Sequential actions: Effects of upcoming perceptual and motor tasks on current actions. Experimental Brain Research, 1e8. Manzone, J., Davarpanah Jazi, S., Whitwell, R. L., & Heath, M. (2016). Biomechanical constraints do not influence pantomime-grasping adherence to Weber's law: A reply to Utz et al. (2015). Vision Research, 130, 31e35. Marks, L. E., & Algom, D. (1998). Psychophysical scaling. Mishkin, M., Ungerleider, L. G., & Macko, K. A. (1983). Object vision and spatial vision: Two cortical pathways. Trends in Neurosciences, 6, 414e417. Namdar, G., Ganel, T., & Algom, D. (2016). The extreme relativity of perception: A new contextual effect modulates human resolving power. Journal of Experimental Psychology: General, 145(4), 509. Smeets, J. B. J., & Brenner, E. (2008). Grasping Weber's law. Current Biology, 18(23), R1089eR1090. Team, J., & others. (2016). JASP (Version 0.7. 5.5). Computer Software. Utz, K. S., Hesse, C., Aschenneller, N., & Schenk, T. (2015). Biomechanical factors may explain why grasping violates Weber's law. Vision Research, 111, 22e30. Wagenmakers, E.-J. (2007). A practical solution to the pervasive problems ofp values. Psychonomic Bulletin & Review, 14(5), 779e804. Whitwell, R. L., & Goodale, M. A. (2016). Real and illusory issues in the illusion debate (Why two things are sometimes better than one): Commentary on Kopiske et al. (2016). Cortex. Q4
Please cite this article in press as: Namdar, G., et al., Dissociable effects of stimulus range on perception and action, Cortex (2017), http:// dx.doi.org/10.1016/j.cortex.2016.12.017
48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94