The engaging nature of interactive gestures.

The social interactions that we experience from early infancy often involve actions that are not strictly instrumental but engage the recipient by eliciting a (complementary) response. Interactive gestures may have privileged access to our perceptual and motor systems either because of their intrinsically engaging nature or as a result of extensive social learning. We compared these two hypotheses in a series of behavioral experiments by presenting individuals with interactive gestures that call for motor responses to complement the interaction ('hand shaking', 'requesting', 'high-five') and with communicative gestures that are equally socially relevant and salient, but do not strictly require a response from the recipient ('Ok', 'Thumbs up', 'Peace'). By means of a spatial compatibility task, we measured the interfering power of these task-irrelevant stimuli on the behavioral responses of individuals asked to respond to a target. Across three experiments, our results showed that the interactive gestures impact on response selection and reduce spatial compatibility effects as compared to the communicative (non-interactive) gestures. Importantly, this effect was independent of the activation of specific social scripts that may interfere with response selection. Overall, our results show that interactive gestures have privileged access to our perceptual and motor systems, possibly because they entail an automatic preparation to respond that involuntary engages the motor system of the observers. We discuss the implications from a developmental and neurophysiological point of view.

Introduction

Gestures are pervasive in our everyday interactions. They are used to communicate and disambiguate meanings (deictic gestures like pointing, symbolic gestures, emblems), to clarify or emphasize discourse (gestures accompanying speech, iconic gestures), and to signify actions (pantomimes). Communicative gestures share a common neural substrate with language [1, 2, 3] and constitute precursors of language acquisition, both ontogenetically [4, 5, 6] and phylogenetically [7, 8, 9, 10].

Importantly, there is a specific subset of gestures that not only have the function of transmitting socially relevant information from a communicator to a receiver, but are also interactive, as they call for a specific response in the observer to complete a joint action. For instance, an open palm, depending on its orientation, may call for a hand-shake, a high-five, or a giving action. Here, we hypothesize that the key feature of interactive gestures is that they transfer social information via purely pre-verbal, non-symbolic features that maximally engage the recipient [11, 12]: this would predict that people perceiving interactive gestures may recruit specific cognitive processes. Interactive gestures may in fact be processed based on a sensorimotor coding that maps the representation of the observed action onto an interactive script that triggers a complementary response. This would imply that when observers perceive a gesture that is a component of hand-shaking, a hand-shaking script would be activated in the sensorimotor system of the observer and facilitate the performance of the complementary hand shaking action (see [13,14,15,16,17,18,19]). In other words, the perception of interactive gestures might entail social affordances.

We use the concept of affordance in the broad sense that is common in Cognitive Neuroscience studies of action. Traditionally, the term has been used to describe the action possibilities provided to an organism by the perceptual properties of the environment (e.g., of objects, [20]). Neuroimaging and lesion studies have corroborated the hypothesis that the perception of objects induces a behavioural response facilitation; it involves a left-lateralized fronto-parietal network responsible for the preparation of object-specific motor responses, which activates regardless of the intention to actually interact with the observed object [21, 22, 23, 24]. In a non-externalist notion of affordances [25, 26, 27], they can thus be considered not only as object properties, but also as sensorimotor patterns of brain activations that determine specific behavioural responses. With social affordances, then, we refer to the opportunities for social interactions potentially evoked by interactive gestures, such as the call for a complementary response that would complete a joint action.

Previous studies have used spatial compatibility tasks to investigate affordance-driven behavioural effects induced by the observation of interactive and communicative gestures. These tasks provide a measure of the automatic activation of responses that have a spatial overlap with the spatial feature of the stimuli [28]. When action representations are spontaneously activated by a visual stimulus, e.g., as a consequence of object affordance, the responses that are compatible with the evoked action are facilitated (e.g., a left response to a stimulus that is presented on the left side of the screen), while the responses that are incompatible with the evoked action show slower response times.

In these studies, interactive gestures have been found to have a facilitatory effect on the selection of a complementary action [29, 30, 31] as compared to non-social control stimuli (e.g, wooden hands, [29]), non-social intransitive gestures (e.g., a fist, [31], [29]), or non-social directional stimuli (e.g. arrows, [30]). For example, [29] report an unexpected reversed compatibility effect for an interactive stimulus (a hand-shake, Experiment 1), which primed a complementary and not an imitative response as compared to a non-social stimulus depicting the same gesture executed by a wooden hand. This effect disappeared if the stimulus was communicative but not interactive (an OK gesture, Experiment 2), but the two results were not directly compared. This incidental finding is in line with our prediction that interactive gestures may recruit specific sensorimotor processing routes. However, this and the other previous studies lacked control stimuli with comparable social relevance, or they averaged responses to interactive and communicative gestures, thus ignoring potential differences between these two types of stimuli. Thus, it is an open question whether processing interactive gestures really involves a different processing route as compared to other socially relevant stimuli.

Capitalizing on the previous literature, we aimed at directly comparing the response facilitation induced by Interactive gestures (‘hand shaking’, ‘requesting’, ‘high-five’) with the response facilitation induced by Communicative gestures (‘Ok’, ‘Thumbs up’, ‘Peace’) by using a spatial compatibility task. Participants were instructed to judge target letters based on their orientation (upright/inverted), while concomitantly observing irrelevant Interactive and Communicative gestures matched for salience (S1 File). We presented (inter)action-relevant (but task-irrelevant) gestures in a location that was compatible or incompatible with the responding hand and tested whether Communicative and Interactive gestures differently influence participants’ accuracy and response speed. We compared the size of the spatial compatibility effect (CE) between the two types of gestures, thus measuring the impact of Interactive vs. Communicative gestures on response selection.

We had the following predictions. First, Communicative gestures should induce a strong CE due to the perceptual salience that characterizes social stimuli, e.g., responding with the left hand should be faster when the Communicative gesture appears in the left visual hemi-field. Second, if Interactive gestures provide a social affordance, they should induce a smaller CE: indeed, the CE may be reduced by the call for a complementary response, e.g., a requesting gesture that appears on the left hemi-field may call for a giving gesture performed with the right hand. In three experiments, we investigated differences in the CE induced by Interactive as compared to Communicative gestures (Experiment 1) and whether they can be truly attributed to social affordance effects (Experiment 2 and 3).

Experiment 1

In the first experiment, we used a spatial compatibility task [32] to test whether Interactive gestures lead to a reversed or smaller CE than Communicative gestures [29]. This finding would indicate that interactive gestures trigger the representation of complementary responses, i.e., exert social affordances. We also investigated whether the predicted reduction of the CE in the Interactive gesture condition is restricted to the dominant hand, which is the effector normally used in a joint action (e.g., hand shaking with the right hand) or whether it generalizes to the non-dominant hand. Investigating this generalization is crucial to shed light on the processes underlying potential social affordance effects. Indeed, an effect restricted to the dominant hand would indicate that the processing of interactive gestures activates overlearned sensorimotor routes. On the contrary, the generalization of social affordance effects to the non-dominant hand may indicate that they do not depend on the activation of overlearned social scripts and may rather result from intrinsic perceptual features that call for a complementary response, independently of previous experience with the specific social script.

Methods of Experiment 1

Participants

We based our sample size on previous studies investigating the modulation of object affordance effects [33, 34], which found a moderate effect size of ηp2 = .21 for the interaction between spatial compatibility and the affordance effect. Using a 2x2 within-subject analysis of variance (ANOVA) the power analysis conducted in G*Power 3.1 [35] revealed that, with α = .05 and statistical power at 1–β = .90, we needed a sample size of N = 27. Thirty participants were recruited to take part in the experiment, of which one was excluded because he did not understand the task instructions, and a second one was excluded because his accuracy and RT data showed a high amount of outliers (see below, final sample 28 participants, 20 f, average age = 26.3 years, SD age = 4.85 years). All experiments in this study were approved by the United Ethical Review Committee for Research in Psychology (EPKEB). All participants reported to be right-handed and to have normal or corrected-to-normal vision. They signed prior informed consent and received monetary compensation. The study was performed in accordance with the Declaration of Helsinki and later amendments.

Stimuli and apparatus

The set of stimuli comprised full-colour pictures of interactive and communicative gestures. Each picture was taken to include the torso of the model forward facing (head excluded) and both arms, with only the hand performing the gesture being visible (see Fig 1). The torso occupied the center of the picture so that the hand performing the gesture was lateralised with respect to the model’s body and directed towards the observer. For each condition (Interactive and Communicative) there were three different gestures (see Fig 1A), performed by a male and female model. For each gesture, we created a right and left-hand version by mirroring the original picture to avoid any low-level perceptual differences (for a total of 24 stimuli: 3 communicative and 3 interactive gestures performed by a female or male actor and presented in right- and left-hand version). The body stimuli were 500 x 296 pixels, 2.5 cm in height and 4.2 cm in width and subtended 2.6° and 4.37° of visual angle at a viewing distance of 55 cm.

Fig 1

A. The images of the gestures used as distractor stimuli. B. The trial-time line. Both experimental stimuli and trial-timeline were identical in Experiment 1, 2 and 3.

To verify that stimuli were matched for salience, we ran a preliminary experiment on an independent sample of 15 participants using a go/no-go task. Results indicated no differences in the salience of Interactive vs. Communicative gestures (S1 File).

The experimental script was run and participants’ responses were recorded using MatLab 16b software running on a Dell Precision T5610 PC with a screen size of 24 inch and display resolution of 1920 x 1080 at 60Hz.

Procedure

Each trial started with a fixation cross presented in the center of the screen for 500 ms. The pre-cue stimulus (gesture) appeared in the center of the screen. The Target (a capital letter T, Sans Serif font, size = 18, either upright or inverted) was presented positioned on the model’s hand at 90 pixels from the center and at 170 pixels from the top of the image. After a stimulus onset asynchrony (SOA) that randomly varied from 300–400 ms from trial to trial, the Target appeared. The time-interval of the SOA was chosen based on previous findings suggesting that 300 ms is the minimum time required for object affordances to modulate the activation of the motor system [36, 33, 37]. Participants were asked to indicate whether the Target was upright or inverted by pressing one of the two assigned keys with either their left or right index finger (see Fig 1B). The target was displayed until participants had responded, for a maximum of 1.5 s; if a response was not detected within 1.5 s the script proceeded to the next trial. After an inter-trial interval (ITI) that varied randomly between 1500–1900 ms after the response, the next trial started. The keyboard was centered on the computer screen, so that the response keys were lateralized with respect to the stimuli presented. The assigned keys were key A (left side of the keyboard) for left hand responses, and key L (right side of the keyboard) for right hand responses.

Experimental design

Participants completed 6 experimental blocks of 48 trials each. There were 144 spatially compatible trials and 144 spatially incompatible trials. Spatial compatibility was coded with respect to the combination between the (left/right) hemi-field of the screen where the target letter and the distractor (the hand gesture) appeared and the required (left/right) hand response, which depended on the target orientation (upright/inverted). The distractor and target position were on the same side of the correct hand response (compatible trials) 50% of the time and on the opposite side (incompatible trials) 50%. The association between Target orientation (upright/inverted) and response (left/right) was counterbalanced across participants. Half of the compatible and incompatible trials displayed interactive gestures and half displayed communicative gestures as distractors. The three different gestures in each condition, performed by female and male actors, appeared with equal frequency across trials and were positioned an equal number of times to the right or left side of the screen. Trial order was randomized within and across blocks. Each block took about 4 minutes, leading to 25 minutes to complete the experiment.

Data analysis

We measured Accuracy (Acc), i.e., the proportion of correct responses, and Response Times (RTs), i.e., the time delay between the instant when the target letter appeared on the screen and the participants' response, measured on correct trials only. Overall, participants were highly accurate: a response was not detected in 0.73% of trials, equal to 61 trials in the whole sample, and errors were equal to 2.79% of the trials (233 in the whole sample). We planned to exclude participants showing outlier values in both the individual mean Acc and individual mean RTs, as identified by the Box and Whisker Plot. One participant was excluded from further analysis according to this criterion in Experiment 1.

For illustrative purposes, we report in Table 1A raw Acc and RT data: here, we calculated the individual mean Acc and RTs for each condition, excluding from the analysis of RTs any outlier values that fell 2.5 SDs above or below the individual mean of each experimental condition.

Table 1

Raw group means of Acc and RTs in each experimental condition for the three experiments.

	Response Times (ms)		Accuracy
Experiment 1	mean	standard deviation	mean	standard deviation
Interactive Compatible	522	68	0.98	0.03
Interactive Incompatible	536	62	0.96	0.04
Communicative Compatible	516	66	0.97	0.02
Communicative Incompatible	542	72	0.96	0.04
Experiment 2
Interactive Compatible	531	82	0.97	0.04
Interactive Incompatible	556	85	0.98	0.05
Communicative Compatible	527	81	0.97	0.04
Communicative Incompatible	567	87	0.98	0.04
Experiment 3
Interactive Compatible	550	69	0.98	0.03
Interactive Incompatible	570	71	0.97	0.04
Communicative Compatible	542	66	0.97	0.04
Communicative Incompatible	587	78	0.97	0.04

Data were analyzed in the statistical programming environment R (R 3.3.3, R Core Team 2014). For the analysis of Acc, generalized linear mixed effects models were used [38, 39]. As Acc is a binary dependent variable, it was submitted to a series of logistic mixed effects regressions using GLMER procedure in “lme4” R package (version 1.1–5, [40]). RTs were analyzed as a continuous dependent variable using linear mixed effects models, fitted using the LMER function in “lme4” R package (version 1.1–15, [40]). In both analyses, the inclusion of fixed effects in the best fitting model was tested with a series of likelihood ratio tests, including only the fixed effects that significantly increased the model’s goodness of fit [41] (see ). Only the results of the best fitting model are reported.

We considered as fixed effects spatial Compatibility (Compatible vs. Incompatible), Gesture-type (Interactive vs. Communicative), and Response-side (Right hand/Left hand), and their interactions. Concerning the random effect structure, by-subjects and by-stimulus-type (S3 Table) random intercepts were included to account for between-subjects and between-stimuli variability (S2 Table). We report here only the parameters of the best fitting model. In the analysis of RTs, we also applied a model criticism procedure to the best fitting model to exclude outlier trials as recommended by [38]. Statistics of the fixed effects of the best fitting model were estimated with the “lmerTest” R package (version 3.0–1, [42]). We report a summary of the fixed effects of the best-fitting models for each variable; for RTs, significance levels are based on Satterthwaite’s degrees of freedom approximation. When appropriate, the post-hoc direct contrasts between the single levels of the significant interactions and main effects were conducted on the best fitting model with the “phia” R package (version 0.2–1, [43]), applying Bonferroni correction for multiple comparisons. All tests of significance were based upon an α level of 0.05.

Results of Experiment 1

Accuracy

The best fitting model only included spatial Compatibility as fixed effect (S2A Table). The results showed a significant main effect of spatial Compatibility (Wald Z = -6.21, p < 0.001) indicating that participants were more accurate on Compatible than on Incompatible trials (adjusted (adj) mean Compatible trials 0.99, SE 0.23; adj mean Incompatible trials 0.97, SE 0.21).

Response times

The best fitting model included spatial Compatibility, Gesture-type and Response-side as fixed effects (S2A Table). 3.21% of the trials were excluded from further analysis as outliers (251 trials in the whole dataset). The results showed a significant main effect of spatial Compatibility (F(1,7516.5) = 117.06, p < .001) and Response-side (F(1, 7516.3) = 25.46, p < .001), while the main effect of Gesture-type was not significant (F(1, 10) = 0.15, p = .71). These effects indicate that responses on Compatible trials (adj mean 511.90 ms, SE 10 ms) were faster than on Incompatible ones (adj mean 538.14 ms, SE 10 ms), and responses with the Right hand (adj mean 518.90 ms, SE 10 ms) were faster than with the Left hand (adj mean 531.14 ms, SE 10 ms). Crucially, the results showed a significant Gesture-type x spatial Compatibility interaction (F(1, 7516.5) = 6.85, p = .009), indicating that the spatial Compatibility effect (CE) for Interactive gestures (Interactive-Compatible, adj mean 516.06 ms, vs. Interactive-Incompatible, adj mean 535.95 ms, p < .001) was smaller than for Communicative gestures (Communicative-Compatible, adj mean 507.74 ms, vs. Communicative-Incompatible, adj mean 540.32 ms, p < .001).

To directly compare the size of the CE between Interactive and Communicative gestures, we also computed an index of the CE for each participant (RT Incompatible–RT Compatible) separately for Gesture-type. A Dependent Sample t-test revealed a significant difference in the CE between Communicative and Interactive gestures (t(27) = -8.86, p < 0.001, d = -1.675), with the effect being bigger for Communicative gestures (mean 17 ms, sd 6 ms) than for Interactive gestures (mean 9 ms, sd 2 ms) (Fig 2).

Fig 2

The figure illustrates the stimulus-type by spatial compatibility interaction effect (CE) that was observed in both Experiment 1 and 2.

Notably, the CE is smaller for Interactive compared to Communicative gestures. Grey lines indicate single-subject values and black thick lines indicate the group means and standard deviations.

Discussion of Experiment 1

In line with our predictions, there was a significant interaction between Gesture-type and spatial Compatibility, indicating that the Interactive gestures led to a reduced CE as compared to Communicative gestures. The reduction of the CE for Interactive vs. Communicative gestures occurred for both the dominant and non-dominant hand. Overall, these results suggest that, in the Interactive gesture condition, the CE induced by the perceptual salience of the social stimuli might be counter-acted by the call for a complementary response, The findings indicate that this effect is not a mere effect of extensive social motor learning as it occurs both in the dominant and non-dominant hand.

Experiment 2

To further assess whether the reduction of CE in the Interactive gesture condition might be ascribed to a social affordance effect, we asked participants to respond to the target letter with the index and middle finger of their right hand rather than with their left and right hand [44]. Indeed, proper affordance effects trigger a response in an effector-specific way and are thus expected to disappear in intra-manual spatial compatibility tasks [44]. On the contrary, if interactive gestures induce a reduction in the CE in an effector-unspecific way, we expect the reduction in CE to be present even when responses are mapped onto fingers of the same hand: this result would suggest that interactive gestures trigger an automatic preparation to respond that precedes the selection of a specific action program. Finally, in Experiment 2 one might expect a general facilitation for interactive gestures presented in the left hemi-field, i.e., the position that should trigger a response with the right hand used to respond in both Compatible and Incompatible trials: as a consequence, in Experiment 2 the visual hemi-field (left or right) where stimuli appeared was tested as possible predictor of performance in the analyses.

Methods of Experiment 2

Thirty individuals (18 female, average age = 27.03 years, SD age = 4.79 years) participated in Experiment 2. All participants reported to be right-handed and to have normal or corrected-to-normal vision. They signed prior informed consent and received monetary compensation. The study was performed in accordance with the Declaration of Helsinki and later amendments.

The procedure was identical to Experiment 1, except that participants were asked to respond to letter orientation with the index and the middle finger of the right hand. The assigned keys were key J for left side responses with the index finger, and the key L for right side responses with the middle finger.

The experimental design was the same as in Experiment 1.

Data analyses were the same as in Experiment 1, with the only difference that we tested the effect of Stimulus-hemifield instead of Response-side. In Experiment 2, the factor Response-side was not tested as possible predictor as we did not expect any difference between the fingers of the same hand. We instead included the factor Stimulus-hemifield in the analysis to investigate whether the side where stimuli appeared could impact on participants' accuracy and response times.

Overall, participants were highly accurate: a response was not detected in 0.66% of trials, equal to 57 trials in the whole sample, and errors were equal to 2.35% of the trials (203 trials in the whole sample). We planned to exclude participants showing outlier values in both the individual mean Acc and individual mean RTs, as identified by the Box and Whisker plot. No participant was excluded according to this criterion in Experiment 2. Raw Acc and RTs data are reported in Table 1 and were calculated as described for Experiment 1.

Results of Experiment 2

The best fitting model only included spatial Compatibility as fixed effect (S2B Table). The results showed a significant main effect of spatial Compatibility (Wald Z = -8.25, p < 0.001), indicating that participants were more accurate on Compatible than on Incompatible trials (adj mean Compatible trials 0.99, SE 0.24; adj mean Incompatible trials 0.97, SE 0.20).

The best fitting model included spatial Compatibility and Gesture-type, but not Stimulus-hemifield, as fixed effects (S2B Table). 2.90% of the trials were excluded from further analysis as outliers (242 trials in the whole dataset). The results showed a significant main effect of spatial Compatibility (F(1,8095.5) = 209.48, p < .001), while the main effect of Gesture-type was not significant (F(1, 10) = 0.42, p = .53). These effects indicate that responses on Compatible trials (adj mean 526.92 ms, SE 13 ms) were faster than on Incompatible ones (adj mean 558.96 ms, SE 13 ms). Crucially, the results showed a significant Gesture-type x spatial Compatibility interaction (F(1, 8095.4) = 12.27, p < .001), indicating that the spatial Compatibility effect (CE) for Interactive gestures (Interactive-Compatible, adj mean 529.29 ms, vs. Interactive-Incompatible, adj mean 553.58 ms, p < .001) was smaller than for Communicative gestures (Communicative-Compatible, adj mean 524.57 ms, vs. Communicative-Incompatible, adj mean 564.37 ms, p < .001).

To directly compare the size of the CE between Gesture-types, we computed an index of the CE for each participant (RT Incompatible–RT Compatible) for Interactive and Communicative gestures. A Dependent Sample t-test revealed a significant difference in the CE between Communicative and Interactive gestures (t(29) = -2.99, p = 0.006, d = -0.547), with the effect being bigger for Communicative gestures (mean 39 ms, sd 26 ms) than for Interactive gestures (mean 25 ms, sd 24 ms) (Fig 2).

Discussion of Experiment 2

In Experiment 2 we replicated evidence of a specific reduction in the CE for Interactive as compared to Communicative gestures. This reduction occurred even when the stimulus-response mapping was intra-manual, and for stimuli presented in both visual hemi-fields: this supports the hypothesis that the reduction of CE for Interactive gestures occurs at a processing level at which the effector performing the possible complementary response is not yet specified. We thus suggest that Interactive gestures might trigger an automatic preparation to act that does not directly translate into a social affordance response, as it does not (only) depend on the activation of specific motor scripts. Importantly, as we found no effect of Stimulus-hemifield, we suggest that performing the task with the effector that is usually involved in the execution of interactive actions (i.e., the right dominant hand) cannot solely account for the pattern of results. To further test this latter point, we aimed to replicate our results in a third experiment where the intra-manual task was performed entirely with the left hand.

Experiment 3

We designed Experiment 3 to show that the reduction of CE we observed in Experiment 2 was not specific to the right dominant hand, i.e. the effector usually involved in the preparation of the complementary response to an interactive gesture. We thus asked participants to perform the same task as described in Experiment 2 by responding with the index and middle finger of their left hand. If interactive gestures induce a reduction of CE as a result of triggering automatic preparation to respond that occurs at higher levels of motor planning (where the specific movement is not detailed yet), we expect the reduction in CE to be present even when responses are mapped onto fingers of the left hand.

Methods of Experiment 3

We conducted a power analysis in G*Power 3.1 [35] to determine the sample size required to detect the observed effect size of the Compatibility x Stimulus type second order interaction (i.e., the interaction between compatibility and type of gesture) of ηp 2 = .25 using a 2x2 within-subject analysis of variance (ANOVA). The analysis revealed that, with α = .05 and statistical power at 1–β = .80, we needed a sample size of N = 19. Twenty individuals (14 female, average age = 22.7 years, SD age = 4.36 years) thus participated in Experiment 3. We collected data from one additional participant (total N = 20) to equally balance the stimulus–response mappings between participants. All participants reported to be right-handed and to have normal or corrected-to-normal vision. They signed prior informed consent and received monetary compensation. The study was performed in accordance with the Declaration of Helsinki and later amendments.

The procedure was identical to Experiment 2, except for the fact that participants were asked to perform the task by responding with two fingers of the left hand. Participants were asked to indicate whether the Target was upright or inverted by pressing one of the two assigned keys with their left hand. The assigned keys were key A (middle finger of the left hand) for left side responses, and key D (index finger of the left hand) for right side responses.

The experimental design was the same as in Experiment 2.

Data analyses

Data analyses were the same as in Experiment 2. Overall, participants were highly accurate: a response was not detected in 0.85% of trials, equal to 49 trials in the whole sample, and errors occurred on 2.71% of the trials, equal to 156 trials in the whole sample. We planned to exclude participants showing outlier values in both the individual mean Acc and individual mean RTs, as identified by the Box and Whisker plot. No participant was excluded according to this criterion in Experiment 3.

Raw Acc and RTs data are reported in Table 1, calculated as described for Experiment 1.

Results of Experiment 3

The best fitting model only included spatial Compatibility as fixed effect (S2C Table). The results showed a significant main effect of spatial Compatibility (Wald Z = -5.54, p < 0.001), indicating that participants were more accurate on Compatible than on Incompatible trials (adj mean Compatible trials 0.99, SE 0.28; adj mean Incompatible trials 0.97, SE 0.26).

The best fitting model included spatial Compatibility and Gesture-type, but not Stimulus-hemifield, as fixed effects (S2C Table). 3.00% of the trials were excluded from further analysis as outliers (167 trials in the whole dataset). The results showed a significant main effect of spatial Compatibility (F(1,5355.2) = 114.39, p < .001), while the main effect of Gesture-type was not significant (F(1, 10) = 0.66, p = .43). These effects indicate that responses on Compatible trials (adj mean 544.07 ms, SE 14 ms) were faster than on Incompatible ones (adj mean 576.89 ms, SE 14 ms). Crucially, the results showed a significant Gesture-type x spatial Compatibility interaction (F(1, 5355.2) = 21.31, p < .001), indicating that the CE for Interactive gestures (Interactive-Compatible, adj mean 547.74 ms, vs. Interactive-Incompatible, adj mean 566.37 ms) was smaller than for Communicative gestures (Communicative-Compatible, adj mean 540.41 ms, vs. Communicative-Incompatible, adj mean 587.40 ms).

To directly compare the size of the CE between Gesture-types, we computed an index of the CE for each participant (RT Incompatible–RT Compatible) for Interactive and Communicative gestures. A Dependent Sample t-test revealed a significant difference in the CE between Communicative and Interactive gestures (t(19) = -6.0, p < 0.001, d = -1.342), with the effect being bigger for Communicative (mean = 42 ms, sd = 13 ms) than for Interactive gestures (mean 18 ms, sd 6 ms).

Discussion of Experiment 3

Experiment 3 provided evidence that interactive gestures induce a reduction of compatibility effect also in participants performing the task with the left hand. This confirms that perceiving interactive gestures interferes with automatic response selection at a more abstract level of spatial mapping that does not concern the effector used to perform the complementary actions, possibly as a consequence of an automatic (and not yet specified) preparation to respond.

General discussion

Understanding gestures performed by our conspecifics is fundamental for the development of our social life, as it enables us to timely and successfully engage in interactions, develop and master language, attribute intentions, and take part in collaborative and cultural activities [10]. Previous research on gestures has mainly focused on the shared cognitive processes underlying the understanding of symbolic gestures and their overlap with other forms of symbolic communication, such as language [1]. Among socially relevant gestures, those that are also interactive, calling for a specific response in the observer to complete a joint action, might constitute privileged stimuli for our perceptual and motor system. Indeed, they potentially engage the observer in actively taking part in a social interaction. In the present study, we hypothesized that interactive gestures might be processed differently than purely communicative gestures for two possible reasons, either because of their intrinsically engaging nature, or as a result of extensive social learning.

We designed a series of behavioural experiments to test and disentangle these two possibilities by investigating the processing of Interactive vs. Communicative gestures in a spatial compatibility task, where we measured how response selection was modulated by the (task-irrelevant) gestures presented in a compatible or incompatible position with respect to participants’ response hand.

The results of the first experiment show that the perception of Interactive gestures leads to a reduced spatial compatibility effect (CE) compared to Communicative gestures. This result supports the hypothesis that different cognitive processes mediate the perception of Interactive and Communicative gestures. As stimuli in the two categories were matched for perceptual salience (S1 File) and social relevance, we can rule out low-level explanations for our pattern of results. The effect we found is in line with evidence of "interference" effects generated by perceived interactive gestures on the execution of pre-planned non-interactive movements [45], and it may seem to be driven by the activation of an interactive and complementary action script (i.e. a social affordance effect). However, the results of the first experiment provide indication that this modulation can be observed when responding with both the dominant and non-dominant hand. This raises the possibility that the effect we observe does not originate from an affordance effect due to extensive social motor learning of a complementary response (e.g. learning to shake hands with the right-dominant hand), but instead relies on an automatic motor preparation preceding the selection of a specific action program. This would suggest that, rather than affording a response to complement a specific joint action script, interactive gestures might produce a more generic motor engagement that prepare us to (inter)act.

The results of our second and third experiments support this latter interpretation. Indeed, they replicate the pattern of results of the first experiment and indicate that the selective reduction of CE for Interactive gestures occurred even when the stimulus-response mapping was intra-manual, when stimuli were presented in both visual hemi-fields, and regardless of whether participants performed the task with their right or left hand. Altogether, the pattern of results of these two additional experiments may seem to be incompatible with the notion of affordance-based motor preparation; however, it has to be noted that our stimuli set was composed of three different Interactive gestures (all requiring dominant hand responses), therefore it is possible that we observed a modulation resulting from the preparation of multiple motor plans, all relevant and coherent to the target goal of completing a joint action. Recent affordance models argue that the process of selecting a motor plan may occur simultaneously for multiple actions relevant to the target via attentional mechanisms (i.e. affordance competition hypothesis, [46]). If multiple motor plans are activated concurrently by the same perceived affordance, it might not be possible to observe a behavioral facilitation or modulation that is specific to any of the motor plans in the very early stages of processing. Furthermore, there is evidence that affordance-like effects are modulated by (other than motor) cognitive processes such as attention allocation [37], and that effector selection and action specification may emerge at late stages of action planning [36, 47, 48].

We suggest that the most parsimonious interpretation of our results is that interactive gestures might directly engage the observer and produce a readiness to interact, which does not (yet) result in the preparation of the specific complementary response. At a neural level, this might entail the automatic recruitment of premotor neural resources involved in response preparation that precede the selection of the effector used to act. By analogy with object affordance, these resources might include the anterior pre-supplementary motor cortex (pre-SMA) that typically codes a yet unspecified readiness for action associated with the perception of affordable targets [49]. These properties of pre-SMA have been widely described by studies in non-human primates [50, 51] and seem to also play an important role in social learning [52, 53]. More broadly, the results of our experiments clearly indicate that Interactive gestures are processed via a different route as compared to equally salient and socially relevant gestures that do not have an affordance for motor interaction. We speculate that, while being the recipient of Communicative gestures that carry symbolic meaning activates semantic processing routes [54], the encoding of Interactive gestures might instead rely on preverbal, non-symbolic/semantic mechanisms associated with earlier developing and more direct processing routes. Therefore, in addition to the recruitment of pre-SMA, the engaging nature of interactive gesture may also depend on the involvement of the fronto-striatal system, which plays a crucial role when people perceive that others are responsive to their social cues [11, 55].

The capacity of detecting possibilities for joint actions with conspecifics is fundamental to develop social and coordination skills [56] and fast discern actions that might require a social response. As such, this ability might be early acquired while infants take part in interactions mediated by some form of infant-directed interactive gesture, allowing them to parse interactive gestures as part of social interaction scripts (i.e. the action script of giving and requesting, see for instance [57]). The results of the present study indicate that the involuntary perception of interactive gestures influences the participants’ responses, as if our perceptual system was equipped with the ability to fast identify opportunity for interactions independently of the actual engagement in a social exchange: these findings pave the way for future investigations to address when the "perceptual advantage" of interactive gestures emerges and what social abilities it might require.

Average response times and accuracy per stimulus at the Go/No-Go task.

(DOCX)

Click here for additional data file.

Model selection for the analysis of accuracy and reaction times in each experiment.

(DOCX)

Click here for additional data file.

The stimulus-type variable that was used as random intercept to account for the between-stimuli variability.

(DOCX)

Click here for additional data file.

(XLSX)

Click here for additional data file.

(XLSX)

Click here for additional data file.

(XLSX)

Click here for additional data file.

(DOCX)

Click here for additional data file.

18 Mar 2020

PONE-D-20-02355

The engaging nature of interactive gestures

PLOS ONE

Dear Dr. curioni,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

As noted by both reviewers, this is a very well-written manuscript and the findings are an important contribution to the current literature on interactive gestures. Please make the suggested changes, including the additional statistical analyses as suggested by Reviewer #2. Also, please incorporate the citations that both reviewers suggest.

We would appreciate receiving your revised manuscript by May 02 2020 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Julie Jeannette Gros-Louis, PhD

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements:

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at http://www.plosone.org/attachments/PLOSOne_formatting_sample_main_body.pdf and http://www.plosone.org/attachments/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For more information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially sensitive information, data are owned by a third-party organization, etc.) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

3. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: This extremely well-written ms. describes an elegant suite of studies that explore the engaging nature of "interactive" rather than "communicative" actions, i.e. actions which require activity on the part of the human observer drawing him/her in to interact. By cleverly combining this with a spatial compatibility task, the interference of such stimuli was objectively assessed. Across different experiments it was shown that interactive gestures impact response selection and spatial compatibility as compared to communicative actions. The authors interprete this as evidence for a priviliged access of interactive gestures to perceptual and action systems.

I applaud the authors to systematically addressing this subtle, but important distinction! I am very impressed by the choice of stimuli and the methodological rigor with which the studies were conducted. I strongly recommend publication of this ms. and only have minor comments:

1) A study by von der Lühe et al. (2016) also investigated the impact of "communicative" actions. The authors may want to make reference to this study and could also use this opportunity to briefly discuss whether interactive gestures might bring about different expectations/predictions in the observer and how this might play out in a predictive coding scheme.

2) Schilbach and colleagues published several papers (2012, 2013), in which they investigated the impact of social gaze cues on spatial compatibility effects. This papers appear to be missing in the bibliography.

3) Finally, it might be tempting to speculate how interactive gestures might be important (motivationally and otherwise) for matters of mental health. It has been suggested that psychiatric disorders could be construed as "disorders of social interaction", where interactive gestures might also figure prominently (Schilbach 2016). Maybe the authors might want to speculate (or not).

Reviewer #2: Curioni and co-authors report an interesting study in which, in three behavioral experiments, they tested the difference between two types of social actions, namely interactive and communicative gestures, in a spatial compatibility task. By measuring the interference effect of observing lateralized task-irrelevant gestures when participants were requested to respond to target stimuli, they showed a reduced spatial compatibility effect in reaction times for interactive gestures (which were chosen to elicit a complementary response in the observers) as compared to communicative gestures, irrespective of the responding hand. Response accuracy showed that participants were more accurate when the position of the distractor and target was on the same side, regardless of the type of stimulus (interactive, communicative).

The manuscript is well written and provides an interesting extension to the available literature. The experimental design is sound and the data were rigorously analyzed.

I have few queries that the Authors should address to improve the clarity of the manuscript.

1) The first point regards the discussion of the obtained results with reference to the potential difference in salience for interactive gestures, given their intrinsically engaging nature, as compared to communicative gestures.

The results of all three experiments consistently show for RTs that interactive gestures lead to a reduction of the spatial compatibility effect as compared to communicative gestures. Therefore, it emerged that for interactive gestures, incompatible trials (distractor and targets on opposite sides) were less incompatible/interfering than for communicative gestures. Given that this effect was not limited to the dominant hand which would have been recruited to perform a complementary response to the presented interactive gesture, the Authors conclude that this could seemingly be a generalized effect. Thus, the effect would not reflect any specification of the required motor plan, but be related to a general motor engagement to socially interact which is activated when we are faced with interactive gestures. This unspecific effect for interactive gestures could be dependent on differences in the adopted types of stimuli, however the Authors report to have controlled for this in a go/no-go control task reported in the supplementary materials (e.g., p.18, line 423: “stimuli in the two categories were matched for perceptual salience and social relevance”). Nonetheless, it is unclear to me how the adopted go/no-go task can disentangle the issue of difference in salience/social relevance between communicative and interactive gestures.

I would then ask the Authors to comment on this aspect as it is relevant for the interpretation of the obtained results. In addition, please better report both in the supplementary materials and in the main text the rationale of having used this type of control task for the validation of the experimental stimuli and to control for perceptual salience and social relevance.

2) In addition, the statistical analysis of the go/no-go task included only the Stimulus number as within-subject factor, however it would be relevant to test the role of Gesture-type to be able to differentiate between the two types of gestures. Plus, in the control task the results for the accuracy are missing, while only the ANOVA results for RTs are reported. Please, add these missing information to the results paragraph (see supplementary materials).

3) Results: please report a measure of effect size in the results.

In relation to this, is the ηp 2 = .25 reported in the sample size calculation (pp.14-15, lines 350-355) as the observed effect size of the Compatibility x Stimulus type second order interaction referring to results of Experiment 1 or 2?

4) Discussion: When discussing the obtained results in the Discussion paragraphs, please specify that they refer to RTs and do not apply also to accuracy data (e.g., p. 11, lines 257-263; “there was a significant interaction between Gesture-type and spatial Compatibility, indicating that the Interactive gestures led to a reduced CE as compared to Communicative gestures”).

5) When discussing the role of interference effects in interactive situations (e.g., p.18, lines 424-427), the Authors may also find of interest a recent paper in which the observation of an interactive request gesture is reported to facilitate the execution of an incongruent, but appropriate, complementary response, while interfering with a congruent, but inappropriate, action (Betti et al., 2019, PeerJ).

Typo: a dot is missing in p.11, line 262.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

31 Mar 2020

Dear Dr. Gros-Louis,

We would like to thank you for giving us the opportunity to submit a revised version of our manuscript entitled “The engaging nature of interactive gestures” and the Reviewers for their thoughtful comments. We have addressed all points raised by the Reviewers. In the point-by-point response letter, the Reviewer’s points are reported in plain font, our replies in bold font and preceded by "A.R." (Authors’ Response). Changes to the main text are highlighted in italics.

We hope that you and the Reviewers will be satisfied with our revisions.

We look forward to hearing from you.

Kind regards

Arianna Curioni, also on behalf of Günther Knoblich, Natalie Sebanz and Lucia Maria Sacheli

Reviewers’ comments

Reviewer #1: This extremely well-written ms. describes an elegant suite of studies that explore the engaging nature of "interactive" rather than "communicative" actions, i.e. actions which require activity on the part of the human observer drawing him/her in to interact. By cleverly combining this with a spatial compatibility task, the interference of such stimuli was objectively assessed. Across different experiments it was shown that interactive gestures impact response selection and spatial compatibility as compared to communicative actions. The authors interprete this as evidence for a priviliged access of interactive gestures to perceptual and action systems.

I applaud the authors to systematically addressing this subtle, but important distinction! I am very impressed by the choice of stimuli and the methodological rigor with which the studies were conducted. I strongly recommend publication of this ms. and only have minor comments:

A.R. We thank Reviewer #1 for her positive evaluation of our manuscript and for the constructive comments that helped us to improve our manuscript. Please find our point by point response below.

1) A study by von der Lühe et al. (2016) also investigated the impact of "communicative" actions. The authors may want to make reference to this study and could also use this opportunity to briefly discuss whether interactive gestures might bring about different expectations/predictions in the observer and how this might play out in a predictive coding scheme.

A.R. We thank Reviewer #1 for offering this interesting discussion point. We now refer to this study in the Introduction, and have amended the main text as follows (page 3):

“It has been already proposed, in fact, that we might process social actions in terms of their sensorimotor consequences in the world, as this would give us a predictive advantage and the possibility to prepare for a timely response (Kilner et al,. 2007; von der Lühe 2016).”

2) Schilbach and colleagues published several papers (2012, 2013), in which they investigated the impact of social gaze cues on spatial compatibility effects. This papers appear to be missing in the bibliography.

A.R. We thank Reviewer #1 for pointing out this gap in our reference list. We agree that the Schilbach et al. studies are indeed relevant for the present manuscript. We have therefore added a reference to these important papers in the introduction (page 4):

“Similar facilitatory effects have been also found in spatial compatibility tasks where social gaze cues were presented to participants prior to the go signal and reduced the compatibility effect (Schilbach et al., 2012, 2011)”

And discuss the present findings in light of the Schilbach et al. studies (page 19):

“Interesting, this would be in line with evidence of a similar facilitatory effect recorded during spatial compatibility tasks when social gaze cues were presented to participants: in a series of studies, Schilbach at al. showed that being ‘looked at’ by a virtual other induces a social engagement that facilitates the selection of the spatially incongruent response, resulting in a reduction of the compatibility effect (Schilbach et al., 2012, 2011).

3) Finally, it might be tempting to speculate how interactive gestures might be important (motivationally and otherwise) for matters of mental health. It has been suggested that psychiatric disorders could be construed as "disorders of social interaction", where interactive gestures might also figure prominently (Schilbach 2016). Maybe the authors might want to speculate (or not).

A.R. We thank Reviewer #1 for raising the interesting point of the possibility that several psychiatric conditions could be conceived at disorders of social cognition impairments (defined as the impairment of the processes involved in understanding and engaging with, rather than merely observing, other people). Notwithstanding the relevance of this topic for understanding the behavioral and neural processes supporting social interactions, we feel that our work does not directly contribute to this debate as we do not compare data from neurotypical and clinical populations. However, we have mentioned at the end of the Discussion (page 20) that the application of our paradigm to psychiatric conditions may be an interesting topic for future studies:

“The results of the present study indicate that the involuntary perception of interactive gestures influences the participants’ responses, as if our perceptual system was equipped with the ability to fast identify opportunity for interactions independently of the actual engagement in a social exchange: these findings pave the way for future investigations to address when the "perceptual advantage" of interactive gestures emerges, what social abilities it might require, and whether it might be impaired in situations where the possibility to engage in successful interactions is also impaired, as it might be the case in psychiatric disorders [62]”

Reviewer #2: Curioni and co-authors report an interesting study in which, in three behavioral experiments, they tested the difference between two types of social actions, namely interactive and communicative gestures, in a spatial compatibility task. By measuring the interference effect of observing lateralized task-irrelevant gestures when participants were requested to respond to target stimuli, they showed a reduced spatial compatibility effect in reaction times for interactive gestures (which were chosen to elicit a complementary response in the observers) as compared to communicative gestures, irrespective of the responding hand. Response accuracy showed that participants were more accurate when the position of the distractor and target was on the same side, regardless of the type of stimulus (interactive, communicative).

The manuscript is well written and provides an interesting extension to the available literature. The experimental design is sound and the data were rigorously analyzed.

I have few queries that the Authors should address to improve the clarity of the manuscript.

A.R. We thank Reviewer #2 for her positive evaluation of our manuscript and for the constructive comments that helped us improve our manuscript. Please find our point by point response below.

1) The first point regards the discussion of the obtained results with reference to the potential difference in salience for interactive gestures, given their intrinsically engaging nature, as compared to communicative gestures.

The results of all three experiments consistently show for RTs that interactive gestures lead to a reduction of the spatial compatibility effect as compared to communicative gestures. Therefore, it emerged that for interactive gestures, incompatible trials (distractor and targets on opposite sides) were less incompatible/interfering than for communicative gestures. Given that this effect was not limited to the dominant hand which would have been recruited to perform a complementary response to the presented interactive gesture, the Authors conclude that this could seemingly be a generalized effect. Thus, the effect would not reflect any specification of the required motor plan, but be related to a general motor engagement to socially interact which is activated when we are faced with interactive gestures. This unspecific effect for interactive gestures could be dependent on differences in the adopted types of stimuli, however the Authors report to have controlled for this in a go/no-go control task reported in the supplementary materials (e.g., p.18, line 423: “stimuli in the two categories were matched for perceptual salience and social relevance”). Nonetheless, it is unclear to me how the adopted go/no-go task can disentangle the issue of difference in salience/social relevance between communicative and interactive gestures.

I would then ask the Authors to comment on this aspect as it is relevant for the interpretation of the obtained results. In addition, please better report both in the supplementary materials and in the main text the rationale of having used this type of control task for the validation of the experimental stimuli and to control for perceptual salience and social relevance.

A.R. We agree with Reviewer #2 that it is important to address the rationale of the control experiment more extensively. For the Communicative gestures set, we have chose stimuli that carry a positive/engaging social meaning for the observer, and that are performed towards the observer. In order to ensure that the two stimuli sets (Interactive and Communicative) did not differ in terms of perceptual salience, we ran a control task where we measured the time and accuracy in detecting each gesture when randomly presented together with all the other gestures. The results showed no difference in response time and accuracy among individual stimuli, and also show no overall difference between stimulus sets (Interactive and Communicative, see response to the Reviewer’s Point N.2). We chose such a perceptual task because it provides an independent behavioural measure of the difficulty of stimulus detection. We preferred this over an explicit validation of the stimuli set, that may have potentially been biased by interindividual differences in interpreting the questions and/or the rating scale. We also would like to point out that, in order to further control for potential differences driven by single stimuli within the two sets, we performed all main analyses including stimulus number in the random structure, so that inter-stimulus differences (e.g., in salience) were controlled for. Importantly, moreover, we would like to drive the reviewer’s attention on the fact that we never find a significant main effect of gesture-type: this already indicates that any alternative explanation of our results in terms of perceptual or salience difference between the two categories (Interactive vs. Communicative) cannot account for our results. We have now clarified the rationale at the basis of our go-no go task in the Method section (page 7):

“To verify that stimuli were matched for salience, we ran a preliminary experiment on an independent sample of 14 participants using a go/no-go task. In this task, we measured the time and accuracy in detecting each gesture of the stimulus set when randomly presented together with all the gestures of the set. We chose such a perceptual task as it provides an independent behavioural measure of stimulus detection. If gestures are matched for salience and valence, detection performance (accuracy and response times) should be comparable across gesture categories and across individual gestures. We preferred this over an explicit validation of the stimuli set because such validations may be influenced by interindividual differences in interpreting the validation questions, and/or rating scales. The results indicated that the salience of Interactive and Communicative gestures was comparable, as this was also supported by a Bayesian statistical analysis (see Supplementary Materials).”

We have also amended the Supplementary material accordingly.

2) In addition, the statistical analysis of the go/no-go task included only the Stimulus number as within-subject factor, however it would be relevant to test the role of Gesture-type to be able to differentiate between the two types of gestures. Plus, in the control task the results for the accuracy are missing, while only the ANOVA results for RTs are reported. Please, add these missing information to the results paragraph (see supplementary materials).

A.R. We are grateful to the Reviewer for pointing out these omissions. We now report in the Supplementary material a table summarizing the RTs and ACC results of the control task (see Supplementary Table S1). We also compared the mean ACC and RTs at the go/no go task between the two stimulus categories by means of a non-parametric test (for ACC data) and a paired-sample t-test (for RTs data). These analyses revealed no difference between categories, as also confirmed by a Bayesian comparison (see Supplementary Results):

“We also compared the mean Accuracy (Acc) and Reaction Times (RTs) between Stimulus Category (Interactive, mean Acc 0.88 +/- 0.12, mean RTs 645.32 +/- 85.37; Communicative, mean Acc 0.91 +/- 0.10, mean RTs 639.58 +/- 90.30). The non-parametric Wilcoxon test comparing mean Acc between Stimulus Categories showed no significant effect (W = 41.5, p = 0.17). The paired-sample t-test comparing mean RTs also showed no significant effect (t(13) = -.57, p = .58).

Finally, we performed a Bayesian Paired-Sample T-Test on RTs data to explore whether the data provided evidence in favor of the null hypothesis (i.e., absence of difference between Stimulus categories). The results showed a Bayesian Factor (BF10) equal to 0.31, indicating a moderate evidence in favor of the null hypothesis, i.e., that the mean RTs showed by the participants at the control task were equal for Communicative and Interactive gestures, suggesting there was no difference in salience between the Stimulus categories.”

3) Results: please report a measure of effect size in the results.

In relation to this, is the ηp 2 = .25 reported in the sample size calculation (pp.14-15, lines 350-355) as the observed effect size of the Compatibility x Stimulus type second order interaction referring to results of Experiment 1 or 2?

A.R. We thank the reviewer for pointing out that the description of the power analysis was unclear. The Power analysis run to establish the required sample-size for the Experiment 3 was indeed based on the results from Experiment 2, as both experiments use an intra-manual task. This point is now clarified at page 15. In this regard, we highlight that all the results reported in the main text are based on mixed model analyses, as indicated in the methods section: we believe this analytical approach is crucial in our design to control for between-subject and between-stimulus variability. Yet, the quantitative estimation of effect size directly from mixed-model analyses is quite problematic. Thus, in order to run the power analysis with G-Power, we re-analyzed the data of the Experiment 2 with a GLM: this analysis showed a significant Stimulus category by Compatibility interaction (F(1,29)= 8.96, p = .006) with an ηp 2 = .24 (this value has been amended, we apologize for the typo).

Moreover, to provide a clear indication of the effect-size, in each experiment we calculated the Cohen’s d values of each effect of interest based on the marginal means and standard deviations of the compatibility effect. These results are the following:

Experiment 1, d = -1.675 (see page 11):

“To directly compare the size of the CE between Interactive and Communicative gestures, we also computed an index of the CE for each participant (RT Incompatible – RT Compatible) separately for Gesture-type. A Dependent Sample t-test revealed a significant difference in the CE between Communicative and Interactive gestures (t(27) = -8.86, p < 0.001, d = -1.675), with the effect being bigger for Communicative gestures (mean 17 ms, sd 6 ms) than for Interactive gestures (mean 9 ms, sd 2 ms) (Fig 2).”

Experiment 2, d = -0.547 (page 14):

“To directly compare the size of the CE between Gesture-types, we computed an index of the CE for each participant (RT Incompatible – RT Compatible) for Interactive and Communicative gestures. A Dependent Sample t-test revealed a significant difference in the CE between Communicative and Interactive gestures (t(29) = -2.99, p = 0.006, d = -0.547), with the effect being bigger for Communicative gestures (mean 39 ms, sd 26 ms) than for Interactive gestures (mean 25 ms, sd 24 ms) (Figure 2).”

Experiment 3, d = -1.342 (page 17):

“To directly compare the size of the CE between Gesture-types, we computed an index of the CE for each participant (RT Incompatible – RT Compatible) for Interactive and Communicative gestures. A Dependent Sample t-test revealed a significant difference in the CE between Communicative and Interactive gestures (t(19) = -6.0, p < 0.001, d = -1.342), with the effect being bigger for Communicative (mean = 42 ms, sd = 13 ms) than for Interactive gestures (mean 18 ms, sd 6 ms).”

4) Discussion: When discussing the obtained results in the Discussion paragraphs, please specify that they refer to RTs and do not apply also to accuracy data (e.g., p. 11, lines 257-263; “there was a significant interaction between Gesture-type and spatial Compatibility, indicating that the Interactive gestures led to a reduced CE as compared to Communicative gestures”).

A.R. We have amended the main text accordingly.

5) When discussing the role of interference effects in interactive situations (e.g., p.18, lines 424-427), the Authors may also find of interest a recent paper in which the observation of an interactive request gesture is reported to facilitate the execution of an incongruent, but appropriate, complementary response, while interfering with a congruent, but inappropriate, action (Betti et al., 2019, PeerJ).

A.R. We now discuss this study on page 18:

“Our results are also in line with evidence for a facilitation of socially relevant motor responses during the observation of interactive requests, even when such responses are kinematically incongruent with the observed actions: this suggests that social affordance effects are strong enough to override the impact of automatic imitation of observed actions [Betti et al., 2019]. “

Typo: a dot is missing in p.11, line 262.

A.R. We have amended the main text accordingly.

8 Apr 2020

The engaging nature of interactive gestures

PONE-D-20-02355R1

Dear Dr. curioni,

We are pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it complies with all outstanding technical requirements.

Within one week, you will receive an e-mail containing information on the amendments required prior to publication. When all required modifications have been addressed, you will receive a formal acceptance letter and your manuscript will proceed to our production department and be scheduled for publication.

Shortly after the formal acceptance letter is sent, an invoice for payment will follow. To ensure an efficient production and billing process, please log into Editorial Manager at https://www.editorialmanager.com/pone/, click the "Update My Information" link at the top of the page, and update your user information. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, you must inform our press team as soon as possible and no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

With kind regards,

Julie Jeannette Gros-Louis, PhD

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

13 Apr 2020

PONE-D-20-02355R1

The engaging nature of interactive gestures

Dear Dr. curioni:

I am pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

For any other questions or concerns, please email plosone@plos.org.

Thank you for submitting your work to PLOS ONE.

With kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Julie Jeannette Gros-Louis

Academic Editor

PLOS ONE