The action-specific account of perception proposes that perceivers see the surrounding environment in terms of their ability to act in it. For example, softball players who are hitting better than others see the ball as bigger (Witt & Proffitt, 2005). As another example, hills appear steeper to perceivers who are fatigued or less physically fit (Bhalla & Proffitt, 1999). The various accounts of these effects (Proffitt, 2006; Proffitt & Linkenauger, 2013; Witt, 2011a) emphasize the role of the perceiver’s own body and abilities. More recently, new research has shown similar patterns with respect to another person’s ability to act. On the face of it, these findings with regard to others’ actions seem consistent with the action-specific account of perception, but their theoretical implications have not yet been fully explored and may be at odds with the account’s main claims.

The effect of another person’s ability to act on perceptual judgments has been documented with respect to reaching ability and blocking ability. Targets presented beyond arm’s reach look closer when a perceiver intends to use a reach-extending tool to reach to the targets (Witt, Proffitt, & Epstein, 2005; see also Davoli, Brockmole, & Witt, 2012; Kirsch & Kunde, 2012; Linkenauger, Witt, Stefanucci, Bakdash, & Proffitt, 2009; Moragado, Gentaz, Guinet, Osiurak, & Palluel-Germain, 2013; Osiurak, Morgado, & Palluel-Germain, 2012; Witt, 2011b, Witt & Proffitt, 2008). Targets beyond arm’s reach also look closer when observing another person using the tool, even when the observer’s own abilities have not changed (Bloesch, Davoli, Roth, Brockmole, & Abrams, 2012). As another example, when a person attempts to block balls with various sized paddles, the balls appear slower when the paddle is bigger and, thus, more effective than when the paddle is smaller (Witt & Sugovic, 2010, 2012, 2013). Ball speed is similarly affected when another person is observed attempting to block the ball (Witt, Sugovic, & Taylor, 2012). This evidence suggests that another person’s ability to act also influences perception.

One of the main claims of the action-specific account is that, as Gibson (1979) theorized, the goal of perception is to express the relationship between the perceiver and the environment. Yet this claim is clearly challenged by the findings that another person’s ability to act can penetrate perception, because the effects are no longer specific to the perceiver–environment relationship. What, then, are the theoretical implications of other-based action-specific effects for the action-specific account? We outline three possibilities.

First, instead of perception expressing the relationship between the perceiver and the environment, perception might express the relationship between affordances (possibilities for action; Gibson, 1979) and the environment, regardless of who is performing the action. In some ways, this is not substantially different from the original claim, since both emphasize the role of affordances. The original action-specific account, however, also emphasizes that these effects are not errors in perception but, rather, are adaptive for planning future actions (see the Discussion section). For example, perceiving a target as smaller when they are shooting well (cf. Lee, Lee, Carello, & Turvey, 2012) could encourage hunters to shoot from their current location, rather than to move closer and risk detection (Witt, 2011a). However, this claim loses validity when one considers that seeing the world in terms of someone else’s ability is unlikely to be adaptive for planning one’s own action.

A second possibility is that apparent action-specific effects could be due to postperceptual processes such as task demands (e.g., Woods, Philbeck, & Danoff, 2009), rather than to changes in perception. Changes in another person’s ability to act could lead to the same modifications of these postperceptual processes as changes in one’s own ability to act. To date, much of the research in this area has demonstrated that postperceptual processes are unlikely to account for many action-specific effects (see Witt & Sugovic, 2013). However, a postperceptual explanation for other-based action-specific effects may be more parsimonious.

The third possibility is that other-based action-specific effects could still reflect an effect of the perceiver’s own abilities on perception. While observing another person perform an action, perceivers might anticipate their own abilities as if they were in the other person’s situation. Anticipation of their own abilities could then influence perception of the other person’s target. In this case, other-based action-specific effects would not challenge the main claim that perception expresses the relationship between the perceiver and the environment, because the effects would still be based on the perceiver’s own abilities. Furthermore, the idea that action-specific effects could be adaptive for planning future actions (e.g., Witt, 2011a) would be preserved, because it is the perceiver’s own abilities that influence perception.

To distinguish between these three possibilities, we must first determine the source of action-related information that influences perception. When observing another person, the perceiver could see the world relative to the other person’s abilities, a result that would challenge claims of the action-specific account either in terms of its adaptive nature (possibility 1) or in terms of its perceptual nature (possibility 2). Alternatively, the perceiver could also see the world relative to his or her own abilities, a result that would be consistent with the action-specific account (possibility 3). To determine whose abilities are relevant, we examined perception of ball speed in a ball-blocking paradigm when a perceiver performed the task and when he or she observed another person perform the task. We grouped participants on the basis of whether their ball-blocking ability was better or worse than their partner’s ability and examined whether perceived speed was a function of the observer’s own abilities or the partner’s abilities.

Method

Sample

Participants were recruited from an ongoing project examining personality and romantic relationship satisfaction. Eligibility for the larger study from which they were drawn included being married within the last 12 months and currently living together. Of the 78 individuals who completed the perception task, 2 were deemed outliers (scores greater than 3 times the interquartile range) and were removed, leaving 76 people in the final sample (38 men, 38 women). On average, the husbands (26.82 years, SD = 4.32) and wives (25.95 years, SD = 4.31) had dated for a mean length of 39 months (SD = 29.4) before marrying, and the average duration of marriage was 3.53 months (SD = 3.05). Participants completed informed consent at the initial baseline appointment; the perception task was completed at the third appointment, 2 weeks later. All participants were compensated for participation in the larger study.

Procedure

The procedure completed by all participants is described in detail in South, Witt, & Sugovic (2013) and is similar to the procedure in Witt et al. (2012; see Supplementary Materials). All participants completed a two-part training phase in which they were exposed to and tested on their ability to discriminate between two anchor speeds: slow (0.18 m/s) and fast (0.74 m/s). Then participants completed the test trials. On each test trial, the white ball (1.6 cm in diameter) moved across the screen from left to right along a diagonal at one of six possible speeds (0.26–0.67 m/s). The actor used a joystick to control the paddle and attempt to block the ball. The paddle was a white rectangle (width = 0.8 cm) set to one of three heights (1.9, 5.8, or 11.8 cm). If the actor was successful, the ball was stopped on the paddle; otherwise, it continued moving past the screen’s edge. Each participant then estimated ball speed by indicating whether the ball moved more like the slow speed or more like the fast speed. Each block contained 36 trials (two repetitions of each paddle size and speed combination). Participants completed four blocks in one role (actor or observer) and then switched with their partners.

Results

Speed bisection responses were summarized by calculating the point of subjective equality (PSE) from the slopes and intercepts from binary regressions for each participant for each paddle size for each role (actor, observer). When participants were in the actor condition, those who blocked the ball more successfully than others estimated the ball to be moving slower (see Fig. 1). A median split was conducted on the basis of mean blocking performance as the actor, thus dividing participants into better performers (n = 37; M = 76.7 %, SD = 3.7 %, range = 72 %–88 % balls successfully blocked) and worse performers (n = 39; M = 60.8 %, SD = 13.0 %, range = 26 %–71 % balls successfully blocked). As was expected, blocking performance significantly differed between the two groups, t(74) = 7.36, p < .001.

Fig. 1
figure 1

Points of subjective equality (PSEs) when participants served as the actor are plotted as a function of paddle size and blocking performance group. Those who blocked fewer balls successfully are labeled as worse blockers, and those who blocked more balls successfully are labeled as better blockers. Lower PSEs correspond to estimating the ball as moving faster. Error bars are 1 SEM calculated between subjects

Full size image

PSEs when participants served as the actor were submitted to a repeated measures ANOVA with paddle size as a repeated factor and performance group as a between-subjects factor. Paddle size significantly influenced the PSEs, F(2, 148) = 49.28, p < .001, η p 2 = .40. Participants estimated the ball as moving slower when the paddle was bigger, and more effective (see Supplementary Materials), than when it was smaller. This replicates prior research with this paradigm (Witt & Sugovic, 2010, 2012). Performance group also significantly influenced the PSEs, F(1, 74) = 6.21, p < .05, η p 2 = .08. Better blockers estimated the ball as moving slower than did worse blockers. In sports-related paradigms, the results of many studies have shown that those who are playing better than others see their sports-specific target as bigger (Gray, 2013; Kwon & Kim, 2012; Lee et al., 2012; Witt & Dorsch, 2009; Witt, Linkenauger, Bakdash, & Proffitt, 2008; Witt & Proffitt, 2005). This is the first time that similar effects have been shown with respect to perceived speed. It is possible that the effect has only emerged for the first time with this paradigm because of the wider range of performance levels (see Supplementary Materials), perhaps due to using a community sample. The interaction between paddle size and performance group was not significant, F(2, 148) = 0.01.

Next, we analyzed the effects of relative blocking performance on apparent speed when observing one’s partner play. Relative blocking performance was calculated as the mean number of balls one successfully blocked when playing minus the mean number of balls successfully blocked by one’s partner. Participants were divided into two groups on the basis of whether their overall blocking performance was better or worse than their partner’s blocking performance (ns = 38; worse than partner, M = −10.1 % relative successfully blocked balls, SD = 9.6 %, range = −41 % to −1 %; better than partner, M = 9.0 %, SD = 8.1 %, range = 1 % to 36 %). As was expected, the better-than-partner group was significantly better at blocking the ball than was the worse-than-partner group, t(74) > 9.

The critical test is whether participants see the ball in terms of their partner’s abilities or in terms of their own abilities as if they were in their partner’s situation. In other words, whose abilities are relevant? If the relevant abilities are those of the person playing the game (the actor), the observer should perceive the ball’s speed relative to the actor’s abilities. This would mean that when observers make perceptual judgments while watching their partner play, those who are better blockers than their partners will see the ball as moving faster, as compared with observers who are worse blockers than their partners; the partner’s lesser abilities would influence perception, and lesser ability corresponds with seeing the ball as faster. In contrast, if the relevant abilities are one’s own abilities, the observer should see the ball according to his or her own abilities. This would mean that when observing their partner play, observers who are better blockers than their partners will see the ball as moving slower, as compared with observers who are worse blockers than their partners, because it is the observer’s own better abilities that influence perception and better ability corresponds with seeing the ball as moving slower.

The data support the latter prediction: Those who were better blockers than their partners perceived the ball to be moving slower, consistent with effects based on their own abilities, rather than their partners’ abilities (see Fig. 2). PSEs when participants served as the observer were submitted to a repeated measures ANOVA with paddle size as a within-subjects factor and relative blocking performance group (better or worse than one’s partner) as a between-subjects factor.Footnote 1 Relative blocking performance group significantly influenced the PSEs, F(1, 74) = 4.94, p < .05, η p 2 = .06. Paddle size also influenced the PSEs, F(2, 148) = 21.05, p < .001, η p 2 = .22. Participants estimated the ball as moving slower when their partners played with the bigger paddle. As the observer, participants’ own task had not changed as a function of paddle size; they simply observed their partners play. Nevertheless, their perception of ball speed was still influenced by the ease with which the ball could be blocked. The interaction was not significant, F(2, 148) = 2.21, p > .11, η p 2 = .03.

Fig. 2
figure 2

Points of subjective equality (PSEs) when participants served as the observer are plotted as a function of paddle size and relative blocking performance group. Those who blocked fewer balls successfully than their partners are labeled as “Worse than Partner,” and those who blocked more balls successfully than their partners are labeled as “Better than Partner.” Note that these groupings are not the same as the groupings in Fig. 1. Lower PSEs correspond to estimating the ball as moving faster. Error bars are 1 SEM calculated between subjects

Full size image

That the perceiver’s own abilities, and not their partner’s abilities, influenced perception was further confirmed in an additional analysis for which both sets of abilities were included as predictors of apparent speed (see Supplementary Materials). In these analyses, we also included sex and individual trial performance (i.e., successful catch or miss) as potential confounds. The perceiver’s own abilities significantly influenced estimated speed as the observer, whereas their partner’s abilities did not.

Discussion

Recent research raises the issue of whether perceivers see the environment exclusively in terms of their own ability to act or whether another person’s abilities can penetrate their perception (Bloesch et al., 2012; Witt et al., 2012). The finding that another person’s ability to act can influence perception poses a challenge to accounts of action-specific effects, because these accounts emphasize the idea that perception expresses the relationship between the perceiver and the environment (Proffitt, 2006; Witt, 2011a). The most extreme interpretation of this alternative viewpoint would question whether action-specific effects are, indeed, even perceptual or whether they are, instead, due to preperceptual processes such as attention (Bloesch et al., 2012) or postperceptual processes such as response generation (Woods et al., 2009). A necessary first step is to determine whether perceivers see the world in terms of another person’s abilities or in terms of their own abilities as if they were in the other person’s situation.

Married couples played a modified version of the computer game Pong and observed their partners play. Performance was manipulated by varying the size of the paddle used to block the ball. After each attempt, both the actor and the observer estimated the speed of the ball. As actors, participants perceived the speed of the ball in terms of their own blocking ability. The ball appeared to be moving faster when the paddle was smaller than when the paddle was bigger. This replicates previous research using this paradigm (Witt & Sugovic, 2010, 2012, 2013). In addition, participants who were better at the blocking task than others also perceived the ball as moving slower. Whereas previous research has found similar effects for perceived size (e.g., Witt & Proffitt, 2005), this is the first demonstration that overall ability also influences perception of speed.

The critical question for the present study was whether observers would perceive ball speed in terms of their own abilities or their partners’ abilities. We grouped participants on the basis of whether they were better or worse at blocking the ball than their partner and analyzed perceived speed as a function of this grouping. Participants perceived ball speed in terms of their own abilities: Those who were relatively better blockers perceived the ball to be moving slower when observing their partners play than did those who were relatively worse blockers. In addition, when both sets of abilities were included as predictors, only the participant’s ability (and not their partner’s ability) significantly influenced speed estimates as the observer. This result clarifies our previous data (Witt et al., 2012), which were ambiguous as to which person’s abilities influenced apparent speed. Although this ambiguity was previously acknowledged (pp. 722–723), we leaned toward the conclusion that people perceive the environment in terms of others’ abilities. We now acknowledge that this claim is incorrect. Instead, the present data suggest that perceivers see the environment in terms of their own abilities to act as if they were in the other person’s situation.

These results have implications for the mechanisms underlying action-specific effects. Previously, motor simulation was proposed as one of these mechanisms (Witt & Proffitt, 2008), but a critical issue is that if motor simulation is the underlying mechanism, what exactly is being simulated? The literature describes two anatomically distinct simulation mechanisms (Waytz & Mitchell, 2011). One is a mirroring mechanism for which the other person’s actions, thoughts, and desires are directly simulated. The other is a self-projection mechanism by which the observer’s own actions, thoughts, and desires are simulated as if the observer is in the same situation as the other person. If motor simulation underlies action-specific effects when a perceiver is observing another, the present results suggest that this mechanism involves a self-projection style simulation rather than a mirror simulation. Waytz and Mitchell proposed unique neural pathways for the two types of simulation, so the present results also have implications for which brain areas are candidates for feeding information “back” to perceptual areas. First, however, neuroimaging data are needed to confirm the involvement of simulation. Nevertheless, the present results are important in determining the information relevant for perception regardless of the specific mechanisms that incorporate this information into perception. In particular, the results demonstrate that the relevant information is specific to the perceiver’s own abilities and not to the observed person’s abilities.

The present results preserve the idea that perception expresses the relationship between the perceiver and the environment, because it is the perceiver’s own abilities that are influential. Several accounts have proposed that action-specific effects are adaptive—specifically, for planning future actions (Cole, Balcetis, & Dunning, 2013; Proffitt, 2006; Witt, 2011a). Seeing a hill as steeper when he or she is fatigued or out of shape (cf. Bhalla & Proffitt, 1999), for example, could help motivate the perceiver to find an alternative route or to select a slower walking pace (Proffitt, 2006). Seeing a fearful object, such as a spider, as closer could help promote faster reactions and responses (Cole et al., 2013). In a hunting scenario, seeing a target as bigger or closer when shooting well (cf. Lee et al., 2012; Witt & Dorsch, 2009; Witt & Proffitt, 2005) could help promote the decision to shoot from one’s current location, as opposed to risking detection by moving closer (Witt, 2011a). All of these accounts depend implicitly, if not explicitly, on the idea that it is the perceiver’s own abilities that factor into perception. While this idea would have been challenged if another person’s abilities had penetrated perception, the present finding that it is the perceiver’s own abilities that are relevant preserves the idea that action-specific effects could be adaptive.

An unresolved issue is whether similar patterns will emerge when the observer has his or her own actions to perform, rather than just observing. If the observer performs an action concurrently with the actor, this will preoccupy simulation processes, so the observed actions of another may not influence perception. In addition, if two people are working together, they may take into account the other person’s abilities. Indeed, prior research has shown this to be the case. In one study, perceivers estimated the weight of a bag of golf balls (Doerrfeld, Sebanz, & Shiffrar, 2012). When they anticipated carrying the bag alone, they perceived the bag to be heavier than when they anticipated carrying the bag with the help of another person. However, if the other person appeared to be incapable of being much help (in this case, the person was on crutches), the bag looked heavier, similar to when the perceivers anticipated carrying the bag alone. Thus, the relationship between the perceiver’s goals and the other person’s abilities seems to modulate these effects. To our knowledge, no research to date has examined these effects in a competitive situation for which the other person’s abilities will have an inverse relationship to the perceiver’s own performance.

In summary, although action-specific effects are apparent when observing another person perform an action, it seems to be the perceiver’s own abilities that influence perception. Perceived speed was a function of the perceiver’s own abilities to block the ball, rather than the partner’s abilities. This finding is consistent with action-specific accounts of perception (Proffitt, 2006; Witt, 2011a), which assert that the world is perceived in terms of the perceiver’s ability to perform the intended action.