ZHANG Hai-hang ZHOU Ji-fan

(Department of Psychology and Behavioral Sciences, Zhejiang University, Hangzhou 310007, China)

1 Introduction

When walking with friends, we can effortlessly keep up with their pace. In a table tennis game, we do not only passively perceive the coming ball, but also automatically guess the strategy of the opponent based on online perception, predict the orbit of the ball, and conceive counterattacks all the time.Even seemingly simple social actions like shaking hands involves understanding others’ actions and goals, together with planning and taking one’s own action. The question is: how can we form the representations of interacting parties and achieve coordination with others?

The ability to form shared representations, for its role in enabling individuals to plan their actions and behave in anticipation of others’actions rather than passively responding, was treated as the cornerstone of social cognition (Sebanz, Knoblich, & Prinz, 2003, Sebanz, Bekkering, & Knoblich, 2006).Direct evidences for shared representation were given by a series of studies of Social Simon Effect, in which they found that participants’action was affected by their coactor even when they were told to perform an independent task (Sebanz, Knoblich et al. 2003；Sebanz, Knoblich et al. 2006；Atmaca, Sebanz, Prinz, & Knoblich, 2008；Dolk, Hommel, Prinz, & Liepelt, 2014；Sebanz, Knoblich, Prinz, & Wascher, 2006；Tsai, Kuo, Hung, & Tzeng, 2008). Simon effect refers to the effect found by Simon (1990) in a spatial compatibility RT(reaction time) task, in which participants do a spatial two-choice response to stimuli based on their colors, but the stimuli contain another irrelevant spatial feature: the side of screen on which they appear.Results showed that people respond faster when the stimulus appears in the same side as the responding hand.Simon effect is also known as spatial compatibility effect.Sebanz et al. (2003) extended this finding to investigate social cognition by distributing the spatial compatibility task of two hands to two people. During the experiment, two participants sat beside each other, and each of them was told to respond to one color only. That is to say, they were told to perform a Go-Nogo task, sitting together and using the same computer screen. It is interesting that the spatial compatibility effect occurred, although participants were not told to represent others, and results even showed that representing others may disturb their own performance. But when there was no co-actor alongside participants, the spatial compatibility effect has no effect on the participants, although what they did is exactly identical to that in co-acting situation. Given the different RT patterns between the alone Go-Nogo task condition and the social Go-Nogo task condition, one may deduce that people automatically represent others in social conditions. Furthermore, the similar RT patterns of the social Go-Nogo task and the individual spatial compatibility task indicated that people represent others just as their own in social context, which implies that we can maintain and manipulate the representations from both parties during live social interaction.

Not with standing the occurrence of shared representations was demonstrated in those mentioned studies, the mechanism underlying them remains largely unknown. How does an individual form the representations of oneself and co-actor？Whether the representations for oneself and others are integrated or separated？ Theoretically, those two postulations both make sense: on one hand, human beings are social animals and survive from natural selection by the help of social interaction (Tomasello, 2009), which implies that representing shared representation sintegratedly make it easy for people to treat group as a whole, enhancing the efficiency of social interaction, as to achieve the priority during evolution process. However, integrated representations may deter people from distinguishing one’s own task from others’, which may cause low efficiency of performance and do harm to the social coordination. And representing shared representations separately could keep people from being distracted by others and focus on their own ongoing tasks. Another support for separated representing process arose from a study indicating that Intelligent Quotient (IQ) is highly related to working memory capacity, which was considered as the ability to control attention and inhibit distraction(Engle, Tuholski, Laughilin, & Conway, 1999;Oberauer, Wittman, Wilhelm, & Shulze, 2002; Unsworth,& Engle, 2004;Zhang, & Fang, 2017). So, it is reasonable to speculate that separately representing shared representations is better for human beings.

Testing the two hypotheses requires a paradigm that can not only demonstrate the storage of two separated representations, but also indicate the following-up manipulating process. A recent study using a Ponzo illusion effect to test the integrated/separated mechanism of visual working memory (VWM) inspired us to conduct the present study(Shen, Xu, Zhang, Shui, Zhang, & Zhou, 2015). In Shen et al.’s (2015)study, participants were required to remember sequentially presented components of the Ponzo illusion figures. Results demonstrated that fragmented visual inputs could be involuntarily integrated into a holistic figure, leading to working memory Ponzo illusion. Based on the same rationale, we created a social working memory Ponzo illusion task to directly distinguish the integrated hypothesis from the separated hypothesis.In this study, two frames of Ponzo figures were distributed to two participants to memorize, and we predict that: if participants represent the task from both parties integratedly, there would be a positive Ponzo effect in memory; while if participants form isolated representation for both parties respectively, there would be no Ponzo effect to be discovered.

In addition, previous findings have suggested that different types of social interactions would influence the formation shared representations(Aiello & Douthitt, 2001, Guerin, 2010;Zajonc, 1965; Greenwald, 1970;Prinz, 1997;Schuch & Tipper, 2007). For example, people are more likely to take others’ tasks into their consideration and form the shared representations in the prosocial situation, but they tend to focus on their own task in the antisocial situation (Hommel, Colzato, & Van Den Wildenberg, 2009). However, there is also evidence showing that people automatically represent others’ tasks as their own, no matter what kind of social relationship they are in (Ruys & Aarts 2010). Here, we set up both typical prosocial relationship (cooperation) and antisocial relationship (competition) to further examine whether people form social representations separately or integratedly, and whether the separated/integrated rules of social representation are affected by different kinds of social relationships.

2 Experiment 1:Ponzo illusion in VWM

To confirm the stability of VWM Ponzo illusion, we firstly replicated the VWM Ponzo illusion(Shen, Xu et al. 2015). All the procedure and settings are the same as those in previous study.

2.1 Participants

18 participants were recruited from undergraduates of Zhejiang University. They had normal or corrected-to-normal vision, and were naive to the aim of the study. They received financial compensation for their participation. The study was approved by the institutional review board at the Department of Psychology and Behavioral Science, Zhejiang University. All our sample sizes were in line with previous study of VWM Ponzo illusion (Shen, Xu et al. 2015).

2.2 Apparatus

The stimuli were presented on a Dell 19-inch CRT monitor with a spatial resolution of 1024 × 768 pixels at a 100-Hz refresh rate. The experiment was written in MATLAB, using the Psychophysics Toolbox extensions (Brainard, 1997). Participants sat approximately 60 cm from the monitor during the experiment.

2.3 Materials

Figure 1 a manifests the Ponzo illusion figure used in this study. The figure contains four white lines: two converging lines with a length of 4.4° visual angle and two identical horizontal lines of 3.9° length. The distance between the upper ends of the converging lines was 1.9° and was 5.7° between the lower ends. The distance between the two horizontal lines was 2.9°.

2.4 Design and Procedure

The experiments used a partial-report working memory design, which required participants to memorize the sequentially presented lines and reproduce one of them after a short time. We adopted a single factor within-subject design with the four target conditions (i.e., one of the four lines designated as the target).

Figure 1 The stimuli and the procedure in Experiment 1

(a) The Ponzo illusion used in the current study, with the length of each line and the distance between lines indicated in visual angle. (b) Sequence of events for each trial.

To begin with each trial, a fixation presented in the center for 500 ms followed by a 1000-ms blank screen. Four-linePonzoillusion figure was divided into two parts that were presented separately in two sequential frames. Figure 1b illustrates the procedure of Experiment 1: the converging lines appeared for 200 ms, then a mask presented for 100 ms. after a 1000-ms inter-stimulus interval with a blank screen, the horizontal lines were presented for 200 ms, followed by a mask. During the task, four lines never appeared simultaneously on the screen.

Participants were required to memorize all the length of all lines. After a 1000 ms blank screen, participants were presented with a probe line. The task was to adjust the length of the probe line to reproduce the target line indicated by the location of the probe line. The probe line always appeared at the same location as one of the four memorized lines and with the same orientation. The starting length of the probe line varied randomly (range ±1°-2.9°) with respect to the length of the target line. The reproduced length reflected the representation in VWM. No feedback was provided after participants’ responses.

There were 240 trials in formal experiment. Each of the four lines had an equal chance of being the target, therefore each line was designated as target for 60 trials. All the trials were presented in a pseudorandom order. To examine the occurrence of the Ponzo illusion, our analysis focused on the trials in which the horizontal lines were assigned to be targets. If the illusion occurred in VWM, the upper line would be perceived as longer than the lower line.

2.4 Results

Figure 2 Results from Experiment 1

The upper line was reproduced significantly longer than the lower line, indicating the occurrence of the VWM Ponzo illusion. Error bars indicate standard errors of the mean.

The results are shown in Figure 2, the reproduced length of the upper line was significantly longer than of the lower line (4.20° vs. 3.91°,t(17) = 3.00,p=0.001,d=0.94), suggesting that the Ponzo illusion occurred in VWM. Hence, we have replicated the VWM Ponzo illusion effect (Shen, Xu et al. 2015).

3 Experiment 2:VWM Ponzo illusion in social situation

Based on the solid VWM Ponzo illusion effect in Experiment 1, here we adopted the VWM Ponzo illusion effect as the index to investigatethe mechanism of sharedrepresentations.

The VWM Ponzo illusion task in Experiment 1 was divided into two complementary parts for two participants: before the fixation appeared, a memory cue was added to instruct which of the two frameswasrequired to be memorized. The memory cue was an Arabic digit (1 or 2), indicating the first and the second frame, respectively. One of the lines in that frame was assigned to be the target; while those two lines in the uncued frame was assigned to his or her co-actor to memorize. Two participants were seated in two separated compartments of the same room.

Pairs of participants were recruited, and they were told to choose the role played in the interactive situation by casting lots. In the cooperative situation, two participants were told that they were“sharing the same fate”: the final award were defined by their average mark 2; while in the competitive situation, two participants were told that they are competing with each other, and final award would be given to whom have better performance.

During the task, participants were given feedbacks in every 40 trails, which contain: the mark of themselves, the mark of co-actors and the average mark of them both.

3.1 Results

Figure 3 Results from cooperation situation and competition situation

In both experiments, the upper line was reproduced significantly longer than the lower line, indicating the occurrence of the VWM Ponzo illusion. Error bars indicate standard errors of the mean.

In the cooperative situation, the results are shown in Figure 3, the reproduced length of the upper line was significantly longer than of the lower line (4.15° vs. 3.98°,t(17)=2.44,p=0.026,d=0.58), suggesting that the VWM Ponzo illusion occurred in the cooperative situation.

In the competitive situation, the reproduced length of the upper line was significantly longer than of the lower line (4.15°vs. 3.94°,t(17)=3.37,p=0.004,d=0.79), suggesting that the VWM Ponzo illusion occurred in the competitive situation.

4 Discussion

In this study, we use variant of VWM Ponzo illusion task to test the mechanism of the shared representations in social interaction. Although two participants were required to memorize part of the Ponzo illusion figures respectively, the VWM Ponzo illusion still existed. The results imply that rather than separately representing both parties in social interaction,people represent the social information of both sides in an integrated way.

To compare the Ponzo illusion effects in the non-social condition with that in the social condition, the reproduced length was submitted to a 3 (situation: individual/ cooperation/competition) × 2 (line position: upper line/lower line) Mixed-design ANOVA. The results revealed no significant interaction,F(2, 51) =0.952,p=0.393, prep=0.036. To clarify the difference ofthe Ponzo illusion effects between individual condition and social condition, we combined the results in both cooperative and competitive conditions and compare them with those in individual condition. A 2 (situation: individual/social) × 2 (line position: upper line/lower line) Mixed-design ANOVA showed no significant interaction,F(1, 52)=1.688,p=0.20, prep=0.031. Compared to that in individual condition (0.29°), the Ponzo illusion effect obtained in social condition (0.19°) was slightly lower, suggesting a weaker integrated effect in social condition. We speculated that this difference may result from different instruction: in the individual condition, participants were required to remember four lines, so they must encode four lines carefully, while in the social conditions,participants were instructed to memorize only a part of the Ponzo figure while neglecting the other part, this procedure should achieve exactly the opposite of integration (Song, Liu, Lu, & Gu, 2017).

The consistent occurrence of VWM Ponzo illusion effects in both cooperative and competitive situation simply that shared representations occurred involuntarily, rather than being controlled by different types of social interactions.Similar results can be traced to the developmental studies of Theory of Mind (ToM),showing even 7-month-old infants have the ability to understand others’ emotion, intention and false belief (Dai, Liang,& Jia, 2017,Kovcs, Téglet, & Endress, 2010). In addition, children can use the perspective-taking and false-belief-understanding abilities to predict other people’s behaviors flexibly (Baillargeon, Scott, & Lin, 2016, Morewedge,& Keysar, 2004, Jing, Gao, Hu, & Geng, 2017). Compared to previous findings, the integrated effects observed in our study extended the study of co-representation from high level social decision-making process to the VWM process. However, our results are inconsistent with those studies which showed the advantages of cooperation. There is a recent developmental study showing that cooperation, but not competition, improves the reasoning abilityof 4-year-old children about others’ diverse desires. In that study, the social contexts were set up as a fishing game in which children actively cooperated or competed with experimenter(Jin, Li, He, & Shen, 2017). While in our study, the social interaction was defined by the top-down instruction of the task, which may lead to low ecological validity, and minor difference of the performance between cooperation and competition.In the future, we should apply the social contexts in reality to our research so as to make it lively.

5 Conclusion

In conclusion, our results reveal that in social conditions, people represent the task of themselves and co-actors in an integrated way, rather than representing them separately. Additionally, the integration process is stable in different types of social interactions.

Reference

Aiello, J. R., & Douthitt, E. A. (2001). Social facilitation from Triplett to electronic performance monitoring.GroupDynamics:Theory,Research,andPractice, 5(3), 163-180.

Baillargeon, R., Scott, R. M., & Lin, Bian. (2016). Psychological reasoning in infancy.AnnualReviewPsychology, 67: 159-186.

Brainard, D. H. (1997). The psychophysics toolbox.SpatialVision,10,433-436.

Brass, M., Bekkering, H., & Prinz, W. (2001). Movement observation affects movement execution in a simple response task.ActaPsychologica,106(1), 3-22.

Dai, J., Liang, J., & Jia, N. (2017). The mechanism of mindreading during general-knowledge question task.,ChineseJournalofAppliedPsychology,23(1), 23-30.

Engle, R. W., Tuholski, S. W., Laughilin, J. E., & Conway, A. R. A. (1999). Working memory, short-term memory, and general fluid intelligence: A latent-variable approach.JournalofExperimentalPsychology:General,128(3): 309-331.

Epley, N., Morewedge, C. K., & Keysar, B. (2004). Perspective taking in children and adults: Equivalent egocentrism but differential correction.JournalofExperimentalSocialPsychology,40(6), 760-768.

Greenwald, A. G. (1970). Sensory feedback mechanisms in performance control: With special reference to the ideo-motor mechanism.PsychologicalReview,77(2), 73-99.

Guerin, B. (2010). Social facilitation: Cambridge:Cambridge University Press.

Jin, X., Li, P., He, Jie., & Shen, M. (2017). Cooperation, but not competition, improves 4-year-old children’s reasoning about others’ diverse desires.JournalofExperimentalChildPsychology,157(2017), 81-94.

Jing, J., Gao, C., Hu, A., & Geng, Q. (2017). Effect of empathy on behaviors of different roles in cyberbullying.ChineseJournalofAppliedPsychology,23(3), 265-275.

Ponzo, M. (1910). Intorno ad alcune illusioni nel campo delle sensazioni tattili, sull’illusione di Aristotele e fenomeni analoghi: Wilhelm Engelmann.

Prinz, W. (1997). Perception and action planning.EuropeanJournalofCognitivePsychology,9(2), 129-154.

Ruys, K. I., & Aarts, H. (2010). When competition merges people’s behavior: Interdependency activates shared action representations.JournalofExperimentalSocialPsychology, 46(6), 1130-1133.

Schuch, S., & Tipper, S. P. (2007). On observing another person’s actions: Influences of observed inhibition and errors.Perception&Psychophysics,69(5), 828-837.

Sebanz, N., Bekkering, H., & Knoblich, G. (2006). Joint action: bodies and minds moving together.TrendsinCognitiveSciences,10(2), 70-76.

Sebanz, N., Knoblich, G., & Prinz, W. (2003). Representing others’ actions:Just like one’s own?Cognition,88(3), B11-B21.

Sebanz, N., Knoblich, G., Prinz, W., & Wascher, E. (2006). Twin peaks: An ERP study of action planning and control in coacting individuals.JournalofCognitiveNeuroscience,18(5), 859-870.

Shen, M., Xu, H., Zhang, H., Shui, R., Zhang, M., & Zhou, J. (2015). The working memory Ponzo illusion: Involuntary integration of visuospatial information stored in visual working memory.Cognition,141,26-35.

Simon, J. R. (1990). The effect of an irrelevant directional cue on human information processing. In R. W. Proctor & T. G. Reeve (Eds.), Stimulius-response compatibility: An interated perspective (pp. 31-86). Amesterdam: North-Holland.

Song, C., Liu, W., Lu, X., & Gu, Q. (2016). Building blocks of visual working memory:Objects, features, or hybrid?ChineseJournalofAppliedPsychology,22(2), 112-126.

SüB, H., Oberauer, K., Wittman, W., Wilhelm, O., & Shulze, R. (2002). Working-memory capacity explains reasoning ability—and a little bit more.Intelligence,30(3): 261-288.

Tomasello, M. (2009).Whywecooperate,The MIT Press.

Tsai, C. C., Kuo, W. J., Hung, D. L., & Tzeng, O. J. (2008). Action co-representation is tuned to other humans.JournalofCognitiveNeuroscience,20(11), 2015-2024.

Unsworth, N., & Engle, R. W. (2005). Working memory capacity and fluid abilities: Examining the correlation between Operation Span and Raven.Intelligence,33(1), 67-81.

Vogel, E. K., McCollough, A. W., & Machizawa, M. G. (2005). Neural measures reveal individual differences in controlling access to working memory.Nature,438(7067), 500-503.

Ybarra, O., Burnstein, E., Winkielman, P., Keller, M. C., Manis, M., Chan, E., & Rodriguez, J. (2008). Mental exercising through simple socializing: Social interaction promotes general cognitive functioning.PersonalityandSocialPsychologyBulletin,34,248-259.

Ybarra, O., Keller, M. C., Chan, E., Garcia, S. M., Sanchez-Burks, J., Rios Morrison, K., & Baron, A. S. (2010).Being unpredictable: Friend or foe matters.SocialPsychologicalandPersonalityScience,1,259-267.

Ybarra, O., Winkielman, P., Yeh, I., Burnstein, E., & Kavanagh, L. (2011). Friends (and sometimes enemies) with cognitive benefits: what types of social interactions boost executive functioning?SocialPsychological&PersonalityScience,2(3), 253-261.

Zajonc, R. B. (1965). Social facilitation: Research Center for Group Dynamics, Institute for Social Research, University of Michigan.

Zhang, X., & Fang, F. (2017).Neural mechanisms of bottom-up attention.ChineseJournalofAppliedPsychology,23(2), 99-109.