Understanding others’ minds: Social context matters

Paula Fischer — PhD Candidate, Cognitive Development Centre, Department of Cognitive Science, Central European University

Imagine that you are walk­ing with your friend through the forest, and sud­denly you find yourselves next to a bush filled with red ber­ries. Let’s sup­pose that you know a lot about dif­fer­ent plants, and you imme­di­ately recog­nise that these ber­ries are not only red ber­ries, but that they are also dan­ger­ous. In fact, they are pois­on­ous. However, you can see the sparkle in your friend’s eyes, and that he is already reach­ing towards the ber­ries to replen­ish his energy levels after the long walk. What do you do? Well, if you would like to save the life of your friend, or at least pre­vent him from an unpleas­ant exper­i­ence, you would warn him. You would do this because you under­stand that he believes that these ber­ries are good to eat, and you know that he wouldn’t go for these ber­ries if he knew that they were dan­ger­ous.

From this example and oth­er every­day exper­i­ences, we can see that humans pos­sess highly soph­ist­ic­ated abil­it­ies to ‘read’ oth­ers’ minds. This abil­ity, called the Theory of Mind (ToM), enables us to attrib­ute men­tal states to oth­ers, and to make pre­dic­tions and draw infer­ences from their beha­vi­or and actions to their men­tal states. It is there­fore essen­tial for social inter­ac­tions, because it under­pins our being able to effect­ively coordin­ate and com­mu­nic­ate with oth­ers. Researchers have been invest­ig­at­ing this ability’s char­ac­ter­ist­ics for dec­ades, and much of this research has focused on when and how it devel­ops. In this post, I will pro­pose that one aven­ue for mak­ing pro­gress in resolv­ing open ques­tions about the devel­op­ment of ToM can be made by appeal­ing to when we use ToM.

Since Dennett (1978) poin­ted out that attrib­ut­ing true beliefs to oth­ers can­not be empir­ic­ally dis­tin­guished from agents simply mak­ing pre­dic­tions about the actions of oth­ers on the basis of their own know­ledge and beliefs about the world, the con­ven­tion­al test for ToM became prob­ing false belief (FB) under­stand­ing. One typ­ic­al way to test for the under­stand­ing of false beliefs in chil­dren is the location-change task (Wimmer & Perner 1983; Baron-Cohen, Leslie & Frith, 1985). In such a stand­ard false belief task, par­ti­cipants are exposed to a story in which the main char­ac­ter has a false belief regard­ing a loc­a­tion of an object (as a second char­ac­ter changed its loc­a­tion while she was absent). When to expli­citly indic­ate where the first char­ac­ter will look for the object, chil­dren typ­ic­ally fail to take into account her false belief before the age of 4, answer­ing (or point­ing) towards the new (actu­al) loc­a­tion of the object (Wimmer & Perner, 1983, Perner, Leekam & Wimmer, 1987).

There has been an ongo­ing debate as to wheth­er the abil­ity to under­stand oth­ers’ (false) beliefs is early devel­op­ing, or wheth­er it devel­ops only from around the age of 4 with the emer­gence of oth­er abil­it­ies, e.g. exec­ut­ive func­tion and lan­guage (see for example Slade & Ruffmann, 2005). Two main lines of research have col­lec­ted evid­ence either for or against these state­ments. One line of research which uses impli­cit meas­ures of false-belief under­stand­ing, mostly influ­enced by Leslie’s the­ory on pre­tence (Leslie, 1987), sug­gests that infants are sens­it­ive to oth­ers’ beliefs from very early on. For example, Onishi and Baillargeon (2005) found evid­ence of false-belief under­stand­ing in 15-month-olds using a viol­a­tion of expect­a­tions paradigm (see Scott & Baillargeon, 2017 for a review on this research). The oth­er line of research instead sug­gests that full-blown ToM devel­ops only after the age of 4. This line of research attempts to explain pos­it­ive find­ings with young­er infants by appeal­ing to either low level cues (e.g. Heyes, 2014), or a min­im­al ToM account (Apperly & Butterfill 2009) which pro­poses that an early devel­op­ing sys­tem is rich enough to rep­res­ent belief-like states only (but not beliefs per se).

How can this puzzle regard­ing early mind read­ing be solved? One may ask: if there is a con­cep­tu­al change around the age of 4, then what exactly hap­pens around that time that allows or trig­gers such change? I will sug­gest that focus­ing on why ToM is cru­cial in sev­er­al aspects of our every­day social lives (from lan­guage devel­op­ment and com­mu­nic­a­tion, to cooper­a­tion and altru­ist­ic beha­viour) may provide a means of answer­ing this ques­tion.

Can the basic abil­ity to track oth­ers’ men­tal states con­trib­ute to lan­guage acquis­i­tion? Some exper­i­ment­al evid­ence sup­ports the hypo­thes­is that, from a rel­at­ively early age, infants are sens­it­ive to semant­ic incon­gru­ity. That is, they under­stand when an object is labelled incon­gru­ently from its real mean­ing (e.g. Friedrich & Friederici, 2005; 2008). A study by Forgács and col­leagues (2018) invest­ig­ated wheth­er infants would track such semant­ic incon­gru­it­ies by oth­ers’ per­spect­ives. They meas­ured 14-months-olds event-related poten­tial (ERP) sig­nals, and found that infants show N400 activ­a­tion (a well-established neuro­psy­cho­lo­gic­al indic­at­or of semant­ic incon­gru­ity) not only when objects are incon­gru­ently labelled from their own view­point, but also from their com­mu­nic­at­ive partner’s point of view (see also Kutas & Federmeier, 2011; Kutas & Hillyard, 1980). These find­ings sug­gest that infants track the men­tal states of social part­ners, keep such attrib­uted rep­res­ent­a­tions updated, and use them to assess oth­ers’ semant­ic pro­cessing. This study can fur­ther be taken as indic­at­ing that rep­res­ent­a­tion­al capa­cit­ies (such as those required for belief ascrip­tion) are present at 14-month-olds in a com­mu­nic­at­ive con­text.

Such belief attri­bu­tion in sim­il­arly young infants can also be observed in ostensive-communicative infer­en­tial con­texts. In a study by Tauzin and Gergely (2018), infants’ look­ing time was meas­ured dur­ing the obser­va­tion of unfa­mil­i­ar com­mu­nic­at­ive agents; chil­dren needed to inter­pret the turn-taking exchange of vari­able tone sequences, which was indic­at­ive of com­mu­nic­at­ive trans­fer of goal rel­ev­ant inform­a­tion from a know­ledge­able to a naïve agent. In their exper­i­ments, infants observed the fol­low­ing inter­ac­tion: one of the agents placed a ball in a cer­tain loc­a­tion, and later saw the ball mov­ing to a dif­fer­ent loc­a­tion. The oth­er agent, who had not observed the location-switch, later tried to retrieve the ball. Based on their look­ing times, infants only expec­ted the ball-retrieving agent to go to where the ball really was if the first agent (who observed the location-switch) com­mu­nic­ated the trans­fer. Based on these find­ings, the authors sug­ges­ted that 13-months-old infants recog­nised these turn-taking exchanges as com­mu­nic­at­ive inform­a­tion trans­fer, sug­gest­ing that they can attrib­ute communication-based beliefs to oth­er agents if they can infer the rel­ev­ant inform­a­tion that is being trans­mit­ted.

Besides play­ing a role in chil­dren com­ing to under­stand import­ant aspects of com­mu­nic­a­tion, ToM may play a cru­cial part in cooper­a­tion and altru­ist­ic beha­viour. The ques­tion as to how ToM relates to, for instance, instru­ment­al help­ing, has received rel­at­ively little atten­tion. One of the first stud­ies prob­ing the rela­tion­ship between false belief under­stand­ing and help­ing comes from Buttelmann, Carpenter and Tomasello (2009). During their exper­i­ments, infants observed a prot­ag­on­ist strug­gling to open a box in order to obtain a toy. In the crit­ic­al part of the exper­i­ment the toy was moved by anoth­er agent from its ini­tial box to a dif­fer­ent box. The prot­ag­on­ist either observed this move, or had left the room.  When the main prot­ag­on­ist had left the room and then tried to open the box which ini­tially con­tained the toy, infants spon­tan­eously helped him by indic­at­ing that he should try to open the altern­at­ive box instead. However, when the main prot­ag­on­ist observed the location-switch, infants helped him open the ini­tial box. This sug­gests that by 18 months of age, help­ing beha­viour is guided by the beliefs of the helpee. This study, amongst oth­ers (see also Matsui & Miura, 2008), sup­port the hypo­thes­is that rep­res­ent­ing oth­ers’ men­tal states is a key fea­ture for help­ing and cooper­at­ing, and that infants are cap­able of tak­ing into account oth­ers’ beliefs when help­ing spon­tan­eously from very early on.

The abil­ity to rep­res­ent oth­ers’ men­tal states plays a cru­cial part in our social lives. Understanding what oth­ers think is import­ant not only for high-level cooper­at­ive or com­pet­it­ive prob­lem solv­ing, but even in smal­ler day-to-day social inter­ac­tions when we need to act fast (e.g., pre­vent­ing our friends from com­ing to harm dur­ing a walk). The stud­ies dis­cussed here sug­gest that from a rel­at­ively early age, humans are able to adjust their help­ing beha­viour on the basis of oth­ers’ beliefs, and the beliefs of oth­ers may shape children’s under­stand­ing of com­mu­nic­at­ive epis­odes. Future research may do well to keep in mind that when it comes to ToM, social con­text seems to mat­ter.



Apperly, I. A., & Butterfill, S. A. (2009). Do humans have two sys­tems to track beliefs and belief-like states?. Psychological review116(4), 953.

Baron-Cohen, S., Leslie, A. M., & Frith, U. (1985). Does the aut­ist­ic child have a “the­ory of mind”?. Cognition21(1), 37–46.

Buttelmann, D., Carpenter, M., & Tomasello, M. (2009). Eighteen-month-old infants show false belief under­stand­ing in an act­ive help­ing paradigm. Cognition112(2), 337–342.

Dennett, D. C. (1978). Beliefs about beliefs [P&W, SR&B]. Behavioral and Brain sci­ences1(4), 568–570.

Forgács, B., Parise, E., Csibra, G., Gergely, G., Jacquey, L., & Gervain, J. (2018). Fourteen-month-old infants track the lan­guage com­pre­hen­sion of com­mu­nic­at­ive part­ners. Developmental sci­ence, e12751.

Friedrich, M., & Friederici, A. D. (2005). Lexical prim­ing and semant­ic integ­ra­tion reflec­ted in the event-related poten­tial of 14-month-olds. Neuroreport16(6), 653–656.

Friedrich, M., & Friederici, A. D. (2008). Neurophysiological cor­rel­ates of online word learn­ing in 14-month-old infants. Neuroreport19(18), 1757–1761.

Heyes, C. (2014). Submentalizing: I am not really read­ing your mind. Perspectives on Psychological Science9(2), 131–143.

Kutas, M., & Federmeier, K. D. (2011). Thirty years and count­ing: Finding mean­ing in the N400 com­pon­ent of the event-related brain poten­tial (ERP). Annual Review of Psychology, 62, 621–647.

Kutas, M., & Hillyard, S. A. (1980). Reading sense­less sen­tences: Brain poten­tials reflect semant­ic incon­gru­ity. Science, 207(4427), 203–205.

Leslie, A. M. (1987). Pretense and rep­res­ent­a­tion: The ori­gins of” the­ory of mind.”. Psychological review94(4), 412

Matsui, T., & Miura, Y. (2008). Pro-social motive pro­motes early under­stand­ing of false belief.

Onishi, K. H., & Baillargeon, R. (2005). Do 15-month-old infants under­stand false beliefs?. sci­ence308(5719), 255–258.

Perner, J., Leekam, S. R., & Wimmer, H. (1987). Three-year-olds’ dif­fi­culty with false belief: The case for a con­cep­tu­al defi­cit. British journ­al of devel­op­ment­al psy­cho­logy5(2), 125–137.

Scott, R. M., & Baillargeon, R. (2017). Early false-belief under­stand­ing. Trends in Cognitive Sciences21(4), 237–249.

Tauzin, T., & Gergely, G. (2018). Communicative mind-reading in pre­verbal infants. Scientific reports8(1), 9534.

Slade, L., & Ruffman, T. (2005). How lan­guage does (and does not) relate to the­ory of mind: A lon­git­ud­in­al study of syn­tax, semantics, work­ing memory and false belief. British Journal of Developmental Psychology23(1), 117–141.

Wimmer, H., & Perner, J. (1983). Beliefs about beliefs: Representation and con­strain­ing func­tion of wrong beliefs in young chil­dren’s under­stand­ing of decep­tion. Cognition13(1), 103–128.