Abstract
The shape of an object restricts its movements and therefore its future location. The rules governing selective sampling of the environment likely incorporate any available data, including shape, that provide information about where important things are going to be in the near future so that the object can be located, tracked, and sampled for information. We asked people to assess in which direction several novel objects pointed or directed them. With independent groups of people, we investigated whether their attention and sense of motion were systematically biased in this direction. Our work shows that nearly any novel object has intrinsic directionality derived from its shape. This shape information is swiftly and automatically incorporated into the allocation of overt and covert visual orienting and the detection of motion, processes which themselves are inherently directional. The observed connection between form and space suggests that shape processing goes beyond recognition alone and may help explain why shape is a relevant dimension throughout the visual brain.
Keywords: Shape, direction, attention, eye movements, motion
The visual world is not static; within it, things are moving and we are often moving ourselves — if not our bodies then at least our eyes which constantly scan the visual scene. Processing dynamic input requires efficient extraction of information about the current state of the environment to make predictions about where important things will be in the near future. We should guide our eyes and attention not to an object’s previous location, but to where it is likely to be once action can be taken. Fortunately, under normal circumstances, an object does not randomly change location from one moment to the next; its future state depends on its past state. An optimized system would be able to use such information to accurately predict an object’s future location or motion path from a single snapshot in time. This could bias both overt and covert visual orienting so that objects can be located, tracked, and sampled even in a dynamic world. Here we test the hypothesis that information derived from an object’s shape enables the brain to make such inferences.
Within the visual system, the dorsal pathway’s role in visual orienting, tracking, and motion analysis is well-established (Andersen, 1997; Colby & Goldberg, 1999; Mountcastle, Lynch, Georgopoulos, Sakata, & Acuna, 1975; Ungerleider & Mishkin, 1982; Van Essen & Gallant, 1994). In addition, some regions of the dorsal stream are responsive to the shape of objects (Grill-Spector & Malach, 2004; Janssen, Srivastava, Ombelet, & Orban, 2008; Konen & Kastner, 2008; Lehky & Sereno, 2007; Murata, Gallese, Luppino, Kaseda, & Sakata, 2000; Oliver & Thompson-Schill, 2003; Red, Patel, & Sereno, 2012; Sakata et al., 1998; Sakata, Taira, Murata, & Mine, 1995; A. B. Sereno & Amador, 2006; A. B. Sereno & Maunsell, 1998; M. E. Sereno, Trinath, Augath, & Logothetis, 2002; Taira, Mine, Georgopoulos, Murata, & Sakata, 1990). The fact that shape selectivity exists in cortical areas beyond the ventral visual stream (Desimone, Albright, Gross, & Bruce, 1984; Gross, Rocha-Miranda, & Bender, 1972; Logothetis & Sheinberg, 1996; Tanaka, Saito, Fukada, & Moriya, 1991) argues against regional specialization for particular stimulus attributes, emphasizing the need to consider function and goal in relation to object properties. Shape information might be integrated with various other cues and tailored to a particular process or task. Indeed, the shape of an object influences processes thought to depend on the dorsal visual stream, such as visual orienting and estimation of motion, in addition to object recognition and categorization, which are classically linked with the ventral visual stream.
For example, the oculomotor system seems able to take into account the global shape of an object during saccade planning (He & Kowler, 1991). This kind of visual orienting does not merely depend on low-level averaging of visual elements but has access to a higher level representation of the object’s shape (Melcher & Kowler, 1999). This shape information may be partially or wholly independent from the representation used for perception (Vishwanath, Kowler, & Feldman, 2000). The shape of an object guides overt and covert attention within the object itself and can, in special cases, push attention away (Driver et al., 1999; Fischer, Castel, Dodd, & Pratt, 2003; Friesen & Kingstone, 1998; Hommel, Pratt, Colzato, & Godijn, 2001; Kuhn & Kingstone, 2009; Tipples, 2002, 2008). An arrow is a prime example. Despite initial thoughts to the contrary (Jonides, 1981), arrows automatically bias orienting (Hommel et al., 2001; Kuhn & Kingstone, 2009; Tipples, 2002, 2008). This may be partially due to repeated association of this particular shape and its referent; something often appears in the direction to which an arrow is pointing. Here, we argue that this association is not arbitrary; initially, the symbol might have been selected because its shape already had an inherent directionality that automatically evoked an orienting bias. This bias might again be derived from the fact that the structure of a real arrow facilitates a stable flight path in a single direction. In general, the shape of objects constrains their movements. It would therefore be beneficial for the visual system to use shape information to predict an object’s probable motion path and to use such predictions for overt and covert visual orienting.
Shape or form cues are integrated into motion calculations, (see Kourtzi, Krekelberg, & van Wezel, 2008, for a review). For example, the oriented trace or streak left by a fast-moving object determines its perceived axis of motion (Burr & Ross, 2002; Geisler, 1999). Dynamic Glass patterns, which contain no coherent motion, can also lead to the perception of movement and affect the tuning of motion selective neurons (Krekelberg, Dannenberg, Hoffmann, Bremmer, & Ross, 2003; Krekelberg, Vatakis, & Kourtzi, 2005; Ross, Badcock, & Hayes, 2000). Likewise, still photographs depicting objects in motion evoke greater activation in motion selective cortical regions than do photographs of stationary objects (Kourtzi & Kanwisher, 2000; Senior et al., 2000). An object’s remembered location is also shifted along its implied path of motion (Freyd & Finke, 1984). This shift is lost when motion selective cortical regions are temporarily deactivated (Senior, Ward, & David, 2002). With some exceptions (e.g. Caplovitz & Tse, 2007; Tse & Logothetis, 2002), most studies on the effects of form or shape cues on motion involved simple non-object-like stimuli (e.g. motion streaks, Glass patterns), or recognizable animate or inanimate objects or scenes depicting familiar events or actions.
The studies described in this paper were stimulated by the idea that shape information existing in dorsal stream regions is tailored to and supports the function of these areas in spatial perception and action guidance (Goodale & Milner, 1992; Milner & Goodale, 1995; Ungerleider & Mishkin, 1982). We explore the role of shape information in visual orienting and motion calculation which have well-known neural substrates in the dorsal stream. In experiments 1 and 2, we establish that people consistently deem novel shapes to “point” in particular directions. We then look at the effects of this shape-derived directionality on visual orienting (experiments 3, 4, and 7) and motion perception (experiments 5, 6, and 7).
Our work shows that objects have intrinsic directionality derived from their shape. This shape information is swiftly and automatically incorporated into the allocation of overt and covert visual orienting and the detection of motion, processes which are inherently directional. While covert attention might be split under some unusual circumstances (Awh & Pashler, 2000; Hahn & Kramer, 1998; Kramer & Hahn, 1995), our eyes only move in one direction at a time. Likewise, a single object only moves in one direction at any given time point. Attention is automatically pushed away from the object in a direction that depends on the object’s shape. This in turn is incorporated into the calculation of the object’s probable path of movement; detection of an object’s direction of motion is facilitated if it is congruent with the inherent shape-derived directionality of the object and hindered if shape directionality and motion directionality oppose each other. Importantly, such form-dependent directional biases are not limited to well-known or over-learned objects or tasks. Instead, they are seen for meaningless shapes, with which people have no prior experience, in a variety of settings and regardless of whether people have any intent or reason to use this directional information. This suggests that shape-related directional biases are ever present and are given weight in predictions or simulations of the upcoming state of the environment or, more specifically, where important objects will likely be located in the immediate future.
Experiments: Methodological Overview
A total of 114 people participated in one of seven experiments. Each person took part only once. They reported normal or corrected to normal vision and were paid for their participation. All participants gave their written consent. The experimental protocol was approved by Brown University’s Institutional Review Board.
The experiments were controlled by a computer console running on the QNX real-time operating system (QSSL; QNX Software Systems). It communicated with a Windows XP PC through a direct high-speed Ethernet connection. This computer ran custom-made software based on OpenGL for graphics display. In experiments 1–3, stimuli were shown on a standard 20″ cathode ray tube monitor (width: 41 cm; height: 30 cm) with 1024 × 768 resolution. In experiments 5–7, they were shown on a high speed 23″ widescreen LCD monitor (width: 51 cm; height: 28 cm) with 1920 × 1080 resolution. In experiment 4, half of the participants were run using the former setup, and half the latter. The monitors’ vertical refresh rate was 100 Hz for all experiments. The displays were placed at a distance of 57 cm in front of the subjects.
Participants were seated in a dark, quiet room in front of a computer screen. A black curtain was draped around them and the computer screen. Participants’ heads were held still by a chin rest. In experiments 3, 4, and 7, people’s eye movements were monitored with an EyeLink 1000 eye tracker (SR Research). A high-speed camera and an infrared light source were desk-mounted under the computer monitor. Eye gaze was monocularly recorded at 1000 Hz. The analog signal was sampled and digitized at 200 Hz. The eye tracker was calibrated by asking participants to saccade to and fixate several small targets that appeared in random locations on the screen. Eye tracking was not performed in experiments 1, 2, 5, and 6.
An alpha level of 0.05 is assumed for all statistical analysis of the data. Statistical tests are two-sided. Results are Greenhouse-Geisser corrected for deviations from sphericity when Mauchly’s test of sphericity is significant. Effect sizes are estimated using Pearson’s r, Cohen’s d (mean difference/standard deviation of difference) and partial eta squared (ηp2). Error bars represent 95% confidence intervals for within-subject comparisons and are calculated using Cousineau’s method (2005) with the correction described by Morey (2008).
Experiment 1: Directionality Assessment
The aim of experiment 1 was to assess the extent to which people agree on the directionality of objects based on their shape alone. We did not want to constrain the interpretation of our results with our preconceptions about what might make a shape directional. We therefore constructed a variety of random shapes with which people had no previous experience and empirically determined their directionality. We asked people to judge where each of the novel shapes pointed or directed them and determined if people’s judgments were more similar than would be expected by chance. If judgments of a majority of the shapes deviate from circular uniformity, we would conclude that directionality is a general property of a wide variety of shapes.
Method
Participants
16 people (9 women) participated in this experiment. Their ages ranged from 18 to 36 (M=25).
Stimuli
80 novel shapes were generated by superimposing two filled polygons. Each polygon was made by fitting a spline to randomly generated coordinates (8 for simple or 16 for complex shapes) on a 2D plane. The algorithm was based on the General Polygon Clipper library (v. 2.32) which is freely available for non-commercial use (Murta, 2000; see also Vatti, 1992). Shapes were scaled to an equal area. Their diameter was approximately 4°. Of the 80 shapes, 20 were made symmetrical by reflecting one side around the y-axis. The contours of the shapes were densely sampled and translated so that the means of their contour coordinates would coincide. Each shape was randomly rotated around this pivot and kept this rotation throughout the experiment and for all participants. The same method was used to make additional shapes for a short practice session. All shapes were shown as white, filled silhouettes. The shapes can be seen in figure 5.
Figure 5. All novel shapes.
Shapes are shown pointing to the right, as judged by people in experiment 2. All novel shapes were used in experiments 1 and 2. Red and orange shapes were used in experiments 3, 5, 6, and 7. Red shapes were used in experiment 4. Shapes were not shown in color during the experiments.
Design
Each person completed 360 trials, out of which 40 were control trials and 320 were experimental trials (80 shapes × 4 repetitions). The trials were spread across five blocks and were shown in a randomized order with the constraint that 8 control trials were shown in each block.
Procedure
People were instructed to look to the center of the screen at the beginning of each trial. A single central shape and a surrounding gray circle (diameter 26°) were presented on a black background. The shape was on for 100 ms but the circle stayed visible throughout the trial. The task is depicted in figure 1.
Figure 1. Task in experiment 1.
People judged the directionality of novel shapes by dragging a line in the direction to which they thought the shapes pointed or directed them.
People used a computer mouse to drag a gray line in the direction to which they thought the shape pointed or directed them; longer lines indicated stronger confidence. The line was drawn in real time from the screen center to the current position of a gray circular cursor (diameter 0.4°) and could be drawn as far as to the surrounding circle. Participants clicked the left mouse button to indicate their response. They were encouraged to not think much about their responses but instead go with their intuition.
In a minority of trials, no shape was shown and participants instead dragged a line to the position of a small disk. All participants performed well on these control trials, ensuring us that they paid attention to the task at hand and could position the line appropriately. Before beginning the experiment, people completed a short practice block.
Results
Each of the 16 participants judged the direction of each shape four times, giving a total of 64 data points for each of the 80 random shapes. Example shapes with their directional judgments can be seen in figure 2.
Figure 2. Example shapes and directional judgments.
Several novel shapes are shown with the endpoints (white circles) of people’s “drag-and-clicks” used for directional judgments in experiment 1.
We tested for circular uniformity of the directional judgments of each shape. Visual inspection of the click endpoints indicated that some of the shapes were unidirectional, some were bidirectional, and yet others were multidirectional. We therefore performed two different statistical tests on each shape: A Raleigh test and Rao’s spacing test. The Raleigh test assumes that the samples are drawn from a von Mises distribution (analogous to the normal distribution for non-circular data) and is useful for detecting deviations from uniformity when a shape has one main direction (Berens, 2009). Rao’s spacing test can detect deviations from a uniform distribution for shapes that are neither unidirectional nor axially bidirectional (Berens, 2009).
A participant’s decision criterion for assigning directionality could evolve over the course of the experiment. For each shape, we therefore tested for significant deviations from circular uniformity using only the first directional judgment of each participant. Instead of using all 64 judgments, we therefore only used 16 data points per shape, effectively lowering our statistical power. Despite this rather conservative way of analyzing the data, the Raleigh test rejected the null hypothesis that the drag-and-clicks were uniformly distributed for 42 out of 80 shapes. Rao’s spacing test was significant for nearly all of the shapes, or 71 out of 80. We therefore conclude that novel, random shapes in general are directional. A majority of completely novel shapes has an inherent directionality, be it unidirectional, bidirectional, or multidirectional. Supplementary figure 1 shows our entire shape set and the corresponding directionality judgments from experiment 1. Supplementary figure 2 shows all directional judgments from experiment 1, regardless of shape. Test statistics can be found in supplementary tables 1 and 2.
Experiment 2: Forced Choice of Directionality
Experiment 1 showed that directional judgments are non-uniform for a majority of randomly shaped novel objects. However, judgments also appeared to be influenced by factors that were independent of, or interacted with, the shape of these objects (supplementary figure 2). Even if the rotation of the shapes was randomly determined, people in general tended to favor an upward and to some lesser extent a downward direction. People might have been following a heuristic akin to “when in doubt, an object is aligned to the axis of gravity”. The benefits of transient visual attention have also been documented to be greater in the upper than the lower visual hemifield (Kristjánsson & Sigurdardottir, 2008), and this could be a contributing factor. It is also possible that the response mode introduced some bias. To minimize such biases, experiment 2 involved a more constrained judgment about the directionality of the same shapes with a new set of subjects.
The main purpose of experiment 2 was to get unbiased measurements of each shape’s perceived directionality so that these measures could be used as predictors of behavior in experiments 3–7. We also wanted to know whether we could assume that directionality was independent of the time of probing. Neurons within dorsal stream regions important for the allocation of attention and eye movements respond selectively to shapes, but these shape responses can change very rapidly over the course of a few hundred milliseconds (H. M. Sigurdardottir & D. L. Sheinberg, unpublished observations). We therefore thought it possible that the perceived direction of a shape could change very rapidly as well and thus we included two different stimulus onset asynchronies in this experiment.
Method
Participants
14 new participants (9 women) completed experiment 2. They were between 18 and 31 years of age (M=23).
Stimuli
The 80 shapes used in experiment 1 were also used in experiment 2. We found the median axis of the directional estimates gathered for each shape in experiment 1. Note that the axis itself has an orientation but not a direction; for example, directional judgments to the left and right would similarly favor a horizontal axis, and up and down directional judgments would count toward a vertical axis. All shapes were then rotated so that this main axis fell on the horizontal meridian. Clockwise or anti-clockwise rotation was chosen for each shape, whichever one led to a rotation of fewer degrees from the shape’s orientation in experiment 1. Each shape was shown in this alignment (original) or reflected across the vertical meridian (mirrored).
Design
Each person completed 320 trials spread across five experimental blocks in a random order. The 80 shapes were shown four times each, twice in the original alignment and twice mirrored to ensure that any possible left-right biases would not systematically influence people’s directional judgments. Each shape was followed by two peripheral disks with a 150 ms or 300 ms stimulus onset asynchrony (80 shapes × 2 alignments × 2 SOAs).
Procedure
The behavioral task from experiment 2 can be seen in figure 3. People were asked to look to the center of the screen at the start of each trial. A white shape (diameter approximately 4°) appeared on a black background in the center of the screen followed by the onset of two gray disks (diameter 2°), one on the left and the other on the right side of the screen (8° eccentricity). The shape and the disks stayed on the screen until the person responded.
Figure 3. Task in experiments 2 and 3.
In experiment 2, a central shape was shown, followed by the appearance of two peripheral disks. Participants indicated whether the shape pointed or directed them to the left or right disk by pressing one of two response buttons. These judgments were used as indicators of each shape’s directionality. In experiment 3, a shape was followed by the display of only a single disk which could appear with equal likelihood in a direction congruent or incongruent with the shape’s inherent directionality. Subjects indicated which side of the screen the disk appeared on by pressing the appropriate button. An example congruent disk is marked here with a white circle; no white circles were shown in the actual experiments.
Participants held a button box with both hands. They were instructed to push the left button if they thought a shape pointed or directed them to the left dot, and push the right button if they thought a shape pointed or directed them to the right dot. They were told to respond as soon as the dots appeared and were informed that there were no correct or incorrect responses for any of the shapes. Each person completed a short practice session with separate shapes.
Results
For each stimulus onset asynchrony (150 ms and 300 ms), we calculated a measure of a shape’s directionality. We did so by determining whether a shape and its mirror image were reported to have opposite directionality. If they did, the participant was said to have determined that a shape had a particular directionality which we arbitrarily call positive and negative (positive: original shape pointed left, mirror image right; negative: vice versa). We then counted the number of participants that indicated that a particular shape had a positive directionality, subtracted the number of participants that reported that the shape had a negative directionality, and divided the difference by the total number of participants. This measure of directionality can theoretically range from -1 (all participants indicate a negative directionality) to 1 (all participants indicate a positive directionality).
As can be seen in figure 4, there is a high correlation between the directionality of the shapes at the two stimulus onset asynchronies (r(78)=0.91, p=3.9×10−32) and the regression line passes through the origin (y-intercept is not statistically different from 0, p=0.475). Therefore, a shape’s directionality appears to be unaffected by the time of probing. The high correlation between the two measures indicates that they are capturing the same construct (i.e. directionality) with some added noise. We therefore combined the measures by taking their average. The measure’s sign was used in all following experiments as a binary statistic indicating each shape’s directionality, i.e. was a shape deemed to be leftward or rightward in its original position? The measure’s absolute value was used in a cross-experiments analysis as an indicator of directionality strength or consensus (see Individual Item Analysis). A two-factor ANOVA with directionality strength as a dependent measure did not reveal any significant effects of either the complexity of the shapes (whether their polygons were made by fitting a spline to 8 or 16 coordinates) or whether they were symmetric or asymmetric (main effect of complexity: F(1,76) = 2.860, p = 0.095; main effect of symmetry: F(1,76) = 1.730, p = 0.192; interaction: F(1,76) = 0.022, p = 0.882).
Figure 4. Directionality of novel shapes.
Each novel shape’s directionality is indicated by a marker. Symmetrical shapes are marked with a light gray diamond, and asymmetrical shapes with a dark gray circle. A diamond embedded in a circle is a marker for two different shapes, one symmetrical and the other asymmetrical. A circle marked with a dot represents two asymmetrical shapes. Directionality can theoretically range from -1 (everyone judges a shape to have negative directionality) to 1 (everyone judges a shape to have positive directionality). Directionality judgments were highly similar at two stimulus onset asynchronies. For details, see Experiment 2: Forced Choice of Directionality.
Experiment 3: Shape-Induced Covert Visual Orienting of Attention
Our claim is that the capability of biasing orienting is a general property of shape, even without explicit training or learning, instead of being limited to a select few over-learned objects. In experiment 3, we therefore used several novel shapes and asked if they automatically pushed visual attention in a particular direction. We would reach this conclusion if people were faster at detecting visual targets when novel shapes pointed to their location even though targets were no more likely to appear there than they were to appear in the opposite direction. We additionally wanted to see whether these effects were time-sensitive. We expected the shape of the objects to rapidly and automatically lead to the formation of an initial hypothesis of where to pay attention, soon to be rejected because relying on the shapes’ directionality was maladaptive for performance in the task. We therefore expected a rapid waxing and waning of the effects of shape-derived directionality on the allocation of spatial attention akin to the time course of transient visual attention (see e.g. Nakayama & Mackeben, 1989).
Method
Participants
Subjects were 20 people (12 women), ages 18 to 28 (M=21).
Stimuli
40 out of 80 shapes used in experiment 2 were used as stimuli in experiment 3. The shapes with the strongest directionality, as determined by responses in experiment 2, were used with the constraints that the proportions of shape types (symmetric or asymmetric, generated from 8 or 16 coordinate polygons) were the same as in the original shape set. Stimuli were displayed as described for experiment 2.
Design
Each person completed 960 trials in a random order. The trials were spread over 10 blocks and were completed in a single day. Each shape was shown equally often in its original alignment and mirrored (see experiment 2). A disk target followed the shape onset with a stimulus onset asynchrony (SOA) of 0 ms, 50 ms, 100 ms, 150 ms, 300 ms, or 500 ms. The design was fully crossed (40 shapes × 2 polarities × 2 disk locations × 6 SOAs) so the shapes predicted neither where nor when a target would appear.
Procedure
Eye position was monitored with an EyeLink 1000 eye tracker (SR Research). Participants had to maintain fixation within 0.65 degrees of the center throughout each experimental trial, otherwise it would abort. Participants held a response button box with both hands. A shape was displayed in the center. A single gray disk target (diameter 2°) appeared with a variable time delay on the horizontal meridian, either on the left or the right side of the screen at an eccentricity of 8°. Shapes did not predict either where or when a target would appear. The task is depicted in figure 3.
People were instructed to press the left button if this target appeared on the left and press the right button if it appeared on the right. They were asked to do this as fast as they could while keeping their responses nearly 100% correct. The task procedure can be seen in figure 3. Before data collection began, participants completed a short practice session with a circular shape.
It should be noted that data from a secondary task were collected from participants in experiment 3. This secondary task was a replication of experiment 2 except that the stimulus onset asynchrony (SOA) was fixed at 150 ms. Before their main session (procedure described above), participants judged the directionality of those 40 out of the original 80 shapes that were not used as stimuli in experiment 3. After their main session, participants judged the directionality of the 40 remaining shapes that were used as stimuli in experiment 3. The data from the secondary task were not used since responses in experiment 3 could be sufficiently predicted based on data collected from an independent group of people who participated in experiment 2, as described in the Results below.
Results
Overall accuracy ranged from 92–100%. Accuracy was slightly, but significantly, greater on congruent (M=98.2%) than incongruent trials (M=97.7%, paired samples t-test, t(19) = 2.301, p=0.033, d=0.51). A trial was considered congruent if a central shape pointed in the direction of a peripheral target, as determined by an independent sample of participants in experiment 2, and incongruent if the shape pointed in the opposite direction. We looked at effects on response times for correct trials only.
13 people completed all 960 trials with full eye tracking. Seven people either did not complete all trials, or completed all trials but we were unable to track their eyes for the whole duration of the experiment. The results for these two groups were qualitatively similar, and similar conclusions would be drawn from statistical analysis on their data. We therefore included data from all subjects in an ANOVA with response time as a dependent measure and two repeated factors, congruency and stimulus onset asynchrony (SOA, the time between the onset of the shape and the target). Response time was considered to be the time between target onset and manual response.
People were significantly faster when the shapes’ directionality was congruent with the target location (figure 6, F(1,19)=22.159, p=1.5×10−4, ηp2=0.54). The mean response time also decreased as more time passed between the onset of the shape and the target (F(1.76,33.40) = 75.91, p=4.5×10−13, ηp2=0.80). The interaction between congruency and stimulus onset asynchrony was only marginally significant (F(5,95)=1.967, p=0.090). Joint tests of the effects of congruency within each level of stimulus onset asynchrony showed that 50 ms was the earliest SOA at which congruency had a significant effect on response time (F(1, 95)=21.32, p=1.5×10−5, Bonferroni corrected threshold for significance: 0.008, d=0.99).
Figure 6. Shape-induced orienting of attention.
Mean response times (RT) from experiment 3 are shown as a function of stimulus onset asynchrony (SOA) and whether the location of a target was congruent (light gray) or incongruent (dark gray) with the inherent directionality of a non-predictive central shape cue.
All participants in experiment 3 were right handed. It is conceivable that congruency effects were mainly driven by trials when the target was on the right and the participants thus responded with their dominant hand. Using only correct trials, we therefore performed another ANOVA with response time as a dependent measure and three repeated factors: Congruency, stimulus onset asynchrony (SOA), and target position (on the left or right). The main effects of congruency (F(1,19)=21.25, p=1.9×10−4, ηp2=0.53), SOA (F(1.71,32.44)=76.60, p=4.0×10−13, ηp2=0.80), and target position (F(1,19)=5.26, p=0.033, ηp2=0.22) were all significant, as was the interaction between SOA and target location (F(3.10,58.89)=6.44, p=3.4×10−5, ηp2=0.25). People were faster on congruent trials, they got faster as SOA increased, they were faster for left than for right targets, and this difference for left and right targets decreased with longer SOAs. There was, however, no significant interaction between congruency and target location (F(1,19)=0.04, p=0.85), nor a significant three-way interaction of congruency, SOA, and target location (F(5,95)=1.03, p=0.41). The congruency effect does therefore not appear to depend on the target’s position or the hand used to report it.
Interestingly, there was enough variability explained by target location that when it was included as a factor in the ANOVA, a significant interaction between congruency and SOA was revealed (F(5,95)=2.41, p=0.042, ηp2=0.11). The dependency of the congruency effect on SOA was close to but not exactly as expected. We had hypothesized that the congruency effects would have a sharp monotonous increase followed by a decrease. Instead, the congruency effects appeared to peak twice, once at the 50 ms SOA and again at the 150 ms SOA. Although surprising, two peaks at approximately those same time points have been reported before for transient visual attention (Nakayama & Mackeben, 1989, see e.g. their fig. 7). We will leave it to future studies to find out whether there might be two processes underlying the effects we see here.
Figure 7. Task in experiment 4.
A central shape validly cued the location of a peripheral target cross shown among distractor plus signs. Participants had to find the cross and report its central color. While all shapes provided accurate information about the upcoming target location, the cued location was congruent with the directionality of half of the shapes but incongruent for the other half. Examples of congruent and incongruent search trials are shown. Cued locations are indicated by yellow dotted circles. The locations pointed to and away from are indicated by blue and red dotted circles, respectively. No dotted circles were actually shown to the participants.
To summarize, people in general are both faster and more accurate at detecting a single target if its location is congruent with the directionality of a non-predictive central shape cue. The congruency effects vary with stimulus onset asynchrony, and are apparent very early on, as early as 50 ms after visual onset of a shape.
Experiment 4: Overcoming Shape-Induced Biases
Experiment 3 showed that the shape of an object rapidly and automatically pushes covert attention in a particular direction. How easily can this bias be overcome? Experiment 3 was deliberately set up to have no cue-target contingencies, making the shape useless with regard to the subjects’ detection task. In experiment 4, all shapes provided accurate information about the location of an upcoming target. However, some cue-target contingencies were in accordance with the shape’s directionality while others conflicted with it. Would experience with these cue-target contingencies make people overcome their initial shape-induced biases?
We designed a task where a target always appeared in the location to which some shapes pointed, while for other shapes it always appeared in the location that they pointed away from. If people are consistently faster at finding the target in the former case than in the latter, even though all shape cues are informative, we would conclude that a shape’s directionality influences behavior not only in a situation when there is nothing else to go on, but also comes into play even when other, more accurate information is available.
Method
Participants
16 people (6 women) between the ages of 18 and 30 (M=22) participated.
Stimuli
Eight simple asymmetric shapes were used as central precues in a visual search task. The shapes were black, had the same area, and had an approximate diameter of 3°. The shapes’ directionality had previously been determined in experiment 2.
People searched for a target cross among distractor plus signs. Distractors were made by overlaying a vertical and a horizontal bar (1.1° × 0.3° each). The target was made in the same way except that one bar was vertically displaced by 0.2°. The search stimuli were then given a random rotation on each trial. The search stimuli were black, except that a small colored circle (diameter 0.1°) was embedded in each of them. The target’s circle color could be red or green and was chosen at random. The color of each distractor’s circle was also randomly determined to be red or green with the constraint that there was at least one distractor disk of each color.
Design
Two sets of four shapes were used in this experiment. Half of the participants were given one set and half were given the other set. Each shape served as a predictive central precue in a visual search task. It cued one of four possible target locations (upper left, upper right, bottom left, or bottom right).
Two shapes were congruent, meaning that the shapes’ inherent directionality was consistent with the direction of the target location that it cued. The other two were incongruent; they cued a target location in a direction opposite that of their inherent directionality. The two congruent shapes cued target locations on one diagonal and the two incongruent shapes cued target locations on the other diagonal (figure 7). The rotation of each shape was the same across all participants with the same shape set but the cue-target contingencies differed; each shape served as a congruent cue for four participants and as an incongruent cue for another four participants.
Central shape cues therefore predicted, with 100% accuracy, where peripheral targets would appear. The correctly predicted location could be congruent or incongruent with the shape’s inherent directionality. Each participant completed 240 search trials spread over four blocks during a single session.
Procedure
Eye position was monitored using an EyeLink 1000 eye tracker (SR Research). The participant’s gaze on a central 0.3° × 0.3° fixation square triggered the start of each search trial; she was then free to move her eyes for the remainder of the trial. The fixation square was replaced by a predictive central shape cue which was visible throughout the trial. The participant was told that a shape would appear on the screen after she had acquired fixation, and that after the shape appeared a search array would show up on the screen. After 500 ms, a square search array with three distractors and one target appeared around the central shape. The search stimuli were all shown at 11° eccentricity. The participant had to find the target and report the color (red or green) of an embedded disk by pushing the button of the corresponding color on her response box. This completely disambiguated the manual response from the directionality of the shape. The procedure is depicted in figure 7.
Participants were instructed to respond quickly but to try to maintain near perfect performance. Auditory feedback was given to indicate whether a response was correct or incorrect.
Results
Mean accuracy ranged from 92% to 98%. People were significantly more accurate at judging the color of a disk embedded in a target if the target was preceded by a shape cue whose directionality was congruent (M=96.2%) rather than incongruent (M=94.8%) with the target’s location (paired samples t-test, t(15)=3.257, p=0.005, d=0.81). Error trials were not further analyzed.
Mean response time was used as a dependent measure in a repeated measures ANOVA with block (1–4) and congruency as factors. The main effects of block (F(3,45)=9.57, p=5.3×10−5, ηp2=0.39) and congruency (F(1,15)=15.05, p=0.001, ηp2=0.50) were significant, but the interaction between the two factors was not significant (F(3,45)=0.98, p=0.409). Overall, response times decreased over the course of the experiment. Subjects were also faster at reporting the attributes of a peripheral target when it was in a location congruent with a central shape cue’s directionality. This effect did not seem to diminish over the course of the experiment.
Because congruent and incongruent shapes were equally predictive of where a target would appear, one might have expected that the performance gap between congruent and incongruent shapes would narrow as people gained more experience with the cue-target contingencies. Although this might potentially happen with longer training, we saw no sign of it and the benefit for congruent shape cues persisted. It thus appears that people intuitively make certain associations more easily than others and that this preference is not easily erased in a single session.
Experiment 5: Shape as a Movement Cue
So far we have shown that the shape of an object is used to rapidly and automatically extract its directionality, and that this in turn guides both overt and covert visual orienting. An unanswered question is why the visual system is set up this way at all. One possible reason is that the shape of an object restricts and thus predicts its movements. A snapshot of the shape of an object might therefore provide valuable information about where it may be moments later. The rules governing selective sampling of the environment should incorporate any available data, including shape, which provides prior information about where important things are going to be in the near future. Informal self-reports of participants in experiments 1 and 2 also indicated that judgments about the directionality of shapes could be related to people’s perceptions about where the things were moving or heading. In experiment 5, we directly examined whether the shape-defined directionality of an object was integrated into calculations about its movement. We would reach this conclusion if people were consistently faster at judging where an object was heading if its direction of motion was congruent with the directionality derived from the object’s shape.
Method
Participants
16 people (7 women) of ages 18–54 (M=27) participated in the experiment.
Stimuli
Stimuli were the same 40 shapes used in experiment 3. The shapes were white and shown on a black background. Each shape extended approximately 1°.
Design
Each person completed 320 experimental trials in two blocks within a single session. All shapes were shown four times within each block in a random order (40 shapes × 2 shape directionalities × 2 movement directions × 2 repetitions).
Procedure
The participant was instructed to look at a fixation disk (white 0.5° diameter) at the beginning of each trial. She was otherwise free to move her eyes. The fixation disk stayed onscreen for 510 ms, and 470 ms later, participants then saw multiple copies of a particular shape lined up in a row across the screen (figure 9). The screen center coincided with the pivot point of the central shape (see experiment 1). The distance between corresponding points of juxtaposed copies of the shape was 2.4°. To create a moving stimulus, the row of shapes was translated 0.8° to either the left or right every 130 ms. On any given trial, the row of shapes therefore appeared to be moving either leftward or rightward.
Figure 9. Example motion stimuli.
To create a moving stimulus in experiments 5, 6, and 7, a row of shapes was translated in a direction that was either congruent or incongruent with the shapes’ inherent directionality. The direction to which the example shapes are most often judged to point is indicated by a black vertical bar, and the opposite direction is indicated by a gray vertical bar.
Shapes were shown equally often pointing to the left or the right; this directionality was defined by an independent sample of people (see experiment 2). The shapes pointed in the direction of motion on half of the trials, and pointed the opposite way on half of the trials. Shape was not a valid predictor of motion.
Participants held a response button box with both hands and were told to press the left button if the shapes were moving to the left, and press the right button if they were moving to the right. A tone sounded when the participant responded. No specific feedback was provided about whether the answer was correct or incorrect.
Results
Mean accuracy ranged from 89–99%. Although accuracy was generally very high, people were significantly more accurate at judging where the shapes were going when the shapes pointed in the direction to which they were moving (congruent: M=97.9%, incongruent: M=95.0%, paired samples t-test, t(15)=4.408, p=0.001, d=1.10).
We calculated the mean response times for correct trials only. Response times were defined with respect to motion onset, which was the time of the first translation of the multi-shape stimulus. All participants were faster at judging where the shapes were going if their movement direction was congruent with their inherent directionality (figure 10). This effect was significant (congruent: M=317 ms, incongruent: M=353 ms, paired samples t-test, t(15)=9.746, p=7.0×10−8, d=2.44).
Figure 10. Shape as a movement cue.
Mean response times (RT) are shown as a function of congruency between shape-derived directionality and direction of motion. When an object moved in the direction to which it pointed, people were faster at judging its direction of motion (Experiment 5: Motion direction, course dotted line with squares), matching its motion to the motion of other objects (Experiment 6: Motion matching, solid line with triangles), and following it with their eyes (experiment 7: Ocular pursuit, fine dotted line with circles). Note that confidence intervals are small.
Experiment 6: Match to Motion
The results from experiment 5 were quite robust; every participant was faster at judging where a shape was going if it pointed in the direction of motion. We interpret this as evidence for the idea that the shape of an object, in particular its shape-derived directionality, is automatically integrated into movement calculations.
In experiment 6, we wanted to address two alternative interpretations. First, we wanted to rule out the possibility that any slight pixel-by-pixel differences between leftward and rightward shapes solely determined an object’s supposed directionality and its behavioral effects. Secondly, it is possible that we were not seeing an effect on motion perception but rather a type of effector priming; certain shape features might afford being grasped by a particular hand, and a button press with that hand might thus become potentiated. The stimuli in experiment 5 were all very small 2D silhouettes that, if perceived as graspable at all, probably all afforded a similar pincer grip; nonetheless, we wanted to rule out this explanation.
To address these possibilities, we designed an experiment where moving shapes had a random starting position, and where manual responses were neither directly related to the direction of motion nor to the directionality of shapes. If people are still faster at judging the direction of motion of an object when it is congruent with the directionality derived from its shape, we would conclude that these alternative interpretations do not sufficiently account for our effects and that, instead, shape-derived directionality is integrated into the calculations of an object’s motion path.
Method
Participants
16 people (8 women) of ages 18 to 34 (M=22) took part in this experiment. One additional participant was excluded because of very low accuracy rate (more than six standard deviations below the mean).
Stimuli
Shape stimuli were as described for Experiment 5 with the addition of a white disk shape (diameter 0.9°).
Design
Participants completed 320 trials each in four blocks within one session. Trials were shown in a pseudo-random order. The design was fully crossed (40 shapes × 2 polarities × 2 shape movement directions × 2 disk movement directions).
Procedure
Procedure was as described for experiment 5 with the following changes. Presentation of a central fixation spot was followed by the appearance of several disk shapes that extended to the screen’s edges. The horizontal starting position of the disks was random but the distance between the centers of adjacent disks was fixed at 2.4°. All disk shapes were translated 0.8° degrees every 130 ms so that they appeared to move either leftward or rightward. The disks disappeared 390 ms after their initial onset. After a 500 ms inter-stimulus interval, participants saw multiple copies of a particular novel shape. Their horizontal starting position was random but the distance between corresponding points on two adjacent shapes was always 2.4°. The shapes could point leftward or rightward, and could move leftward or rightward, as detailed in experiment 5 (see also figure 9).
Participants indicated whether each novel shape was moving in the same direction as the disks (match) or in a direction opposite that of the disks (non-match). Participants responded with their right hand using a two-button box. The button box was aligned so that one button was nearer the person and the other was farther away. Half of the participants pushed the closer button to indicate a match and the button farther away to indicate a non-match, and vice versa for the other half of the participants. Participants completed some practice trials with other shapes randomly picked from the rest of the original shape dataset used in experiment 2.
Results
Mean accuracy ranged from 84–100%. Participants were on average more accurate on trials where the shape’s directionality was congruent with the shape’s own direction of motion; this difference did not reach statistical significance (congruent: M=95.6%, incongruent: M=94.1%, paired samples t-test, t(15)=1.676, p=0.114). Error trials were not analyzed further.
Response times were defined as the time between the novel objects’ motion onset and button press. People were significantly faster when novel shapes pointed in the direction to which they were moving (congruent: M=542 ms; incongruent: M=567 ms; paired samples t-test, t(15)=7.244, p=3×10−6, d=1.81, figure 10).
We regressed the objects’ starting position against response time. By starting position we refer to the location of the pivot point (see experiment 1) of the central shape in the first frame relative to the direction of motion; for example, if the pivot is 1° to the right of the screen center, but the shape is moving leftward, then the shape’s starting position is considered to be −1° relative to the motion direction. For each participant, we calculated the slope of the best fitting line (least squares method) and did so separately for congruent and incongruent trials. The participants’ mean slopes for congruent (M=0 ms) and incongruent trials (M=7 ms) were neither significantly different from zero (single-sample t-test; congruent trials: t(15)=0.116, p=0.909; incongruent trials: t(15)=1.561, p=0.139), nor were they significantly different from each other (paired-samples t-test; t(15)=0.977, p=0.344). Starting position was not found to be a significant factor contributing to response times in this task.
People are therefore faster at judging the direction of movement of an object if its shape is congruent with the object’s motion path. This cannot be attributed solely to pixel-by-pixel differences between leftward- and rightward-pointing objects because their starting position was randomly varied. The effect cannot be attributed to effector priming either; the effect was found even though people used one hand only and the button presses were orthogonal to the objects’ direction of motion and shape-derived directionality.
Experiment 7: Shape Effects in Oculomotor Programming
Experiments 5 and 6 showed that shape can play a significant role in motion perception. However, we are especially interested in the contribution that shape information can make to action guidance, in particular oculomotor guidance, considering that shape selectivity has been found in important oculomotor centers of the brain (Janssen et al., 2008; Konen & Kastner, 2008; Lehky & Sereno, 2007; Peng, Sereno, Silva, Lehky, & Sereno, 2008; A. B. Sereno & Maunsell, 1998; M. E. Sereno et al., 2002). Given the numerous dissociations between perception and action (Goodale, 2008; Goodale & Milner, 1992; Milner & Goodale, 1995, 2010), including oculomotor behavior (Mack, Fendrich, Chambers, & Heuer, 1985; Spering & Gegenfurtner, 2008; Spering & Montagnini, 2011; Wong & Mack, 1981), we thought it important to test whether shape affects the programming of eye movements in addition to perception. Additionally, we wished to compare the effects of novel random shapes with those of arrows, which are both familiar and highly directional, and with filled circles, which should be adirectional, to see whether shape-derived directionality mainly helps or hinders performance relative to situations in which no bias should be present. People were asked to follow a row of moving shapes with their eyes. They were free to use both saccadic and smooth eye movements for this ocular pursuit task. We expected eye movements in the direction to which the shapes were pointing to be facilitated, and eye movements in the opposite direction to be hindered.
Method
Participants
16 people participated (7 women). Their age ranged from 18 to 24 (M=21).
Stimuli
We used 40 novel shapes, as described for experiment 5, and four additional shapes: three differently shaped arrows and a filled circle. All shapes, including the arrows and the circle, had the same area and an approximate diameter of 1°.
Design
Each person completed two experimental blocks for a total of 184 trials in random order. All directional shapes (40 novel shapes, 3 arrows) were shown 4 times each (2 motion directions × 2 shape directionalities). Circular shapes were shown on control trials (2 motion directions × 3 repetitions).
Procedure
The experimental procedure was as described for experiment 5 (see also figure 9) with the following changes. At the beginning of each trial, the screen center always coincided with the center of area of the central shape in each multi-shape stimulus. People were told to follow the shapes’ movement (leftward or rightward) with their eyes. Eye position was tracked; a trial started once a person had acquired fixation on a central fixation spot. Instead of responding to the direction of motion with a button press, the trial ended once people’s eyes reached one of two invisible circular regions, a target region in the direction of motion (correct response) or a distractor region in a direction opposite that of the real motion of the shapes (incorrect response). The circular regions were centered on the horizontal meridian at 6.0° eccentricity with a radius of 3.0°. Trials were considered valid if the subjects’ horizontal eye position within the first 130 ms after stimulus onset was no further than 0.65° from the screen’s center, and vertical eye position was no further than 1.0° from the horizontal meridian throughout the trial. Furthermore, trials were considered valid only if people reached one of the circular regions within 2000 ms of motion onset. On average, 79.2% of trials were deemed valid and we base our analysis on these valid trials only.
Results
Participant’s mean accuracy for novel shapes ranged from 62% to 100%. This great range of performance was surprising since the task was mainly designed to measure response time and not accuracy levels. Accordingly, here we saw a much greater difference between the accuracy in congruent (M=92.9%) and incongruent (M=84.2%) novel shape trials than in our previous experiments where accuracy was closer to ceiling. People were significantly more accurate on congruent than incongruent novel shape trials (paired samples t-test, t(15)=3.865, p=0.002, d=0.97). Response times were calculated relative to motion onset on correct trials only. People were significantly faster at reaching the target region, located in the direction of motion, if the novel shapes’ inherent directionality was congruent with the direction of motion (congruent: M=267 ms, incongruent: M=291 ms, paired samples t-test, t(15)=5.719, p=4.1×10−5, d=1.43, figure 10). We note that the effect of congruency on both accuracy and response time remains significant even when invalid trials are included in the analysis.
We compared the effects of novel shapes with the effects of arrows. People were far more accurate when the arrows pointed in the direction of motion than if they pointed in the opposite direction (congruent: M=100.0%, incongruent: M=53.0%, paired samples t-test, t(15)=9.918, p=5.6×10−8, d=2.48) and almost twice as fast (congruent: M=240 ms, incongruent: M=405 ms, paired samples t-test, t(14)=5.231, p=1.3×10−4, d=1.35; one person had no correct incongruent trials and was therefore not included in the RT measures). People were also significantly faster and more accurate for congruent arrows than they were for congruent novel shapes, and they were significantly slower and less accurate for incongruent arrows than they were for incongruent novel shapes (paired samples t-tests, all ps<0.003, all ds > 0.96). As expected, arrows are therefore particularly effective stimuli for orienting guidance.
Finally, we compared novel shapes to filled circles (which have no directionality). The mean accuracy (M=88.9%) and response times (278 ms) for circles fell half-way in between those of congruent and incongruent novel shapes. The differences in accuracy for circles and congruent shapes were not reliably smaller or larger than the differences in accuracy for circles and incongruent shapes (t(15)=0.221, p=0.828). Response time differences for circles and congruent shapes were not reliably smaller or larger than those for circles and incongruent shapes (t(15)=0.129, p=0.899). The effects of directionality therefore appear to be more or less symmetrical; the more congruent a shape’s directionality is with the direction of motion, the faster and more accurate the oculomotor behavior, and the more incongruent a shape’s directionality, the slower and more error-prone is the behavior. Shape-derived directionality appears to be a strong enough motion cue that the stimuli can be perceived to move, and are thus initially pursued, in the direction opposite that of the “real” motion. Overall, our results support the hypothesis that novel shapes have an automatic effect on oculomotor programming.
Individual Item Analysis
The possibility remained that our results were driven only by a few atypical novel shapes, with the rest of them contributing nothing to the effects. For example, it was possible that by random chance, a few of our shapes looked like arrows and that these atypical shapes were the sole driving force behind our results. To rule out this possibility, we analyzed congruency effects for individual shapes. We did so with data collected for novel shapes in experiment 3 (detection), experiment 5 (motion direction), experiment 6 (motion matching), and experiment 7 (ocular pursuit). For experiments 3, 5, and 6, we calculated the mean response time on incongruent and congruent trials for each shape for each participant, calculated response time savings by subtracting the former from the latter, and finally found the mean response time savings for each shape across participants within a particular experiment. We included all trials regardless of people’s responses to get adequate sampling of responses to each shape. Data from one participant in experiment 3 was excluded because she did not complete both congruent and incongruent trials for all of the shapes. For experiment 7, most subjects had at least one shape with either no valid congruent trials or no valid incongruent trials; we therefore collapsed across subjects and calculated mean response times savings for each novel shape. Collapsing across subjects allowed us to include only correct trials for response time calculations and still retain enough trials for each of the novel shapes. Accuracy had a much greater range in experiment 7 than in our other experiments, providing us with the opportunity of also looking at accuracy savings found by subtracting the percent of correct incongruent trials from the percent of correct congruent trials for each novel shape.
Results
The response time savings for the 40 novel shapes were positively correlated across all four tasks (figure 11). Accuracy savings in the ocular pursuit task (experiment 7) were also positively correlated with response time savings from all four tasks (figure 11). Assuming that any one measure is a somewhat noisy estimate of the same construct, that is to say the strength of a shape-derived directional bias, we combined all five measures into a single measure of an overall congruency advantage. We did so by dividing each original savings measure by its standard deviation and then took the average for each shape across the five scaled measures.
Figure 11. Congruency effect correlation matrix.
This matrix shows the relationship between directionality strength (experiment E2) and savings measures from various tasks (experiments E3, E5, E6, and E7, summarized in Individual Item Analysis). Numbers indicate Pearson’s r. Ellipses are the contours of a bivariate normal distribution with a correlation r (Murdoch & Chow, 1996).
The five original measures of savings were positively correlated with directionality strength as defined by the degree of consensus reached on the directionality of shapes in experiment 2 (figure 11). The overall congruency advantage scores were also significantly correlated with directionality strength (r(38)=0.489, p=0.001). As can be seen in figure 12, the behavioral effects were not due to a few outlier shapes; instead they were graded and related to the shapes’ directionality strength. Regressing directionality strength against the congruency measure also revealed that the y-intercept (congruency advantage: −0.260) was not significantly different from 0 (p=0.423), indicating, unsurprisingly but reassuringly, that an adirectional shape would be expected to induce no directional behavioral bias.
Figure 12. Behavioral effects of individual shapes.
Each marker is in the shape of the corresponding novel object shown in experiments E3, E5, E6, and E7. Asymmetric shapes are shown in black, and symmetric shapes are shown in gray. All shapes are shown pointing to the right, as judged by participants in experiment E2. Overall, the stronger the consensus is on a shape’s directionality, the greater the behavioral advantage is on congruent relative to incongruent trials.
The analysis of individual novel shapes shows that the stronger the directionality of a shape, the greater its behavioral biasing effects will, in general, be. This analysis also shows that our results were not driven by few very atypical shapes. Instead, congruency effects were found for a great number of shapes across various tasks. We find it parsimonious to conclude that the effects are not solely explained by resemblance to specialized stimuli such as arrows, but that the visual system instead automatically assigns directionality to many different shapes, and that this drives or biases further visual processing and guides behavior.
General Discussion
We hypothesized that the visual system uses information about shape to swiftly and automatically extract the directionality of virtually any object without explicit training or learning. We explored this idea in several related experiments. A majority of randomly generated novel shapes were reliably judged to have one or more main directions (experiments 1, 2). This inherent shape-derived directionality was found to automatically guide both overt (experiment 3) and covert (experiments 4, 7) visual orienting of attention. The effect was rapid (experiment 3), resistant to experience (experiment 4), and was integrated into the assessment of an object’s movement (experiments 5, 6, 7).
Our results show that an object can rapidly and automatically push attention away from itself due to its shape. This appears to be the rule and not the exception. These biasing effects are likely to be direct instead of coming about through explicit interpretation or semantics; our objects were not symbolic, they were novel and meaningless. These orienting shifts do not need to be explicitly learned or trained. They are not easily overridden or overwritten by experience, persist even when they are not useful, and are found in various tasks and situations.
The fact that our effects arise without any particular training does not necessarily indicate that experience has no role in establishing them in the first place. Indeed, previously a directional and non-spatial visual stimuli such as color patches can start to automatically bias covert (Dodd & Wilson, 2009) and overt (Van der Stigchel, Mills, & Dodd, 2010) visual orienting once they have often been paired with a behaviorally relevant thing or action in a particular direction. The same is true for Arabic numerals where low numbers shift overt and covert visual attention to the left while high numbers shift it to the right (Dehaene, Bossini, & Giraux, 1993; Fischer et al., 2003; Fischer, Warlop, Hill, & Fias, 2004). While there might indeed be a true, spatial mental number line (Dehaene, Izard, Spelke, & Pica, 2008; Zorzi, Priftis, & Umilta, 2002), the associations between directions and these particular shapes are presumably relatively arbitrary and might come about through the cultural tradition of reading from left to right, and thus shifting your eyes and attention in the same direction (Dehaene et al., 1993; Shaki & Fischer, 2008).
The time course of learned, arbitrary visual orienting appears to be relatively slow (Fischer et al., 2003; Van der Stigchel et al., 2010) compared to the rapid effects found for novel shapes in the current study. The shape-induced biases we see arise so early that they are presumably not dependent on recurrent feedback but likely arise from an initial bottom-up sweep of visual information. The difference might be that, unlike color patches or digits, the mapping from shape to space is not arbitrary. Colors are non-spatial and digits do not line up on any obvious spatial dimension; the shape of digits, presumably, changes completely arbitrarily going from 0 to 9. On the other hand, the directionality of a shape might lie on a dimension in a yet unknown multidimensional shape space.
Precisely documenting this shape space is beyond the scope of this paper. After participation in experiments 1 and 2, we nonetheless asked people whether they thought they had used a particular strategy or rule to complete the tasks. We summarize these informal self-reports with the hope that it will help generate hypotheses for future experiments that parametrically vary stimulus properties to address what, exactly, determines the direction of a given shape.
Several different strategies were reported. Often people reported using some geometric properties of the shapes: Direction of a large, long, tapered, or sharp protrusion, overall taper of shape, direction of the average of more than one protrusion, direction opposite a small protrusion and between two cupping protrusions, direction toward the meeting point of two tilted lines. Some reported taking into account a center of mass, like they were weighing the object, or dividing the shape into subparts and going with the direction of the part with the greatest mass or area. Some reported taking into account a general axis or an axis of symmetry. Some reported ignoring small protrusions. Some said that they had trouble judging the directionality of shapes that were blob-like or smoothly curved.
When asked, many noted that at least some of the shapes resembled real things, such as arrows, planes or flowers, but in particular animate things such as bugs, marine life, birds, space aliens, or parts of animate things such as faces, heads, mouths, antennae, tails, legs, hands and fingers. Some reported that they tended to go with the direction in which the shapes appeared to be moving or heading, or where they were facing, especially if the shapes appeared to be biological. Judgments of the shapes’ animacy do nonetheless appear to be unrelated to the strength of their directionality; novel shapes that are deemed to look like some kind of existing or hypothetical creature, animal, or person, are not any more or less likely to have a strong directionality (H. M. Sigurdardottir, M. M. Shnayder, & D. L. Sheinberg, unpublished observations). Finally, some participants just reported that they did whatever felt right and that they were not consciously using any particular strategy.
In short, people seem to use various properties of objects to judge their directionality. Strategies span from taking into account particular features of the object’s parts to using summary statistics of the whole shape to noting body structure and plausible movement patterns. The fact that people report so many different strategies or even no strategy at all suggests that several different form or shape characteristics might all come together to influence the judged directionality of an object, and that people might not necessarily have conscious access to the rules that they use to make such judgments. The algorithm used by the visual system to derive an object’s directionality is therefore currently unknown, and there might be more than one mechanism at work.
We can nonetheless theorize about the mechanisms behind our results. One possibility is that our effects are driven by axis-based shape processing (see e.g. Blum, 1967: Kimia & Appear, 2003: Lin, 1996). There is already some evidence that the visual system can use axis-based shape representations and that this affects perceptual sensitivity within an object (Hung, Carlson, & Connor, 2012; Kimia & Appear, 2003; Kovács, Fehér, & Julesz, 1998; Kovács & Julesz, 1994). In figure 13, we have included an example shape and one scenario of how a shape’s axis might affect target detectability outside its boundaries. In this example, a shape’s topological skeleton is found by gradual erosion of the object’s boundaries without breaking it apart (in this case using the bwmorph function of MATLAB’s Image Processing Toolbox). The skeleton is then pruned by cutting off its smallest branches; in computer vision, regularization of a skeleton is commonly applied to reduce noise because small changes in the boundary of an object can lead to great changes in its skeleton (Shaked & Bruckstein, 1998). The visual system might explicitly assign a direction of flow along an axis segment as supposed by some axis models such as the shock map (Kimia & Appear, 2003). Through extension of the axes of the skeleton, perhaps through rules similar to those hypothesized to support collinear facilitation or contour completion (for a review, see Loffler, 2008), it is also possible that the object is grouped more strongly with targets on one side than another. For example, the association field model assumes that contours are formed by the linking of information across neighboring neural receptive fields tuned to similar orientations (Field, Hayes, & Hess, 1993; Ledgeway, Hess, & Geisler, 2005). The fact that directionality affects the perception of motion is at least consistent with the role of collinear facilitation since it may subserve not only contour formation but appears to influence motion perception as well; the speed of collinear sequences is overestimated (Seriès, Georges, Lorenceau, & Frégnac, 2002) and a vertical line moved horizontally toward a stationary horizontal line can be misinterpreted as the movement of the latter line since it is parallel to the direction of motion (Metzger, Spillmann, Lehar, Stromeyer, & Wertheimer, 2006). Real-world objects can be viewed as spatiotemporal events, and their motion can be thought of as a change in the objects’ boundaries over both space and time. It might therefore be expected that mechanisms which support boundary completion in space might also be involved in boundary completion over time, where the shape of an object’s current boundaries is used to predict its future state.
Figure 13. Hypothetical example of how a shape could affect target detectability.
a) A shape’s topological skeleton is found and pruned. Through extension of the axes of the skeleton, perhaps through rules similar to those thought to support collinear facilitation or contour completion, it is possible that the shape is grouped more strongly with targets on one side than another (here, more with a right than a left target). b) This grouping might be stronger for collinear targets (top) than non-collinear targets (middle and bottom).
If mechanisms such as those underlying collinear facilitation and contour integration are involved, then a number of predictions can be made (although there is some disagreement on the relation between collinear facilitation and contour integration, see Loffler, 2008; Williams & Hess, 1998). First, the biasing effects of a shape might be expected to interact with the qualities of the target. The greatest facilitation would be expected for targets that are collinear with the directionality of the shape, and little facilitation would be expected for targets orthogonal to or tilted relative to the shape’s directionality (Polat & Bonneh, 2000; Polat & Sagi, 1994). Second, there might be no congruency effects when an orthogonal distractor is placed between a target and a shape (Dresp, 1993). Third, because the detectability of a contour increases with the number of elements making up a path (Braun, 1999), a “daisy chain” of shapes could induce stronger congruency effects than a single shape; this is one possible reason why the effects in our motion paradigms (experiments 5–7) seemed particularly robust. Fourth, the congruency effect would be expected to change in magnitude and even sign with the relative distance between the shape and the target (Polat & Sagi, 1993, 1994). Fifth, the congruency effect should reach its peak at a later time point with increased distance between the shape and the target (Cass & Spehar, 2005). It would in general be very interesting to document further how shape-induced biases develop in both space and time, where target detectability would be probed not just at several different time points but at various distances and directions from a shape.
In addition to, or instead of, the mechanisms discussed above, the rules linking shape and space might be more explicitly derived from the complex but non-random way in which the shape of an object restricts its movements and therefore its probable future location. Our stimuli were two-dimensional silhouettes, but real objects exist and move in a fully three-dimensional world. If an object is assigned a directionality for the purpose of predicting its future location, then real-world objects might have a directionality defined in not just two but three dimensions. If all other things are equal, an object is likely to move in a path of least resistance to air flow. Preliminary work from our laboratory suggests that directional judgments might be related to a shape’s aerodynamic properties. The greater the consensus reached on the directionality of a shape, the better its path of least resistance approximated the shape’s empirically defined directionality (S. Boger and S. M. Michalak, unpublished observations). Further work on the role of aerodynamics is warranted. The current results show that the visual system is able to link the appearance of an object with its possible path of motion. Directional information derived from shape can be used to guide the eyes and attention to the object’s future location so that it can be tracked, examined, and acted on.
Our experiments were based on the hypothesis that the shape of an object affects the weights given to locations in a spatial priority map (Bisley & Goldberg, 2010; Fecteau & Munoz, 2006; Itti & Koch, 2001). Overt and covert visual attention would be guided to the location of peak activity within the map once activity reaches a particular threshold, and this attentional orienting signal would in turn bias other processes such as motion perception (Cavanagh, 1992; Stelmach, Herdman, & McNeil, 1994; Treue & Maunsell, 1996). Such a tight link between shape, attentional priority, and motion perception is biologically plausible; posterior parietal regions which play an important role in target selection and visual orienting (Andersen, Snyder, Batista, Buneo, & Cohen, 1998; Arcizet, Mirpour, & Bisley, 2011; Bisley & Goldberg, 2010; Colby & Goldberg, 1999; Gottlieb, Kusunoki, & Goldberg, 1998; Silver, Ress, & Heeger, 2005) are furthermore selective for the shape of objects (Janssen et al., 2008; Konen & Kastner, 2008; Lehky & Sereno, 2007; Red et al., 2012; A. B. Sereno & Amador, 2006; A. B. Sereno & Maunsell, 1998) and their activity is predictive of the perceived motion direction of ambiguous motion stimuli, even to a greater extent than activity within the classical motion regions MT and MST (Williams, Elfar, Eskandar, Toth, & Assad, 2003). The behavioral experiments reported here were also directly prompted by our own electrophysiological work where we recorded activity of single neurons within these posterior parietal regions. This line of research showed that rapid and automatic neural responses to novel visually presented shapes, responses which previously had no known function, could be directly tied to the allocation of spatial attention and eye movements (H. M. Sigurdardottir & D. L. Sheinberg, unpublished observations).
There are, however, other possibilities. For example, motion processing might have a primary role, where shape directly affects the calculation of motion and overt and covert attention is then guided in the direction of movement. Also, if the shape and the target are grouped into one perceptual whole, then the effects reported here might not strictly be considered only spatial and the enhancement for the target to which a shape points could be closely related to object-based attention (Driver & Baylis, 1989; Duncan, 1984; Egly, Driver, & Rafal, 1994). The mechanisms behind the behavioral results reported here need to be further studied.
It is worth noting that earlier attempts to find effects of shape directionality on orienting apparently failed (Zusne & Michels, 1964). Zusne and Michels (1964) did not find evidence for the idea that people would preferentially follow the main direction of a shape with their eyes. The discrepancy between this and the current study could be due to the fact that Zusne and Michels (1964) did not empirically define the shapes’ directionality. Wolfe, Klempe, and Shulman (1999) also failed to find evidence for the hypothesis that varying an object’s polarity, which roughly corresponds to our idea of directionality, led to efficient visual search. They concluded that there is little evidence for the preattentive processing of an object’s polarity. We do not think that our results necessarily contradict those of Wolfe et al. (1999). As these authors themselves acknowledge, it is hard to interpret negative findings. More to the point, we are not claiming that directionality is an attribute that supports efficient visual search, or an almost instantaneous readout (e.g. pop-out) of some particular information. This kind of fast information detection might be fundamentally different from what we are talking about here, which is a stimulus-driven, rapid and seemingly automatic shift in information sampling. An object’s directionality also pushes attention away from the object itself. There is no specific reason why a strongly directional shape should itself be particularly rapidly detected in a visual search.
The affordance competition hypothesis (Cisek, 2007; Cisek & Kalaska, 2010) states that sensory information leads to the specification of current action possibilities which then compete with each other for ultimate selection for behavior (for uses of the word affordance, see also Gibson, 1986; McGrenere & Ho, 2000). Our results could be interpreted within this framework, as the shape of an object may lead to the specification of orienting affordances, or the possible ways to look and pay attention, some of which have a greater chance of being selected than others. Regardless, it is conceivable that the biases we report here can act as a front end to more traditionally defined affordance effects that involve physical interactions through reach and grasp (Bub & Masson, 2010; Cisek, 2007; Tucker & Ellis, 1998; for further discussion on the possible interplay between attention and affordance, see e.g. Anderson, Yamagishi, & Karavia, 2002; Handy, Grafton, Shroff, Ketay, & Gazzaniga, 2003; Riggio et al., 2008; Vainio, Ellis, & Tucker, 2007). Under most circumstances, people look where they are about to act, so eye gaze precedes hand movements both inside the laboratory and in real-world tasks (Ballard et al., 1992; Hayhoe, 2000; Hayhoe & Ballard, 2005; Land & Hayhoe, 2001; Land, Mennie, & Rusted, 1999). Eye orientation also directly affects where people reach (Enright, 1995), and dorsal stream posterior parietal regions important for saccade and reach planning appear to share a common eye-centered coordinate frame (Cohen & Andersen, 2002). Visual attention also appears to be directed from one object to another when familiar, manipulable objects are positioned in a manner that facilitates their interaction, such as when a hammer and a nail are seen together in a position that would allow the hammer to strike the nail (Green & Hummel, 2006; Riddoch et al., 2011; Roberts & Humphreys, 2011; Yoon, Humphreys, & Riddoch, 2010). It would be very interesting to see what, if any, role the effects reported here play in such paired-object affordance effects.
While eye and hand are clearly coupled, orienting biases such as those that we see here are in all likelihood not identical to reach and grasp affordances. It is not clear if unfamiliar 2D silhouettes on a computer screen afford reaching and grasping at all, and the effects are found even when the eyes and not the hands are used as effectors. It is also reasonable to assume that shape-induced orienting biases are evoked by objects beyond reach, such as birds in flight. In some cases, orienting biases might even directly oppose reach and grasp affordances. For example, when using a bottle or teapot, people tend to look at the bottle mouth or spout, or at the container into which they are pouring, instead of looking at their hands or the point of contact (Hayhoe, 2000; Land et al., 1999). Certain types of affordances and the effects reported here might nonetheless share the fundamental property of being “recognition free”, involving a more-or-less direct coupling between vision and action.
Being able to circumvent recognition does not necessitate complete isolation from it. Within the field of computer vision, the detection of a shape’s orientation is an often-used image preprocessing step applied before image registration and recognition (El-Sayed, Abdel-Kader, & Ramadan, 2010), and a number of algorithms have been developed to automatically detect the orientation and/or directionality of shapes (Cortadellas, Amat, & de la Torre, 2004; El-Sayed et al., 2010; Lin, 1996; Martinez-Ortiz & Zunic, 2010; Tzimiropoulos, Mitianoudis, & Stathaki, 2009; Zunic & Rosin, 2009; Zunic, Rosin, & Kopanja, 2006). The systematic and rapid extraction of an object’s directionality could also serve a role in human object recognition (see e.g. Leek & Johnston, 2006; Maki, 1986) by facilitating the search for and alignment to an existing object template or model.
In some cases, especially when objects are unfamiliar or if they are seen from an unfamiliar viewpoint, visual recognition is viewpoint-dependent (Rock, 1973; Tarr & Bülthoff, 1998; Tarr & Pinker, 1989; Tarr, Williams, Hayward, & Gauthier, 1998). When a previously seen object is encountered again from another viewpoint, the new object instance (or, alternatively, an internal reference frame; Robertson, Palmer, & Gomez, 1987) is thought to go through an iterative transformation, such as a mental rotation (Carpenter & Just, 1978; Shepard & Metzler, 1971; Zacks, 2008; but see Farah & Hammond, 1988; Hayward, Zhou, Gauthier, & Harris, 2006; Turnbull & McCarthy, 1996) or alignment (Huttenlocher & Ullman, 1987), that orients the observed object with either a previously seen view or a privileged, canonical view (Jolicoeur, 1985, 1990; Jolicoeur & Landau, 1984; Palmer, Rosch, & Chase, 1981; Robertson et al., 1987; Tarr & Pinker, 1989). Stored object views have a particular orientation and handedness (Tarr & Pinker, 1989) which can be thought of as having a specific directionality. The shortest path between a new and stored view could conceivably be calculated based on the angular difference between the directionality of the stored and observed object.
If a sufficient match to a stored representation is not found, directionality could be used to standardize the building of a new representation that is not completely dependent on the viewpoint from which an object happens to be first seen; the visual input could for instance be transformed and stored in a canonical directionality, such as upright. There indeed appears to be a favored view from which an object is most readily recognized (Palmer et al., 1981; see also Blanz, Tarr, & Bülthoff, 1999; Turnbull, Laws, & McCarthy, 1995), and damage to the parietal cortex can lead to specific deficits in recognizing objects from other, more unconventional views (Warrington & Taylor, 1973). The loss of the ability to automatically extract an object’s directionality could hypothetically lead to such a deficit by preventing the correct normalization to a canonical object representation. The suggested route to recognition is just one of potentially many possible ways to identify an object (Jolicoeur, 1990; Lawson, 1999; Vanrie, Béatse, Wagemans, Sunaert, & Van Hecke, 2002), some of which may not rely in any way on a shape’s directionality. Independent of these speculations, here we have shown that shape influences processes beyond recognition, and these findings may provide insight into why object form may be processed in parallel throughout the visual brain.
Supplementary Material
Figure 8. Congruency effects over the course of learning.
Mean response times (RT) are shown for experimental blocks in experiment 4. A central shape cue predicted, with 100% accuracy, where a peripheral target would appear. The correctly predicted location could be congruent (light gray) or incongruent (dark gray) with the shape’s inherent directionality.
Acknowledgments
This work was supported by National Science Foundation Grant IIS-0827427 (David L. Sheinberg), National Science Foundation Grant SBE-0542013 (Temporal Dynamics of Learning Center), NIH Grant R01EY14681 (David L. Sheinberg), and International Fulbright Science and Technology Award (Heida M. Sigurdardottir). We want to thank Dan Brooks for helpful discussions on the reported experiments.
References
- Andersen RA. Neural mechanisms of visual motion perception in primates. Neuron. 1997;18(6):865–872. doi: 10.1016/S0896-6273(00)80326-8. [DOI] [PubMed] [Google Scholar]
- Andersen RA, Snyder LH, Batista AP, Buneo CA, Cohen YE. Sensory Guidance of Movement. Chichester, West Sussex, England: John Wiley & Sons, Ltd; 1998. Posterior parietal areas specialized for eye movements (LIP) and reach (PRR) using a common coordinate frame. In Novartis Foundation; pp. 109–128. [DOI] [PubMed] [Google Scholar]
- Anderson SJ, Yamagishi N, Karavia V. Attentional processes link perception and action. Proceedings of the Royal Society of London Series B: Biological Sciences. 2002;269(1497):1225–1232. doi: 10.1098/rspb.2002.1998. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Arcizet F, Mirpour K, Bisley JW. A pure salience response in posterior parietal cortex. Cerebral Cortex. 2011;21(11):2498–2506. doi: 10.1093/cercor/bhr035. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Awh E, Pashler H. Evidence for split attentional foci. Journal of Experimental Psychology: Human Perception and Performance. 2000;26(2):834–846. doi: 10.1037/0096-1523.26.2.834. [DOI] [PubMed] [Google Scholar]
- Ballard DH, Hayhoe MM, Li F, Whitehead SD, Frisby J, Taylor J, Fisher R. Hand-eye coordination during sequential tasks [and discussion] Philosophical Transactions of the Royal Society of London Series B: Biological Sciences. 1992;337(1281):331–339. doi: 10.1098/rstb.1992.0111. [DOI] [PubMed] [Google Scholar]
- Berens P. CircStat: A MATLAB toolbox for circular statistics. Journal of Statistical Software. 2009;31(10):1–21. Retrieved from URL: http://www.jstatsoft.org/v31/i10. [Google Scholar]
- Bisley JW, Goldberg ME. Attention, intention, and priority in the parietal lobe. Annual Review of Neuroscience. 2010;33:1–21. doi: 10.1146/annurev-neuro-060909-152823. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Blanz V, Tarr MJ, Bülthoff HH. What object attributes determine canonical views? Perception-London. 1999;28(5):575–600. doi: 10.1068/p2897. [DOI] [PubMed] [Google Scholar]
- Blum H. A transformation for extracting new descriptors of shape. In: Wathen-Dunn W, editor. Models for the Perception of Speech and Visual Form. MIT Press; Cambridge, MA: 1967. pp. 362–380. [Google Scholar]
- Braun J. On the detection of salient contours. Spatial Vision. 1999;12(2):211–225. doi: 10.1163/156856899X00120. [DOI] [PubMed] [Google Scholar]
- Bub DN, Masson MEJ. Grasping beer mugs: On the dynamics of alignment effects induced by handled objects. Journal of Experimental Psychology: Human Perception and Performance. 2010;36(2):341. doi: 10.1037/a0017606. [DOI] [PubMed] [Google Scholar]
- Burr DC, Ross J. Direct evidence that “speedlines” influence motion mechanisms. Journal of Neuroscience. 2002;22(19):8661–8664. doi: 10.1523/JNEUROSCI.22-19-08661.2002. Retrieved from URL: http://www.jneurosci.org/content/22/19/8661.abstract?sid=61c663bf-d404-4d58-b13a-5e0417339241. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Caplovitz G, Tse P. Rotating dotted ellipses: Motion perception driven by grouped figural rather than local dot motion signals. Vision Research. 2007;47(15):1979–1991. doi: 10.1016/j.visres.2006.12.022. [DOI] [PubMed] [Google Scholar]
- Carpenter P, Just M. Eye fixations during mental rotation. In: Senders JW, Fisher DF, Monty RA, editors. Eye movements and the higher psychological functions. Hilldale NJ: Lawrence Erlbaum Associates; 1978. pp. 115–133. [Google Scholar]
- Cass JR, Spehar B. Dynamics of collinear contrast facilitation are consistent with long-range horizontal striate transmission. Vision Research. 2005;45(21):2728–2739. doi: 10.1016/j.visres.2005.03.010. [DOI] [PubMed] [Google Scholar]
- Cavanagh P. Attention-based motion perception. Science. 1992;257(5076):1563–1565. doi: 10.1126/science.1523411. [DOI] [PubMed] [Google Scholar]
- Cisek P. Cortical mechanisms of action selection: The affordance competition hypothesis. Philosophical Transactions of the Royal Society of London Series B: Biological Sciences. 2007;362(1485):1585. doi: 10.1098/rstb.2007.2054. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Cisek P, Kalaska J. Neural mechanisms for interacting with a world full of action choices. Annual Review of Neuroscience. 2010;33:269–298. doi: 10.1146/annurev.neuro.051508.135409. [DOI] [PubMed] [Google Scholar]
- Cohen YE, Andersen RA. A common reference frame for movement plans in the posterior parietal cortex. Nature Reviews Neuroscience. 2002;3(7):553–562. doi: 10.1038/nrn873. [DOI] [PubMed] [Google Scholar]
- Colby CL, Goldberg ME. Space and attention in parietal cortex. Annual Review of Neuroscience. 1999;22:319–349. doi: 10.1146/annurev.neuro.22.1.319. [DOI] [PubMed] [Google Scholar]
- Cortadellas J, Amat J, de la Torre F. Robust normalization of silhouettes for recognition applications. Pattern Recognition Letters. 2004;25(5):591–601. doi: 10.1016/j.patrec.2003.12.003. [DOI] [Google Scholar]
- Cousineau D. Confidence intervals in within-subject designs: A simpler solution to Loftus and Masson’s method. Tutorials in Quantitative Methods for Psychology. 2005;1(1):42–45. Retrieved from URL: http://www.tqmp.org/Content/vol01-1/p042/p042.pdf. [Google Scholar]
- Dehaene S, Bossini S, Giraux P. The mental representation of parity and number magnitude. Journal of Experimental Psychology: General. 1993;122(3):371. doi: 10.1037/0096-3445.122.3.371. [DOI] [Google Scholar]
- Dehaene S, Izard V, Spelke E, Pica P. Log or linear? Distinct intuitions of the number scale in Western and Amazonian indigene cultures. Science. 2008;320(5880):1217–1220. doi: 10.1126/science.1156540. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Desimone R, Albright TD, Gross CG, Bruce C. Stimulus-selective properties of inferior temporal neurons in the macaque. Journal of Neuroscience. 1984;4(8):2051–2062. doi: 10.1523/JNEUROSCI.04-08-02051.1984. Retrieved from URL: http://www.jneurosci.org/content/4/8/2051.short. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Dodd MD, Wilson D. Training attention: Interactions between central cues and reflexive attention. Visual Cognition. 2009;17(5):736–754. doi: 10.1080/13506280802340711. [DOI] [Google Scholar]
- Dresp B. Bright lines and edges facilitate the detection of small light targets. Spatial vision. 1993;7(3):213–225. doi: 10.1163/156856893X00379. [DOI] [PubMed] [Google Scholar]
- Driver J, Baylis GC. Movement and visual attention: the spotlight metaphor breaks down. Journal of Experimental Psychology: Human Perception and Performance. 1989;15(3):448–456. doi: 10.1037/0096-1523.15.3.448. [DOI] [PubMed] [Google Scholar]
- Driver J, Davis G, Ricciardelli P, Kidd P, Maxwell E, Baron-Cohen S. Gaze perception triggers reflexive visuospatial orienting. Visual Cognition. 1999;6(5):509–540. doi: 10.1080/135062899394920. [DOI] [Google Scholar]
- Duncan J. Selective attention and the organization of visual information. Journal of Experimental Psychology: General. 1984;113(4):501–517. doi: 10.1037/0096-3445.113.4.501. [DOI] [PubMed] [Google Scholar]
- Egly R, Driver J, Rafal RD. Shifting visual attention between objects and locations: evidence from normal and parietal lesion subjects. Journal of Experimental Psychology: General. 1994;123(2):161–177. doi: 10.1037/0096-3445.123.2.161. [DOI] [PubMed] [Google Scholar]
- El-Sayed E, Abdel-Kader RF, Ramadan RM. Orientation of multiple principal axes shapes using efficient averaging method. 2010 IEEE International Symposium on Signal Processing and Information Technology (ISSPIT); 2010. pp. 377–381. [DOI] [Google Scholar]
- Enright J. The non-visual impact of eye orientation on eye hand coordination. Vision Research. 1995;35(11):1611–1618. doi: 10.1016/0042-6989(94)00260-S. [DOI] [PubMed] [Google Scholar]
- Farah MJ, Hammond KM. Mental rotation and orientation-invariant object recognition: Dissociable processes. Cognition. 1988;29(1):29–46. doi: 10.1016/0010-0277(88)90007-8. [DOI] [PubMed] [Google Scholar]
- Fecteau JH, Munoz DP. Salience, relevance, and firing: A priority map for target selection. Trends in Cognitive Sciences. 2006;10(8):382–390. doi: 10.1016/j.tics.2006.06.011. [DOI] [PubMed] [Google Scholar]
- Fischer MH, Castel AD, Dodd MD, Pratt J. Perceiving numbers causes spatial shifts of attention. Nature Neuroscience. 2003;6(6):555–556. doi: 10.1038/nn1066. [DOI] [PubMed] [Google Scholar]
- Field DJ, Hayes A, Hess RF. Contour integration by the human visual system: Evidence for a local “association field”. Vision Research. 1993;33(2):173–193. doi: 10.1016/0042-6989(93)90156-Q. [DOI] [PubMed] [Google Scholar]
- Fischer MH, Warlop N, Hill RL, Fias W. Oculomotor bias induced by number perception. Experimental Psychology. 2004;51(2):91–97. doi: 10.1027/1618-3169.51.2.91. [DOI] [PubMed] [Google Scholar]
- Freyd JJ, Finke RA. Representational momentum. Journal of Experimental Psychology: Learning, Memory, and Cognition. 1984;10(1):126. doi: 10.1037/0278-7393.10.1.126. [DOI] [PubMed] [Google Scholar]
- Friesen C, Kingstone A. The eyes have it! Reflexive orienting is triggered by nonpredictive gaze. Psychonomic Bulletin & Review. 1998;5(3):490–495. doi: 10.3758/bf03208827. [DOI] [Google Scholar]
- Geisler WS. Motion streaks provide a spatial code for motion direction. Nature. 1999;400(6739):65–69. doi: 10.1038/21886. [DOI] [PubMed] [Google Scholar]
- Gibson J. The ecological approach to visual perception. Hillsdale, NJ: Lawrence Erlbaum Associates Inc; 1986. [Google Scholar]
- Goodale MA. Action without perception in human vision. Cognitive Neuropsychology. 2008;25(7–8):891–919. doi: 10.1080/02643290801961984. [DOI] [PubMed] [Google Scholar]
- Goodale MA, Milner AD. Separate visual pathways for perception and action. Trends in Neurosciences. 1992;15(1):20–25. doi: 10.1016/0166-2236(92)90344-8. [DOI] [PubMed] [Google Scholar]
- Gottlieb JP, Kusunoki M, Goldberg ME. The representation of visual salience in monkey parietal cortex. Nature. 1998;391(6666):481–484. doi: 10.1038/35135. [DOI] [PubMed] [Google Scholar]
- Green C, Hummel JE. Familiar interacting object pairs are perceptually grouped. Journal of Experimental Psychology: Human Perception and Performance; Journal of Experimental Psychology: Human Perception and Performance. 2006;32(5):1107–1119. doi: 10.1037/0096-1523.32.5.1107. [DOI] [PubMed] [Google Scholar]
- Grill-Spector K, Malach R. The human visual cortex. Annual Review of Neuroscience. 2004;27:649–677. doi: 10.1146/annurev.neuro.27.070203.144220. [DOI] [PubMed] [Google Scholar]
- Gross CG, Rocha-Miranda CE, Bender DB. Visual properties of neurons in inferotemporal cortex of the macaque. Journal of Neurophysiology. 1972;35(1):96–111. doi: 10.1152/jn.1972.35.1.96. Retreived from URL: ftp://lsr-ftp.nei.nih.gov/lsr/ArchiveDB/p0001526.pdf. [DOI] [PubMed] [Google Scholar]
- Hahn S, Kramer AF. Further evidence for the division of attention among non-contiguous locations. Visual Cognition. 1998;5(1–2):217–256. doi: 10.1080/713756781. [DOI] [Google Scholar]
- Handy TC, Grafton ST, Shroff NM, Ketay S, Gazzaniga MS. Graspable objects grab attention when the potential for action is recognized. Nature Neuroscience. 2003;6(4):421–427. doi: 10.1038/nn1031. [DOI] [PubMed] [Google Scholar]
- Hayhoe M. Vision using routines: A functional account of vision. Visual Cognition. 2000;7(1–3):43–64. doi: 10.1080/135062800394676. [DOI] [Google Scholar]
- Hayhoe M, Ballard D. Eye movements in natural behavior. Trends in Cognitive Sciences. 2005;9(4):188–194. doi: 10.1016/j.tics.2005.02.009. [DOI] [PubMed] [Google Scholar]
- Hayward WG, Zhou G, Gauthier I, Harris IM. Dissociating viewpoint costs in mental rotation and object recognition. Psychonomic Bulletin & Review. 2006;13(5):820–825. doi: 10.3758/BF03194003. [DOI] [PubMed] [Google Scholar]
- He P, Kowler E. Saccadic localization of eccentric forms. Journal of the Optical Society of America A. 1991;8(2):440–449. doi: 10.1364/JOSAA.8.000440. [DOI] [PubMed] [Google Scholar]
- Hommel B, Pratt J, Colzato L, Godijn R. Symbolic control of visual attention. Psychological Science. 2001;12(5):360–365. doi: 10.1111/1467-9280.00367. [DOI] [PubMed] [Google Scholar]
- Hung CC, Carlson ET, Connor CE. Medial Axis Shape Coding in Macaque Inferotemporal Cortex. Neuron. 2012;74(6):1099–1113. doi: 10.1016/j.neuron.2012.04.029. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Huttenlocher DP, Ullman S. Object recognition using alignment. Proceeding of the 1st International Conference on Computer Vision; London. Washington, D.C: IEEE; 1987. pp. 102–111. [Google Scholar]
- Itti L, Koch C. Computational modelling of visual attention. Nature Reviews Neuroscience. 2001;2(3):194–203. doi: 10.1038/35058500. [DOI] [PubMed] [Google Scholar]
- Janssen P, Srivastava S, Ombelet S, Orban GA. Coding of shape and position in macaque lateral intraparietal area. Journal of Neuroscience. 2008;28(26):6679–6690. doi: 10.1523/JNEUROSCI.0499-08.2008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Jolicoeur P. The time to name disoriented natural objects. Memory & Cognition. 1985;13(4):289–303. doi: 10.3758/BF03202498. [DOI] [PubMed] [Google Scholar]
- Jolicoeur P. Identification of disoriented objects: A dual-systems theory. Mind & Language. 1990;5(4):387–410. doi: 10.1111/j.1468-0017.1990.tb00170.x. [DOI] [Google Scholar]
- Jolicoeur P, Landau MJ. Effects of orientation on the identification of simple visual patterns. Canadian Journal of Psychology/Revue canadienne de psychologie. 1984;38(1):80–93. doi: 10.1037/h0080782. [DOI] [PubMed] [Google Scholar]
- Jonides J. Voluntary versus automatic control over the mind’s eye’s movement. In: Long JB, Baddeley AD, editors. Attention and performance IX. Hillsdale, New Jersey: Erlbaum; 1981. pp. 187–203. [Google Scholar]
- Kimia BB, Appear T. On the role of medial geometry in human vision. Journal of Physiology-Paris. 2003;97(2–3):155–190. doi: 10.1016/j.jphysparis.2003.09.003. [DOI] [PubMed] [Google Scholar]
- Konen CS, Kastner S. Two hierarchically organized neural systems for object information in human visual cortex. Nature Neuroscience. 2008;11(2):224–231. doi: 10.1038/nn2036. [DOI] [PubMed] [Google Scholar]
- Kourtzi Z, Kanwisher N. Activation in human MT/MST by static images with implied motion. Journal of Cognitive Neuroscience. 2000;12(1):48–55. doi: 10.1162/08989290051137594. [DOI] [PubMed] [Google Scholar]
- Kourtzi Z, Krekelberg B, van Wezel RJ. Linking form and motion in the primate brain. Trends in Cognitive Sciences. 2008;12(6):230–236. doi: 10.1016/j.tics.2008.02.013. [DOI] [PubMed] [Google Scholar]
- Kovács I, Fehér Á, Julesz B. Medial-point description of shape: a representation for action coding and its psychophysical correlates. Vision Research. 1998;38(15):2323–2333. doi: 10.1016/S0042-6989(97)00321-0. [DOI] [PubMed] [Google Scholar]
- Kovács I, Julesz B. Perceptual sensitivity maps within globally defined visual shapes. Nature. 1994;370(6491):644–646. doi: 10.1038/370644a0. [DOI] [PubMed] [Google Scholar]
- Kramer AF, Hahn S. Splitting the beam: Distribution of attention over noncontiguous regions of the visual field. Psychological Science. 1995;6(6):381–386. doi: 10.1111/j.1467-9280.1995.tb00530.x. [DOI] [Google Scholar]
- Krekelberg B, Dannenberg S, Hoffmann KP, Bremmer F, Ross J. Neural correlates of implied motion. Nature. 2003;424(6949):674–677. doi: 10.1038/nature01852. [DOI] [PubMed] [Google Scholar]
- Krekelberg B, Vatakis A, Kourtzi Z. Implied motion from form in the human visual cortex. Journal of Neurophysiology. 2005;94(6):4373–4386. doi: 10.1152/jn.00690.2005. [DOI] [PubMed] [Google Scholar]
- Kristjánsson Á, Sigurdardottir HM. On the benefits of transient attention across the visual field. Perception. 2008;37:474–764. doi: 10.1068/p5922. [DOI] [PubMed] [Google Scholar]
- Kuhn G, Kingstone A. Look away! Eyes and arrows engage oculomotor responses automatically. Attention, Perception, & Psychophysics. 2009;71(2):314–327. doi: 10.3758/APP.71.2.314. [DOI] [PubMed] [Google Scholar]
- Land MF, Hayhoe M. In what ways do eye movements contribute to everyday activities? Vision Research. 2001;41(25):3559–3565. doi: 10.1016/S0042-6989(01)00102-X. [DOI] [PubMed] [Google Scholar]
- Land MF, Mennie N, Rusted J. The roles of vision and eye movements in the control of activities of daily living. Perception. 1999;28(11):1311–1328. doi: 10.1068/p2935. [DOI] [PubMed] [Google Scholar]
- Lawson R. Achieving visual object constancy across plane rotation and depth rotation. Acta Psychologica. 1999;102(2):221–245. doi: 10.1016/S0001-6918(98)00052-3. [DOI] [PubMed] [Google Scholar]
- Ledgeway T, Hess RF, Geisler WS. Grouping local orientation and direction signals to extract spatial contours: Empirical tests of “association field” models of contour integration. Vision Research. 2005;45(19):2511–2522. doi: 10.1016/j.visres.2005.04.002. [DOI] [PubMed] [Google Scholar]
- Leek DEC, Johnston SJ. A polarity effect in misoriented object recognition: The role of polar features in the computation of orientation-invariant shape representations. Visual Cognition. 2006;13(5):573–600. doi: 10.1080/13506280544000048. [DOI] [Google Scholar]
- Lehky SR, Sereno AB. Comparison of shape encoding in primate dorsal and ventral visual pathways. Journal of Neurophysiology. 2007;97(1):307–319. doi: 10.1152/jn.00168.2006. [DOI] [PubMed] [Google Scholar]
- Lin J. The family of universal axes. Pattern Recognition. 1996;29(3):477–485. doi: 10.1016/0031-3203(95)00095-X. [DOI] [Google Scholar]
- Loffler G. Perception of contours and shapes: low and intermediate stage mechanisms. Vision Research. 2008;48(20):2106–2127. doi: 10.1016/j.visres.2008.03.006. [DOI] [PubMed] [Google Scholar]
- Logothetis N, Sheinberg D. Visual object recognition. Annual Review of Neuroscience. 1996;19(1):577–621. doi: 10.1146/annurev.ne.19.030196.003045. [DOI] [PubMed] [Google Scholar]
- Mack A, Fendrich R, Chambers D, Heuer F. Perceived position and saccadic eye movements. Vision Research. 1985;25(4):501–505. doi: 10.1016/0042-6989(85)90152-X. [DOI] [PubMed] [Google Scholar]
- Maki RH. Naming and locating the tops of rotated pictures. Canadian Journal of Psychology/Revue canadienne de psychologie. 1986;40(4):368. doi: 10.1037/h0080104. [DOI] [PubMed] [Google Scholar]
- Martinez-Ortiz C, Zunic J. Curvature weighted gradient based shape orientation. Pattern Recognition. 2010;43(9):3035–3041. doi: 10.1016/j.patcog.2010.03.026. [DOI] [Google Scholar]
- McGrenere J, Ho W. Affordances: Clarifying and evolving a concept. Proceedings of Graphics Interface. 2000;2000:179–186. Retrieved from URL: http://www.graphicsinterface.org/proceedings/2000/177/ [Google Scholar]
- Melcher D, Kowler E. Shapes, surfaces and saccades. Vision Research. 1999;39(17):2929–2946. doi: 10.1016/S0042-6989(99)00029-2. [DOI] [PubMed] [Google Scholar]
- Metzger W, Spillmann LT, Lehar ST, Stromeyer MT, Wertheimer MT. Laws of seeing. MIT Press; 2006. [Google Scholar]
- Milner AD, Goodale MA. The visual brain in action. Oxford, UK: Oxford University Press; 1995. [Google Scholar]
- Milner AD, Goodale MA. Cortical visual systems for perception and action. In: Gangopadhyay N, Madary M, Spicer F, editors. Perception, action, and consciousness: Sensorimotor dynamics and two visual systems. New York: Oxford University Press; 2010. pp. 71–95. [Google Scholar]
- Morey RD. Confidence intervals from normalized data: A correction to Cousineau (2005) Tutorials in Quantitative Methods for Psychology. 2008;4(2):61–64. Retrieved from URL: http://tqmp.org/Content/vol04-2/p061/p061.pdf. [Google Scholar]
- Mountcastle VB, Lynch JC, Georgopoulos A, Sakata H, Acuna C. Posterior parietal association cortex of the monkey: Command functions for operations within extrapersonal space. Journal of Neurophysiology. 1975;38(4):871–908. doi: 10.1152/jn.1975.38.4.871. Retrieved from URL: http://jn.physiology.org/content/38/4/871.long. [DOI] [PubMed] [Google Scholar]
- Murata A, Gallese V, Luppino G, Kaseda M, Sakata H. Selectivity for the shape, size, and orientation of objects for grasping in neurons of monkey parietal area AIP. Journal of Neurophysiology. 2000;83(5):2580–2601. doi: 10.1152/jn.2000.83.5.2580. Retrieved from URL: http://jn.physiology.org/content/83/5/2580.short. [DOI] [PubMed] [Google Scholar]
- Murdoch DJ, Chow ED. A graphical display of large correlation matrices. The American Statistician. 1996;50(2):178–180. doi: 10.1080/00031305.1996.10474371. [DOI] [Google Scholar]
- Murta A. A general polygon clipping library. Advanced Interfaces Group, Department of Computer Science, University of Manchester; 2000. Online resource. URL: http://www.cs.man.ac.uk/~toby/alan/software/gpc.html. [Google Scholar]
- Nakayama K, Mackeben M. Sustained and transient components of focal visual attention. Vision Research. 1989;29(11):1631–1647. doi: 10.1016/0042-6989(89)90144-2. [DOI] [PubMed] [Google Scholar]
- Oliver RT, Thompson-Schill SL. Dorsal stream activation during retrieval of object size and shape. Cognitive, Affective, & Behavioral Neuroscience. 2003;3(4):309–322. doi: 10.3758/CABN.3.4.309. [DOI] [PubMed] [Google Scholar]
- Palmer S, Rosch E, Chase P. Canonical perspective and the perception of objects. In: Long J, Baddeley A, editors. Attention and performance IX. Hillsdale NJ: Erlbaum; 1981. pp. 135–151. [Google Scholar]
- Peng X, Sereno ME, Silva AK, Lehky SR, Sereno AB. Shape selectivity in primate frontal eye field. Journal of Neurophysiology. 2008;100(2):796–814. doi: 10.1152/jn.01188.2007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Polat U, Bonneh Y. Collinear interactions and contour integration. Spatial vision. 2000;13(4):393–401. doi: 10.1163/156856800741270. [DOI] [PubMed] [Google Scholar]
- Polat U, Sagi D. Lateral interactions between spatial channels: suppression and facilitation revealed by lateral masking experiments. Vision Research. 1993;33(7):993–999. doi: 10.1016/0042-6989(93)90081-7. [DOI] [PubMed] [Google Scholar]
- Polat U, Sagi D. The architecture of perceptual spatial interactions. Vision Research, 34. 1994;(1):73–78. doi: 10.1016/0042-6989(94)90258-5. [DOI] [PubMed] [Google Scholar]
- Red SD, Patel SS, Sereno AB. Shape effects on reflexive spatial attention are driven by the dorsal stream. Vision Research. 2012;55:32–40. doi: 10.1016/j.visres.2011.12.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Riddoch MJ, Pippard B, Booth L, Rickell J, Summers J, Brownson A, Humphreys GW. Effects of action relations on the configural coding between objects. Journal of Experimental Psychology: Human Perception and Performance. 2011;37(2):580–587. doi: 10.1037/a0020745. [DOI] [PubMed] [Google Scholar]
- Riggio L, Iani C, Gherri E, Benatti F, Rubichi S, Nicoletti R. The role of attention in the occurrence of the affordance effect. Acta Psychologica. 2008;127(2):449–458. doi: 10.1016/j.actpsy.2007.08.008. [DOI] [PubMed] [Google Scholar]
- Roberts KL, Humphreys GW. Action-related objects influence the distribution of visuospatial attention. The Quarterly journal of experimental psychology. 2011;64(4):669–688. doi: 10.1080/17470218.2010.520086. [DOI] [PubMed] [Google Scholar]
- Robertson LC, Palmer SE, Gomez LM. Reference frames in mental rotation. Journal of Experimental Psychology: Learning, Memory, and Cognition. 1987;13(3):368–379. doi: 10.1037/0278-7393.13.3.368. [DOI] [PubMed] [Google Scholar]
- Rock I. Orientation and form. New York: Academic Press; 1973. [Google Scholar]
- Ross J, Badcock DR, Hayes A. Coherent global motion in the absence of coherent velocity signals. Current Biology. 2000;10(11):679–682. doi: 10.1016/S0960-9822(00)00524-8. [DOI] [PubMed] [Google Scholar]
- Sakata H, Taira M, Kusunoki M, Murata A, Tanaka Y, Tsutsui K. Neural coding of 3D features of objects for hand action in the parietal cortex of the monkey. Philosophical Transactions of the Royal Society of London Series B: Biological Sciences. 1998;353(1373):1363–1373. doi: 10.1098/rstb.1998.0290. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Sakata H, Taira M, Murata A, Mine S. Neural mechanisms of visual guidance of hand action in the parietal cortex of the monkey. Cerebral Cortex. 1995;5(5):429–438. doi: 10.1093/cercor/5.5.429. [DOI] [PubMed] [Google Scholar]
- Senior C, Barnes J, Giampietroc V, Simmons A, Bullmore E, Brammer M, David A. The functional neuroanatomy of implicit-motion perception or ‘representational momentum’. Current Biology. 2000;10(1):16–22. doi: 10.1016/S0960-9822(99)00259-6. [DOI] [PubMed] [Google Scholar]
- Senior C, Ward J, David AS. Representational momentum and the brain: An investigation into the functional necessity of V5/MT. Visual Cognition. 2002;9(1–2):81–92. doi: 10.1080/13506280143000331. [DOI] [Google Scholar]
- Sereno AB, Amador SC. Attention and memory-related responses of neurons in the lateral intraparietal area during spatial and shape-delayed match-to-sample tasks. Journal of Neurophysiology. 2006;95(2):1078–1098. doi: 10.1152/jn.00431.2005. [DOI] [PubMed] [Google Scholar]
- Sereno AB, Maunsell JH. Shape selectivity in primate lateral intraparietal cortex. Nature. 1998;395(6701):500–503. doi: 10.1038/26752. [DOI] [PubMed] [Google Scholar]
- Sereno ME, Trinath T, Augath M, Logothetis NK. Three-dimensional shape representation in monkey cortex. Neuron. 2002;33(4):635–652. doi: 10.1016/S0896-6273(02)00598-6. [DOI] [PubMed] [Google Scholar]
- Seriès P, Georges S, Lorenceau J, Frégnac Y. Orientation dependent modulation of apparent speed: a model based on the dynamics of feed-forward and horizontal connectivity in V1 cortex. Vision Research. 2002;42(25):2781–2797. doi: 10.1016/S0042-6989(02)00302-4. [DOI] [PubMed] [Google Scholar]
- Shaked D, Bruckstein AM. Pruning medial axes. Computer Vision and Image Understanding. 1998;69(2):156–169. doi: 10.1006/cviu.1997.0598. [DOI] [Google Scholar]
- Shaki S, Fischer MH. Reading space into numbers: A cross-linguistic comparison of the SNARC effect. Cognition. 2008;108(2):590–599. doi: 10.1016/j.cognition.2008.04.001. [DOI] [PubMed] [Google Scholar]
- Shepard RN, Metzler J. Mental rotation of three-dimensional objects. Science. 1971;171(3972):701–703. doi: 10.1126/science.171.3972.701. [DOI] [PubMed] [Google Scholar]
- Silver MA, Ress D, Heeger DJ. Topographic maps of visual spatial attention in human parietal cortex. Journal of Neurophysiology. 2005;94(2):1358–1371. doi: 10.1152/jn.01316.2004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Spering M, Gegenfurtner KR. Contextual effects on motion perception and smooth pursuit eye movements. Brain research. 2008;1225:76–85. doi: 10.1016/j.brainres.2008.04.061. [DOI] [PubMed] [Google Scholar]
- Spering M, Montagnini A. Do we track what we see? Common versus independent processing for motion perception and smooth pursuit eye movements: A review. Vision Research. 2011;51(8):836–852. doi: 10.1016/j.visres.2010.10.017. [DOI] [PubMed] [Google Scholar]
- Stelmach LB, Herdman CM, McNeil K. Attentional modulation of visual processes in motion perception. Journal of Experimental Psychology: Human Perception and Performance. 1994;20(1):108. doi: 10.1037/0096-1523.20.1.108. [DOI] [Google Scholar]
- Taira M, Mine S, Georgopoulos AP, Murata A, Sakata H. Parietal cortex neurons of the monkey related to the visual guidance of hand movement. Experimental Brain Research. 1990;83(1):29–36. doi: 10.1007/BF00232190. [DOI] [PubMed] [Google Scholar]
- Tanaka K, Saito H, Fukada Y, Moriya M. Coding visual images of objects in the inferotemporal cortex of the macaque monkey. Journal of Neurophysiology. 1991;66(1):170–189. doi: 10.1152/jn.1991.66.1.170. Retrieved from URL: http://jn.physiology.org/content/66/1/170.short. [DOI] [PubMed] [Google Scholar]
- Tarr MJ, Bülthoff HH. Image-based object recognition in man, monkey and machine. Cognition. 1998;67(1):1–20. doi: 10.1016/S0010-0277(98)00026-2. [DOI] [PubMed] [Google Scholar]
- Tarr MJ, Pinker S. Mental rotation and orientation-dependence in shape recognition. Cognitive Psychology. 1989;21(2):233–282. doi: 10.1016/0010-0285(89)90009-1. [DOI] [PubMed] [Google Scholar]
- Tarr MJ, Williams P, Hayward WG, Gauthier I. Three-dimensional object recognition is viewpoint dependent. Nature Neuroscience. 1998;1(4):275–277. doi: 10.1038/1089. [DOI] [PubMed] [Google Scholar]
- Tipples J. Eye gaze is not unique: automatic orienting in response to uninformative arrows. Psychonomic Bulletin & Review. 2002;9(2):314–318. doi: 10.3758/BF03196287. [DOI] [PubMed] [Google Scholar]
- Tipples J. Orienting to counterpredictive gaze and arrow cues. Attention, Perception, & Psychophysics. 2008;70(1):77–87. doi: 10.3758/PP.70.1.77. [DOI] [PubMed] [Google Scholar]
- Treue S, Maunsell JH. Attentional modulation of visual motion processing in cortical areas MT and MST. Nature. 1996;382(6591):539–541. doi: 10.1038/382539a0. [DOI] [PubMed] [Google Scholar]
- Tse PU, Logothetis NK. The duration of 3-D form analysis in transformational apparent motion. Attention, Perception, & Psychophysics. 2002;64(2):244–265. doi: 10.3758/BF03195790. [DOI] [PubMed] [Google Scholar]
- Tucker M, Ellis R. On the relations between seen objects and components of potential actions. Journal of Experimental Psychology: Human Perception and Performance. 1998;24(3):830. doi: 10.1037/0096-1523.24.3.830. [DOI] [PubMed] [Google Scholar]
- Turnbull OH, Laws KR, McCarthy RA. Object recognition without knowledge of object orientation. Cortex: A Journal Devoted to the Study of the Nervous System and Behavior. 1995;31(2):387–395. doi: 10.1016/s0010-9452(13)80371-1. [DOI] [PubMed] [Google Scholar]
- Turnbull OH, McCarthy RA. When is a view unusual? A single case study of orientation-dependent visual agnosia. Brain Research Bulletin. 1996;40(5):497–502. doi: 10.1016/0361-9230(96)00148-7. [DOI] [PubMed] [Google Scholar]
- Tzimiropoulos G, Mitianoudis N, Stathaki T. A unifying approach to moment-based shape orientation and symmetry classification. IEEE Transactions on Image Processing. 2009;18(1):125–139. doi: 10.1109/TIP.2008.2007050. [DOI] [PubMed] [Google Scholar]
- Ungerleider LG, Mishkin M. Two cortical visual systems. In: Ingle DJ, Goodale MA, Mansfield RJW, editors. Analysis of visual behavior. Cambridge, MA: MIT Press; 1982. pp. 549–586. [Google Scholar]
- Vainio L, Ellis R, Tucker M. The role of visual attention in action priming. The Quarterly Journal of Experimental Psychology. 2007;60(2):241–261. doi: 10.1080/17470210600625149. [DOI] [PubMed] [Google Scholar]
- Van der Stigchel S, Mills M, Dodd MD. Shift and deviate: Saccades reveal that shifts of covert attention evoked by trained spatial stimuli are obligatory. Attention, Perception, & Psychophysics. 2010;72(5):1244–1250. doi: 10.3758/APP.72.5.1244. [DOI] [PubMed] [Google Scholar]
- Van Essen DC, Gallant JL. Neural mechanisms of form and motion processing in the primate visual system. Neuron. 1994;13(1):1–10. doi: 10.1016/0896-6273(94)90455-3. [DOI] [PubMed] [Google Scholar]
- Vanrie J, Béatse E, Wagemans J, Sunaert S, Van Hecke P. Mental rotation versus invariant features in object perception from different viewpoints: An fMRI study. Neuropsychologia. 2002;40(7):917–930. doi: 10.1016/S0028-3932(01)00161-0. [DOI] [PubMed] [Google Scholar]
- Vatti BR. A generic solution to polygon clipping. Communications of the ACM. 1992;35(7):56–63. doi: 10.1145/129902.129906. [DOI] [Google Scholar]
- Vishwanath D, Kowler E, Feldman J. Saccadic localization of occluded targets. Vision Research. 2000;40(20):2797–2811. doi: 10.1016/S0042-6989(00)00118-8. [DOI] [PubMed] [Google Scholar]
- Warrington EK, Taylor AM. The contribution of the right parietal lobe to object recognition. Cortex: A Journal Devoted to the Study of the Nervous System and Behavior. 1973;9(2):152–164. doi: 10.1016/s0010-9452(73)80024-3. [DOI] [PubMed] [Google Scholar]
- Williams CB, Hess RF. Relationship between facilitation at threshold and suprathreshold contour integration. JOSA A. 1998;15(8):2046–2051. doi: 10.1364/JOSAA.15.002046. [DOI] [PubMed] [Google Scholar]
- Williams ZM, Elfar JC, Eskandar EN, Toth LJ, Assad JA. Parietal activity and the perceived direction of ambiguous apparent motion. Nature Neuroscience. 2003;6(6):616–623. doi: 10.1038/nn1055. [DOI] [PubMed] [Google Scholar]
- Wolfe JM, Klempen NL, Shulman EP. Which end is up? Two representations of orientation in visual search. Vision Research. 1999;39(12):2075–2086. doi: 10.1016/S0042-6989(98)00260-0. [DOI] [PubMed] [Google Scholar]
- Wong E, Mack A. Saccadic programming and perceived location. Acta Psychologica. 1981;48(1):123–131. doi: 10.1016/0001-6918(81)90054-8. [DOI] [PubMed] [Google Scholar]
- Yoon EY, Humphreys GW, Riddoch MJ. The paired-object affordance effect. Journal of Experimental Psychology: Human Perception and Performance. 2010;36(4):812–824. doi: 10.1037/a0017175. [DOI] [PubMed] [Google Scholar]
- Zacks JM. Neuroimaging studies of mental rotation: A meta-analysis and review. Journal of Cognitive Neuroscience. 2008;20(1):1–19. doi: 10.1162/jocn.2008.20013. [DOI] [PubMed] [Google Scholar]
- Zorzi M, Priftis K, Umilta C. Brain damage: Neglect disrupts the mental number line. Nature. 2002;417(6885):138–139. doi: 10.1038/417138a. [DOI] [PubMed] [Google Scholar]
- Zunic J, Rosin P. An alternative approach to computing shape orientation with an application to compound shapes. International Journal of Computer Vision. 2009;81(2):138–154. doi: 10.1007/s11263-008-0149-1. [DOI] [Google Scholar]
- Zunic J, Rosin P, Kopanja L. On the orientability of shapes. IEEE Transactions on Image Processing. 2006;15(11):3478–3487. doi: 10.1109/TIP.2006.877527. [DOI] [PubMed] [Google Scholar]
- Zusne L, Michels KM. Nonrepresentational shapes and eye movements. Perceptual and Motor Skills. 1964;18(1):11–20. doi: 10.2466/pms.1964.18.1.11. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.