Towards and away from the body: The relevance of the direction of use in the coding of object-related actions

Abstract

This study examines whether the perception of an object automatically activates the representation of the direction of use of that object. To this aim, we carried out two experiments. In Experiment 1, participants were asked to explicitly categorise objects as used either away from the body (AB, for example, a hammer) or towards the body (TB, for example, a toothbrush). In Experiment 2, participants were asked to judge whether the same objects were natural or manmade. In both experiments, they were asked to respond by moving a joystick backwards (i.e., TB) or frontwards (i.e., AB). Therefore, their response could either be congruent (i.e., backward response with TB objects, frontward response with AB objects) or incongruent (i.e., backward response with AB objects, frontward response with TB objects) with the direction of object use. Results from Experiment 1 showed that in the congruent condition, participants were faster in judging the direction of object use than those in the incongruent condition (congruency effect). Crucially, results from Experiment 2 showed the presence of a congruency effect even when the direction of object use was task-irrelevant. However, this effect was found only for TB objects. These results suggest that the perception of TB objects automatically activates the direction of object use with respect to the body, thus showing evidence of direct connection between perception and action. A specific role of the body might account for different action representation processes involved in TB and AB object-related actions.

Keywords

Object use object semantic away from the body towards the body

Introduction

The ability to use objects efficiently is fundamental for most of our daily activities. Among others, this ability is enabled by two types of semantic knowledge: object function and manipulation (Buxbaum et al., 2000; Buxbaum & Saffran, 2002; Ruotolo, Kalénine, & Bartolo, 2020; for a review, see Osiurak & Badets, 2016). Object function knowledge refers to “what an object is used for,” whereas object manipulation knowledge refers to “how an object is manipulated.” The former belongs to conceptual knowledge (e.g., knowing that a hammer is used to hit a nail), whereas the latter is related to the expected sensorimotor consequences associated with object use (e.g., knowing that a hammer is used by holding it and performing oscillatory movements with the forearm). Function and manipulation information are closely intertwined in the action system (van Elk et al., 2014). Indeed, function knowledge is reinforced by typical object use (i.e., manipulation), while correct manipulation is guided by the purpose of object use (linked to its function). The process that connects functional (more abstract) and manipulation (sensorimotor) representations lies in the action goal, which is defined as “the spatial location towards which an action is directed” (van Elk et al., 2014, p. 223). In fact, the representation of the landing point of an action contributes to defining the intention to act (e.g., “I have to brush my teeth” or “I have to fix a nail on the wall”) and determines the features of the upcoming movement (i.e., how to manipulate and direct the chosen object, van Elk et al., 2014; see also Rosenbaum et al., 1992, 1995). Therefore, the representation of the action goal is of central importance in the transition from the intention to use an object to action planning. However, the cognitive processes underlying the representation of an action goal remain elusive.

To shed light on this topic, van Elk and colleagues (2008, 2009) asked participants in their study to observe pictures of actors holding objects with a correct/incorrect grip and near a correct/incorrect goal (e.g., microphone near mouth/ear). Selected objects were among those typically used towards the body (TB, for example, a microphone; “towards the face,” specifically) and away from the body (AB, for example, pincers). Participants had to judge whether the objects “would typically be moved toward the face” (i.e., explicit task) and whether they were made of plastic (i.e., implicit task). Their response consisted in keypresses executed with TB or AB movements. Results showed that participants were faster with stimuli showing a correct goal than with those showing a correct grip. These results indicate that when people observe how an object is used (hereafter “object-use action”), the action goal is processed even earlier than the object’s correct manipulation. Furthermore, participants were faster when the response movement was congruent with the direction of typical object use in the explicit task. However, this was not the case in the implicit task. According to the authors, this indicates that the action goal is not automatically retrieved when an object is seen, but it needs to be explicitly processed.

However, results from studies by van Elk and colleagues (2008, 2009) seem to contrast with the evidence that the mere view of an object automatically triggers the motor programme to interact with it efficiently, without any explicit intention to do so (e.g., Chainay & Humphreys, 2002; Ellis & Tucker, 2000; Tucker & Ellis, 1998). For example, it has been shown that participants are faster in judging generic semantic features of objects (e.g., “natural or manmade stimulus?”) when they are presented with the handle oriented towards the hand used to perform the task (orientation effect); or when providing the response with a power/precision grip that matches the grip required to grasp the presented object (size effect). Moreover, other studies have shown that participants are faster when the type of executed action is compatible with the functional use of the objects presented (e.g., a pointing gesture towards a calculator, as in the study by Bub et al., 2008 see also Bub & Masson, 2010; Masson et al., 2011). In sum, all these studies suggest that the direction of object use, that is, the action goal, might be implicitly processed once an object is seen.

As previously stated, van Elk and colleagues (2009) did not find any implicit representation of the direction of object use, which could be due to the fact that the objects were always presented as used by an actor and were not oriented towards the participants. We hypothesise that the actor’s presence might have affected the participants’ responses in both tasks. In fact, when participants had to judge the direction of use (i.e., whether moved towards the face), they could take the actor’s body as a reference point (i.e., towards the face of the actor). However, this information was not relevant in the implicit task where the objects’ visual features had to be judged. Here, the absence of an automatic activation of the direction of object use might stem from the impossibility to manipulate the object, owing to its orientation (i.e., oriented towards the actor), and/or from the visual cues regarding the direction of object use (i.e., the spatial relation between the object and the actor’s body; particularly crucial for objects that are used TB and AB). The presence of the actor combined with different task instructions might thus have promoted the emergence of a congruency effect only in the explicit task (for the effect of task instruction on participants’ responses, see Bartolo et al., 2020). The exclusion of the actor should thus allow us to control this aspect.

Therefore, the aim of this study was to verify whether the representation of the direction of use of the object is automatically activated when an object is recognised or whether it is necessary to represent the associated action explicitly. To this end, we re-adapted the experimental paradigm by van Elk et al. (2009) and conducted two experiments, each involving a semantic categorisation task, where the objects are presented alone. In Experiment 1 (explicit task), the direction of object use was task-relevant, meaning that it was explicitly coded (i.e., judging whether the object is used TB or AB). In Experiment 2 (implicit task), the direction of object use was task-irrelevant and could thus only be implicitly coded (i.e., judging whether the object was natural or manmade). Each experiment involved three-dimensional (3D) images depicting objects typically used TB (e.g., a toothbrush) and AB (e.g., a hammer), presented on a computer screen. Responses were provided by moving the vertical handle of a joystick AB or TB, leading to a congruent (e.g., AB response + AB object) or incongruent (e.g., TB response + AB object) condition. Response times (RTs) and accuracy were measured.

We expected that participants would be faster and more accurate in the congruent condition than in the incongruent condition. Moreover, if the direction of use of the objects required explicit processing, the congruency effect should then appear in the explicit (Experiment 1) but not in the implicit task (Experiment 2), as found by van Elk et al. (2009). Conversely, the presence of a congruency effect in the implicit task would support the idea that the direction of object use is automatically activated, even when it is task-irrelevant.

Experiment 1

The aim of Experiment 1 was to verify the presence of a congruency effect between an explicit coding of the direction of use of an object (i.e., TB or AB action) and the direction of the movement required to provide the answer (frontwards: FW; backwards: BW). Participants were required to represent the direction of object use intentionally by asking them to judge whether the presented objects were used TB or AB. Participants were divided into two groups (between-subject design): congruent versus incongruent. The congruent group provided judgements through a movement that was congruent with the direction of object use (e.g., BW movement for TB objects). The incongruent group provided judgements through a movement that was incongruent (e.g., BW movement for AB objects) with the direction of object use. In line with van Elk and colleagues (2009), we expected faster RTs in the congruent group than in the incongruent group.

Method

Participants

An a priori power analysis was conducted with G*Power, version 3.1.9.4 (Faul et al., 2009), to determine the appropriate sample size for each experiment. As we used a between-subject design in Experiment 1 (congruent, incongruent), we chose the following parameters: Cohen’s effect size f = 0.35 (between medium and large, van Elk et al., 2009), α = .05, Power = .90, number of groups = 2, number of measurements = 2 (objects: TB, AB), and correlation among measurements = 0.5. The minimum sample size was 24 participants. The final sample included 30 French right-handed healthy participants (18 women; age range: 18–30; M = 23.50; SD = 2.70).

Participants were assigned to 2 groups of 15 participants each, according to the direction of the response (see Figure 1). In the congruent group (12 women; mean age = 23.25, SD = 2.35), participants were instructed to respond backwards (BW) if they thought that the objects were used TB, and frontwards (FW) if they thought that the objects were used AB. In the incongruent group (7 women; mean age = 23.69, SD = 2.94), participants were instructed to respond frontwards (FW) if they thought that the objects were used TB, and backwards (BW) if they thought that the objects were used AB. Finally, natural (NAT) stimuli were included as distractors requiring no response, as in a go/no-go task.

Figure 1.

Experimental flow from Experiment 1.

All participants were recruited at the University of Lille (France). They had normal or corrected-to-normal vision. The experimental protocol was carried out in agreement with the Declaration of Helsinki (World Medical Association, 2013) and written informed consent was obtained from all participants.

Materials and stimuli

Stimuli consisted of images obtained from 3D models of objects of everyday use and natural elements, presented on a black background. Models were built by using the SketchUp software (version 17.1.174) and drawn from the Microsoft Office PowerPoint 3D images database. The colour and proportions of objects were set as matching those of real objects. There were 60 stimuli in total, divided between manmade (30) and natural (30) stimuli. Natural stimuli included 15 images of animals (e.g., fox, pig, bear; pets such as dogs and cats were not included) and 15 images of plants (e.g., tree, leaf, flower). Manmade stimuli were divided into two sub-categories of 15 stimuli each: towards the body (TB) objects, used with one hand and towards the body (e.g., toothbrush); away from the body (AB) objects, used with one hand and away from the body (e.g., a hammer). Stimuli used towards the body were a toothbrush, a comb, a hair-dyer, a baseball cap, an ice cream cone, a straw, a coffee mug, a cigarette, a cotton bud, eyebrow tweezers, a body thermometer, a microphone, a hairbrush, a canteen, and a lollipop. Away from the body objects were a hammer, a paint brush, a key, a light bulb, scissors, a stapler, a scraper, a TV remote control, a bottle of oil, a kettle, a spanner, a meat tenderiser, a screwdriver, a carafe, and pincers. Objects were all presented with the graspable part oriented for a right-handed use.

Task performance involved the use of a joystick (Thurstmaster T-Flight stick X) with a vertical handle to record participants’ responses and a laptop Lenovo V130 (screen size: 15.6″, 1,920 × 1,080 pixels). The handle of the joystick was 19 cm high, with an ergonomic grip located between 10 and 15 cm from the base and a diameter around 3.5 cm.

Procedure

Participants entered a quiet experimental room and were asked to sit on a chair placed 86 cm from the computer screen. The joystick was positioned 16 cm from the closest table edge to the participants. Chair, joystick, and laptop were all aligned with the participants’ medio-sagittal axis. Once seated, participants received written instructions that were also repeated verbally by the experimenter. The OpenSesame software (version 3.2.6) was used for stimuli presentation and data recording. To make sure that all stimuli were recognised during the task and that participants had no difficulties in retrieving the actions associated with the functional use of the objects, the latter were presented one at a time and participants were asked to indicate their name and function (e.g., “a toothbrush is used to brush one’s teeth”). For natural stimuli, they had to decide only whether the images represented plants or animals (indicating only the category). Once the naming phase was completed, the experiment started.

Participants had to perform an explicit semantic categorisation task. They were asked to judge the direction of use of the object (TB or AB). The response was provided by movements directed TB or AB, performed with the right hand. Exact instructions were as follows: “Do not respond when natural elements are presented. Move the joystick frontwards [or backwards] when an object used toward the body is displayed on the screen; in the other direction otherwise. Try to be as fast and accurate as you can.” In total, each group was submitted to 60 trials (30 trials with natural elements + 30 trials with manmade objects: 15 AB + 15 TB) presented in random order. Each trial started with a fixation cross (1 s) and then the stimulus was displayed on the screen. As soon as the stimulus appeared, participants had to respond as accurately and quickly as possible. The stimulus remained on the screen until participants provided their response, for a maximum of 3 s. After that response, a 1-s interstimulus interval (blank screen) appeared and another trial started (see Figure 2). RTs and accuracy were recorded.

Figure 2.

Experimental flow of one trial (Experiments 1 and 2). A fixation cross appeared for 1 s, followed by the stimulus presentation (time limit: 3 s). Participants were instructed to respond as soon as the stimulus (Natural or Manmade) appeared on screen, by performing the appropriate movement. After the participant’s response, a 1-s blank screen was presented before the beginning of the next trial. Only in Experiment 1, no response was required for Natural stimuli.

Results

Two separate 2 × 2 analyses of variance (ANOVAs) were carried out on accuracy and RTs. The Group was the between factor (congruent vs incongruent) and Object type was the within factor (TB vs AB). Normality of data distribution was verified by a skewness and kurtosis analysis, and equality of variance by a Levene’s test.

Accuracy

Data were normally distributed (skewness and kurtosis values across all variables: average skewness = -1.29 SD = .45 range: -1.61 to -.98; average kurtosis = 1.76 SD = 1.57 range: .64 to 2.87, Chou & Bentler, 1995). The equality of variance assumption was respected (Levene’s test, p > .05). No significant effect was found. In 93% of the trials, participants categorised TB objects as used TB and in 90% of trials, AB objects as used AB.

RTs

RTs above or below 3 SD were excluded (1.6% of total). Data were normally distributed (skewness and kurtosis values across all variables: average skewness = 0.39 SD = 0.007 range: 0.38 to 0.39; average kurtosis = −0.67 SD = 0.15 range: −0.78 to −0.56). The equality of variance assumption was respected (Levene’s test, p > .05). The 2 × 2 ANOVA showed a main effect of Group, F(1, 28) = 17.35, p < .001, $η_{p}^{2} = . 38$ , with the congruent group being faster (M = 1,026.91, SD = 155.24) than the incongruent one (M = 1,271.71, SD = 202.79).

A main effect of Object type was also observed: F(1, 28) = 14.94, p < .001, $η_{p}^{2} = . 34$ . The effect was due to RTs being faster with TB (M = 1,098.85, SD = 198.02) than with AB objects (M = 1,199.77, SD = 227.46) (p < .001) (see Figure 3). No interaction effect was found, F(1, 28) = 0.18, p = .67, ns. Furthermore, correlation analyses between accuracy and RT did not show any speed-accuracy trade-off.

Figure 3.

Comparison of mean RTs for each object category (Experiment 1): objects used towards body (TB) and away from body (AB). Error bars denote confidence intervals.

Discussion

In Experiment 1, participants had to judge the direction of use of objects (TB or AB) through movements carried out in a congruent direction or an incongruent direction (BW or FW) with that of object use. An explicit representation of the action associated with object use was thus required. Results showed a faster categorisation in the congruent group than in the incongruent group, in line with the results by van Elk and colleagues (2009), even though in the current study objects were presented alone. Moreover, results showed longer RTs with AB objects than with TB ones. This could be due to the different representational processes involved in AB and TB movements (Ruotolo, Iachini et al., 2020). TB objects require imaging the relationship between the object and the participant’s own body, whereas AB objects also require imaging the relationship between the object and another object in the external world (e.g., the hammer is used to hit a nail). This additional representation may have slowed down the processing of AB objects.

Nevertheless, these results do not allow us to conclude that action semantics about the direction of object use can automatically be activated. Besides, one could argue that the congruency effect in the explicit task is due to the use of abstract spatial coding, rather than due to the activation of motor processes (Phillips & Ward, 2002). In other words, it is possible that when participants recognise how the object is typically used (e.g., by moving it away from the body in the case of a hammer), the abstract concept of away/towards is activated. In turn, this explicit representation could be associated with the hand movement–directed AB/TB. To address these issues, a second experiment was carried out in which the representation of the direction of object use was implicitly induced.

Experiment 2

In Experiment 2, we tested whether the direction of object use (TB or AB) can be activated when an object is recognised, but the coding of the action associated with object use is not explicitly required (implicit task). To this end, in a within-subject design, participants were presented with images of manmade objects and natural elements and asked to judge whether these stimuli were natural or manmade. As in Experiment 1, responses were provided by movements executed in a congruent or incongruent direction (BW, FW) with that of object use. If the direction of object use was automatically activated, we expected faster RTs in the congruent condition than in the incongruent condition.

Method

Participants

Another sample of participants was recruited for Experiment 2. Sample size estimation was carried out with the following parameters: Cohen’s effect size f = 0.25, α = .05, Power = .90, number of groups = 1, number of measurements = 6 (three object types: TB, AB + Natural stimuli × 2 Movement directions), correlation among measurements = 0.5. The minimum sample size was 24 participants. The final sample comprised 26 right-handed healthy participants (16 women; age range: 18–30; M = 23.89; SD = 2.60).

Procedure

In Experiment 2, participants were asked to perform a semantic categorisation task: they had to judge whether each stimulus represented a natural or a manmade element. The experimental setting and stimuli were the same as in the previous experiment. However, unlike in Experiment 1, a within-subject design was applied and the presentation of the stimuli was organised in blocks. AB and TB objects were presented in separate blocks along with the natural stimuli, because natural objects do not provide any “toward the body”/“away from the body” information. Therefore, each block contained 15 objects (AB or TB) and 15 natural elements (a sub-group randomly selected each time from the 30 natural stimuli). A total of four blocks was obtained: blocks 1 and 2 contained the same 15 TB objects, but different natural stimuli; blocks 3 and 4 contained the same 15 AB objects, but different natural stimuli. Before each block, participants were instructed about how to respond: “Move the joystick frontwards [or backwards] when a manmade object is displayed on the screen; in the other direction otherwise. Try to be as fast and accurate as you can.” As a consequence, participants responded to the TB objects of one block with a movement congruent with the direction of object use (backwards in the congruent condition) and to the TB objects of the other block with an incongruent movement (frontwards in the incongruent condition). The same applied to the blocks containing AB objects. The order of blocks was counterbalanced across participants (see Figure 4).

Figure 4.

Example of experimental flow from Experiment 2.

Participants responded with the right hand to all 120 trials (60 trials per condition: congruent and incongruent). RTs and accuracy were recorded.

Results

Normality of data distribution was verified by a skewness and kurtosis analysis, and the assumption of sphericity by a Mauchly’s test.

Accuracy

Normality distribution was respected (average skewness: −1.92, SD = 1.15, range: −3.92 to −1.10; average kurtosis: −4.16, SD = 1.56, range: −0.85 to 16.49), except in one case (AB objects with BW responses). Non-parametric analyses were performed: a Friedman test for Object type (TB, AB, and NAT) and a Wilcoxon signed-rank test for Movement type (BW and FW). The Friedman test did not show any significant difference among the three object types, χ²(2, N = 26) = 1.94, p = .38, ns. The Wilcoxon signed-rank test did not show any difference between the two types of answers (z = 1.58, p = .11). As expected, participants had no difficulties in distinguishing between natural and manmade objects (Mdn = 98%).

RTs

Stimuli for which RTs were above or below 3 SD were excluded (3.6% of total). 3 × 2 repeated ANOVA measures were conducted on RTs with Object type (TB, AB, NAT) and Movement direction (BW, FW) as factors. Data were normally distributed (average skewness = 0.68, SD = 0.58, range: −0.10 to 1.37; average kurtosis = 1.81, SD = 1.93, range: −0.23 to 4.62). The sphericity assumption was violated for the Object type factor (Mauchly’s test, p < .05). In this case only, degrees of freedom were corrected using Greenhouse-Geisser estimates of sphericity. We found a main effect of Object type: F(1.57, 39.46) = 24.80, p < .001, $η_{p}^{2} = . 49$ . Tukey’s honestly significant difference (Tukey’s HSD) post hoc test showed that the effect was due to judgements for TB objects being slower (M = 659.06, SD = 101.98) than those for AB objects (M = 609.97, SD = 84.20) and NAT stimuli (M = 609.21, SD = 77.81) (p < .001). Results showed no significant difference between AB objects and NAT stimuli (Figure 5). The effect of Movement direction did not reach significance, F(1, 25) = 3.70, p = .07.

Figure 5.

Comparison of mean RTs for each stimulus category from Experiment 2 (NAT = natural stimuli) (**p < .001).

The interaction between object type and movement direction was also significant: F(2, 50) = 3.87, p < .05, $η_{p}^{2} = . 13$ . When participants responded to TB objects with a movement directed FW (incongruent condition, M = 676, SD = 108.84), they were slower than in all other conditions (all at least p < .05) (see Figure 6). Moreover, participants were slower responding to TB objects with a movement BW (congruent condition; M = 642.12, SD = 93.67) than responding to AB objects with a movement BW (incongruent condition; M = 608.86, SD = 83.01). No significant difference emerged between FW and BW responses, with AB objects. In sum, a significant difference between congruent and incongruent responses was observed for TB objects only, with faster responses in the congruent condition than in the incongruent condition (see Table 1 for details).

Figure 6.

Comparison of mean RTs for each stimulus type (TB, AB, NAT) by movement direction (backwards, frontwards) from Experiment 2 (*p < .05; **p < .001).

Table 1.

RTs from the Experiment 2 (implicit task).

	TB	AB	NAT
Mov. Backwards (BW)
M	642.12	608.86	613.17
SD	93.67	83.01	69.51
Mov. Frontwards (FW)
M	676	611.08	605.24
SD	108.84	87.01	86.52

RTs: response times; TB: towards the body; AB: away from the body; NAT: natural.

Discussion

In Experiment 2, participants had to judge whether the stimuli were natural or manmade, by using a movement that was congruent (congruent condition: FW response for an AB object; BW response for a TB object) or incongruent (incongruent condition: FW response for a TB object; BW response for an AB object) with the direction of object use. Crucially, judgements did not explicitly require the access to the semantic knowledge regarding the direction of object use, which therefore could be processed only implicitly.

Results showed that participants were faster in categorising the objects when the response movement was congruent rather than incongruent with the typical direction of object use. However, this was true only for TB objects. Therefore, contrary to the findings of van Elk and colleagues (2009), these results indicate that the direction of object use can be implicitly represented, at least for TB objects.

Finally, the fact that a congruency effect appeared only for TB objects excludes any notion of correspondence between abstract spatial codes (BW-TB; FW-AB). Instead, it is possible that the processing of TB objects activates motor processes centred on the body that are not activated during the processing of AB objects. This last point will be discussed in more detail in the general discussion.

General discussion

In this study, we aimed at finding whether the action goal of an object (the final goal towards which an action is directed) is automatically activated as soon as the object is recognised, or if that action goal needs to be explicitly represented. In two experiments (Experiments 1 and 2), participants were presented with objects typically used towards (TB) and away from the body (AB). In Experiment 1, the action goal of each object was explicitly retrieved by categorising it as typically used towards or away from the body (explicit condition), whereas in Experiment 2, the action goal could be coded only implicitly, because the objects were categorised as natural or manmade (implicit condition). Crucially, in both experiments, participants responded by using a movement that was either congruent or incongruent with the typical direction of object use (backward movement for TB and frontward movement for AB objects in the congruent trials; BW movement for AB and FW movement for TB objects in the incongruent trials).

Results in the explicit condition (Experiment 1) showed that AB objects were categorised more slowly than TB objects, whereas in the implicit condition (Experiment 2), AB objects were categorised faster than TB objects. More importantly, and in line with van Elk and colleagues (2009), results in the explicit condition showed that participants were faster in categorising both TB and AB objects with a movement that was congruent with the typical direction of object use. However, contrary to van Elk and colleagues (2009), a congruency effect was also observed in the implicit condition for TB objects only. The crucial difference between this study and that by van Elk and colleagues (2009) is that the objects were not shown as manipulated by an actor. They were presented alone and in a position that made them easily manipulable. This might have promoted the implicit activation of the action goal associated with object use. This result therefore expands previous findings by showing that the direction of use can automatically be processed when a TB object is recognised.

The differences observed between TB and AB objects in the two experimental conditions could be explained by the double role played by the body in the semantic representation of TB and AB objects. Indeed, when using TB objects, the body is both a stable reference frame used to represent the position of the object and the target towards which the action is directed. This double function seems to give the body a fundamental role in the representation of the action with TB objects, thus promoting the establishment of habitual patterns of action (Wong et al., 2016, 2017). For example, the use of a toothbrush requires the same pattern of movements towards a stable part of the body. Indeed, most TB object-related actions are specifically linked to “hygienic or nutritional activities” (Bartolo et al., 2019; Halsband et al., 2001). Therefore, the primacy of the body might facilitate the activation of motor properties linked to TB objects, even when not explicitly required by the task. Conversely, the use of AB objects requires a shift towards various spatial locations in the external environment as targets of the action. As a consequence, action planning with AB objects might reflect the habitual necessity of considering the multiple spatial relations between elements of the environment in which the action is performed. Ruotolo, Iachini, et al. (2020) suggested that actions with TB objects are more linked to motor imagery processes, whereas actions with AB objects additionally involve visuo-spatial imagery processes useful to represent objects in the external environment (see also Kosslyn et al., 2001; Sirigu & Duhamel, 2001). This higher cognitive load could explain why the time spent in categorising AB objects was longer than for TB objects, when the actions were overtly represented (Experiment 1). De facto, the role played by the usual context in which AB and TB objects are used is crucial.

Several studies have shown that the functional context in which objects are presented can modulate participants’ responsiveness. For example, Wokke and colleagues (2016) showed that participants responded faster to the appearance of objects typically used in the kitchen when they were presented in a functionally congruent context (a kitchen) rather than in a functionally incongruent context (a workshop). Moreover, Vastano et al. (2017) showed that participants were facilitated in judging the relationship between an object and the place/context where it is used when the object functionally matched the context (e.g., spatula-kitchen) rather than in a case of mismatch (e.g., spatula-workshop) (for similar results, see Borghi et al., 2012). In this study, body parts may have represented relevant contextual information for TB objects (e.g., comb and hair), whereas for AB objects, the contextual information was usually represented by another object (e.g., hammer and nail) or by the external environment (e.g., nail and wall). As body parts were directly available in the experimental environment, contextual information could be gathered by participants while performing the task for TB objects, whereas contextual information was not immediately available and needed to be retrieved for AB objects.

These considerations suggest that under explicit request (Experiment 1) and despite the absence of a relevant context for AB objects, a functional semantic representation was activated for both types of objects. This representation includes the multiple processes involved in action planning (representation of spatial relations, motor processes, and contextual information). All these aspects need time to develop, thus allowing the emergence of congruency effects for both AB and TB objects (Phillips & Ward, 2002).

What happens on an implicit level (Experiment 2)? Previous studies have shown that the activation of a sensorimotor representation takes precedence compared to a functional and/or contextual representation (Ruotolo, Kalénine, & Bartolo, 2020). The access to this representation might be more facilitated by the recognition of TB rather than AB objects. As a consequence, processing time is longer for TB than for AB objects in the implicit condition. These speculations seem to be supported by the various patterns of RTs observed in Experiments 1 and 2. In fact, participants took about twice the time to categorise objects in the explicit condition (~1,149.31 ms) than in the implicit condition (~634.52 ms). This indicates that more information is processed in the explicit condition. Similarly, the RT difference between TB (659.06 ms) and AB (609.97 ms) objects in the implicit condition may indicate that the recognition of TB objects reflects the activation of sensorimotor information, which is not automatically activated by the recognition of AB objects. This hypothesis needs to be addressed further in future studies.

In conclusion, results from this study suggest that, on an explicit level, both TB and AB objects may activate representations involving sensorimotor and possibly functional/contextual information. However, the presence of the body as a stable reference seems to ensure that the sensorimotor representations are activated at an implicit level, but only by TB objects. Results with AB objects support the dual-route theory of action by Yoon and Humphreys (2005). According to this model, action knowledge or semantic knowledge can be activated selectively and a congruency or incongruency effect appears only when action knowledge is accessed. TB objects suggest that a direct route between perception and action (cf. Riddoch et al., 1989) is activated by objects used towards the body, whose action goal can be implicitly activated as soon as such an object is recognised.

Footnotes

Acknowledgements

We thank all the participants for their contribution to this study. We also thank Emilie Cooke-Martageix for amending the English.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Angela Bartolo

References

Bartolo

Claisse

Gallo

Ott

Sampaio

Nandrino

J. L.

(2019). Gestures convey different physiological responses when performed toward and away from the body. Scientific Reports, 9(1), 1–10.

Bartolo

Della Sala

Cubelli

(2020). Effect of test instructions: The example of the pantomime production task. Brain and Cognition, 139, Article 105516.

Borghi

A. M.

Flumini

Natraj

Wheaton

L. A.

(2012). One hand, two objects: Emergence of affordance in contexts. Brain and Cognition, 80(1), 64–73.

Bub

D. N.

Masson

M. E.

(2010). Grasping beer mugs: On the dynamics of alignment effects induced by handled objects. Journal of Experimental Psychology: Human Perception and Performance, 36(2), 341–358.

Bub

D. N.

Masson

M. E.

Cree

G. S.

(2008). Evocation of functional and volumetric gestural knowledge by objects and words. Cognition, 106(1), 27–58.

Buxbaum

L. J.

Saffran

E. M.

(2002). Knowledge of object manipulation and object function: Dissociations in apraxic and nonapraxic subjects. Brain and Language, 82(2), 179–199.

Buxbaum

L. J.

Veramonti

Schwartz

M. F.

(2000). Function and manipulation tool knowledge in apraxia: Knowing “what for” but not “how.” Neurocase, 6(2), 83–97.

Chainay

Humphreys

G. W.

(2002). Privileged access to action for objects relative to words. Psychonomic Bulletin & Review, 9(2), 348–355.

Chou

C.-P.

Bentler

P. M.

(1995). Estimates and tests in structural equation modeling. In Hoyle

R. H.

(Ed.), Structural equation modeling: Concepts, issues, and applications (pp. 37–55). SAGE.

10.

Ellis

Tucker

(2000). Micro-affordance: The potentiation of components of action by seen objects. British Journal of Psychology, 91(4), 451–471.

11.

Faul

Erdfelder

Buchner

Lang

A. G.

(2009). Statistical power analyses using G* Power 3.1: Tests for correlation and regression analyses. Behavior Research Methods, 41(4), 1149–1160.

12.

Halsband

Schmitt

Weyers

Binkofski

Grützner

Freund

H. J.

(2001). Recognition and imitation of pantomimed motor acts after unilateral parietal and premotor lesions: A perspective on apraxia. Neuropsychologia, 39(2), 200–216.

13.

Kosslyn

S. M.

Thompson

W. L.

Wraga

Alpert

N. M.

(2001). Imagining rotation by endogenous versus exogenous forces: Distinct neural mechanisms. Neuroreport, 12(11), 2519–2525.

14.

Masson

M. E.

Bub

D. N.

Breuer

A. T.

(2011). Priming of reach and grasp actions by handled objects. Journal of Experimental Psychology: Human Perception and Performance, 37(5), 1470–1484.

15.

Osiurak

Badets

(2016). Tool use and affordance: Manipulation-based versus reasoning-based approaches. Psychological Review, 123(5), 534–568.

16.

Phillips

J. C.

Ward

(2002). SR correspondence effects of irrelevant visual affordance: Time course and specificity of response activation. Visual Cognition, 9(4–5), 540–558.

17.

Riddoch

M. J.

Humphreys

G. W.

Price

C. J.

(1989). Routes to action: Evidence from apraxia. Cognitive Neuropsychology, 6(5), 437–454.

18.

Rosenbaum

D. A.

Loukopoulos

L. D.

Meulenbroek

R. G.

Vaughan

Engelbrecht

S. E.

(1995). Planning reaches by evaluating stored postures. Psychological Review, 102(1), 28–67.

19.

Rosenbaum

D. A.

Vaughan

Barnes

H. J.

Jorgensen

M. J.

(1992). Time course of movement planning: Selection of handgrips for object manipulation. Journal of Experimental Psychology: Learning, Memory, and Cognition, 18(5), 1058–1073.

20.

Ruotolo

Iachini

Ruggiero

Scotto

Tella

Ott

Bartolo

(2020). The role of mental imagery in pantomimes of actions towards and away from the body. Psychological Research. Advance online publication. https://doi.org/10.1007/s00426-020-01359-4

21.

Ruotolo

Kalénine

Bartolo

(2020). Activation of manipulation and function knowledge during visual search for objects. Journal of Experimental Psychology: Human Perception and Performance, 46(1), 66–90.

22.

Sirigu

Duhamel

J. R.

(2001). Motor and visual imagery as two complementary but neurally dissociable mental processes. Journal of Cognitive Neuroscience, 13(7), 910–919.

23.

Tucker

Ellis

(1998). On the relations between seen objects and components of potential actions. Journal of Experimental Psychology: Human Perception and Performance, 24(3), 830–846.

24.

van Elk

van Schie

H. T.

Bekkering

. (2008). Conceptual knowledge for understanding other’s actions is organized primarily around action goals. Experimental Brain Research, 189(1), 99–107.

25.

van Elk

van Schie

H. T.

Bekkering

. (2009). Action semantic knowledge about objects is supported by functional motor activation. Journal of Experimental Psychology: Human Perception and Performance, 35(4), 1118–1128.

26.

van Elk

van Schie

H. T.

Bekkering

. (2014). Action semantics: A unifying conceptual framework for the selective use of multimodal and modality-specific object knowledge. Physics of Life Reviews, 11(2), 220–250.

27.

Vastano

Finn

Barnes-Holmes

(2017). Functional relations modulate the responsiveness to affordances despite the impact of conflicting stimulus-response mappings. Frontiers in Psychology, 8, Article 1951.

28.

Wokke

M. E.

Knot

S. L.

Fouad

Ridderinkhof

K. R.

(2016). Conflict in the kitchen: Contextual modulation of responsiveness to affordances. Consciousness and Cognition, 40, 141–146.

29.

Wong

A. L.

Goldsmith

Forrence

A. D.

Haith

A. M.

Krakauer

J. W.

(2017). Reaction times can reflect habits rather than computations. Elife, 6, Article e28075.

30.

Wong

A. L.

Goldsmith

Krakauer

J. W.

(2016). A motor planning stage represents the shape of upcoming movement trajectories. Journal of Neurophysiology, 116(2), 296–305.

31.

World Medical Association. (2013). World Medical Association Declaration of Helsinki: Ethical principles for medical research involving human subjects. Journal of the American Medical Association, 310(20), 2191–2194.

32.

Yoon

E. Y.

Humphreys

G. W.

(2005). Direct and indirect effects of action on object classification. Memory & Cognition, 33(7), 1131–1146.