Children’s grasp of the aspectuality of beliefs: the Sefo task revisited

Understanding the aspectuality of belief is regarded by many leading developmental psychologists as a hallmark of full-blown theory of mind. As Hannes Rakoczy (2017, p. 692), who has devoted much work to the experimental investigation of early understanding of aspectuality, has recently put it, “crucially, aspectuality is not just an accidental or peripheral but an absolutely fundamental and essential property of beliefs and other propositional attitudes: there is no grasp of what propositional attitudes are without some basic grasp of their aspectuality.” Providing experimental evidence for understanding the aspectuality of belief in early human childhood is also widely regarded to be a challenging task.

The goal of this post is to argue that, even though it was not the authors’ primary goal, an elegant (2010) study by Victoria Southgate, Coralie Chevallier and Gergely Csibra on early mindreading, widely known as ‘the Sefo task’, does shed light on understanding the aspectuality of belief in toddlers. But first, let me briefly explain what the aspectuality of belief is.

Much philosophy of language and mind has focused on a puzzle first discussed by the logician Gottlob Frege. This puzzle has two sides, one of which illustrates the aspectuality of belief; the other illustrates the intensionality of so-called de dicto belief-ascriptions.

Consider the following pattern of valid inference:

(1a) Cicero was a Roman orator.
(1b) Cicero=Tully.
(1c) Tully was a Roman orator.

If (1a) and (1b) are true propositions, then so must be (1c). The truth of (1a) and (1b), in other terms, entails the truth of (1c). Similarly, if (1a) is false (and (1b) is true), then (1c) must be false. This inferential pattern is generally valid: in most statements, you can replace a referring expression such as a proper name by a co-referential expression without altering the truth-value of the initial statement.

However, the validity of this inferential pattern no longer obtains in the case of belief-ascription statements such as (2a):

(2a) Sally believes that Cicero was a Roman orator.
(2b) Cicero=Tully.
(2c) Sally believes that Tully was a Roman orator.

While the truth of (1a)-(1b) entails the truth of (1c), the truth of (2a)-(2b) does not entail the truth of (2c). This demonstrates the intensionality of belief-ascription (2a). If Sally doesn’t know that ‘Cicero’ and ‘Tully’ refer to one and the same individual, then she may hold the belief that Cicero was a Roman orator without believing or even while disbelieving that Tully was a Roman orator. The fact that Cicero was a Roman orator is exactly the same fact as the fact that Tully was a Roman orator. However, this unique fact can be mentally represented in at least two distinct ways, e.g. as the belief that Cicero was a Roman orator or as the belief that Tully was a Roman orator, according to whether one and the same individual is being represented under the ‘Cicero’ mode of presentation or under the ‘Tully’ mode of presentation. This illustrates the aspectuality of belief. While aspectuality is a psychological property of beliefs, intensionality is a linguistic property of belief-reports (or belief-ascriptions). The intensionality of belief-ascriptions is the linguistic reflection of the aspectuality of the attributed beliefs: these are the two sides of Frege’s puzzle.

The point highlighted by Rakoczy’s quote is that only if an individual understands that one and the same fact can be represented by at least two different beliefs can she be credited with the capacity to attribute genuine beliefs to others. Only if one understands the aspectuality of belief can one be credited with a genuine theory of mind.

The experimental investigation of early understanding of the aspectuality of belief has yielded discrepant findings. Some of the evidence is compatible with early understanding of aspectuality. In one study, Scott and Baillargeon (2009) used as stimuli two toy penguins, one made of a single piece and the other made of two pieces that can be assembled or disassembled. When the two pieces of the two-piece penguin are assembled together, the two-piece penguin is visually indistinguishable from the one-piece penguin. They found that 18-month-olds can attribute to an agent the false belief that the two-piece penguin is the one-piece penguin. In this case, the agent’s false belief is about two distinct objects with an identical aspect rather than about a single object with two distinct aspects. Still, the capacity to understand the content of an agent’s false belief that two visually indistinguishable objects are one and the same object comes close to the capacity to understand the content of an agent’s false belief that two distinct properties of a single object belong to two different objects. In a similar vein, Buttelmann, Suhrke and Buttelmann (2015) found that 18-month-olds can represent the content of an agent’s false belief that one of two properties of an object fails to belong to this object.

However, most studies of early understanding of aspectuality so far have been based on explicit (not implicit) false-belief tasks, in which children are directly asked a question. In these explicit tasks, children know that a funny toy has two aspects: for example, a bunny toy can be transformed into a carrot toy, a pen is also an eraser or a ball is also a rattle. In these studies, the children know that the toy has two aspects, but the mistaken agent thinks that each aspect is a property of a distinct object. These explicit false-belief tasks about an object-identity are often referred to by psychologists as aspectual tasks. For a while, most studies found that explicit aspectual tasks are more challenging for children than explicit non-aspectual false-belief tasks about an object’s location (cf. Apperly and Robinson, 1998; 2003; Sprung, Perner, & Mitchell, 2007). One significant step was taken by Rakoczy, Fizke, Bergfeld and Schwartz (2015): after suitably simplifying the aspectual tasks, they found that the performances of 4-year-olds in explicit false-belief aspectual tasks and in explicit false-belief tasks about an object’s location were reliably correlated. Rakoczy and colleagues’ conclusion is well captured by the title of their paper: “Explicit [of full-blown] theory of mind is even more unified than previously assumed: belief ascription and understanding aspectuality emerge together
in development.”

A couple of weeks ago, Dan and I were having lunch and we were talking about how best to experimentally investigate early understanding of the aspectuality of belief. It occurred to me that perhaps the wonderful (2010) paper by Southgate, Chevallier and Csibra, entitled “Seventeen-month-olds appeal to false beliefs to interpret others’ referential communication,” contains the fundamental ingredients for such an investigation. Here is how.

As is typical of studies on false-belief understanding, the first study in this article involves a false-belief condition and a true-belief condition. In both conditions, a first experimenter, Vicky, brought a pair of unfamiliar objects (e.g. a green watering can spout and a red lemon squeezer) for the children to play with. After a short while, Vicky took the objects, placed each in one of two boxes, and closed the lids. She then told the toddlers that she had to go out of the room for a minute and left. At that point, a second experimenter, Coralie, appeared from behind curtains, greeted the infants and approached the boxes. She opened the boxes, switched the objects, closed the boxes, and crept back behind the curtains.

In the false-belief condition, Vicky returned to the room just after Coralie had disappeared behind the curtains. Vicky sat on the floor behind the two boxes and pointed towards, say, the left box and said, ‘Do you remember what I put in here? There’s a sefo in here. There’s a sefo in this box. Shall we play with the sefo?’ She then reached forward and simultaneously opened both boxes, without looking. At this point, the contents of the boxes were visible to the infant, but not to Vicky. She then said to the infant, ‘Can you get the sefo for me?’, while looking directly at the infant, and not signalling either box.

In the true-belief condition, the procedure was identical to the one used in the false-belief condition except that Vicky returned to the room just after Coralie had removed the toys from their initial boxes and she watched as Coralie placed each of the objects in the opposite boxes. Coralie then disappeared back behind the curtains and Vicky sat behind the two boxes. She then did exactly what she had done in the false-belief condition, opened the boxes and asked the infants ‘Can you get the sefo for me?’

What Southgate and colleagues found was that in the true-belief condition, the toddlers gave Vicky the toy that was in the box that she pointed to, but in the false-belief condition they gave her the toy that was in the other box. Suppose Vicky was pointing to the box on the left. In the true-belief condition, toddlers gave Vicky the toy that she is verbally referring to and that is in the demonstrated box, i.e. the toy in the box on the left. In the false-belief condition, toddlers gave her the toy that she is verbally referring to but that is not in the demonstrated box, i.e. the toy in the box on the right.

Now that I have described the true- and the false-belief conditions of the Sefo study, I want to do two things. First, I want to formally convince philosophically-minded skeptical readers that in the false-belief condition, Vicky’s predicament is a genuine instance of Frege’s puzzle. Secondly, I want to offer a plausible interpretation of the inferences that enable toddlers to give Vicky the toy in the non-demonstrated box or the toy in the demonstrated box, according to whether she holds a false belief or a true belief about the toy’s location.

In the false-belief condition, reflective adults would correctly ascribe to Vicky a true belief as in (3a). They would further accept the identity statement (3b). Both (3a) and (3b) are true propositions. Although (3c) is the result of replacing one description of the toy in (3a) by a co-referential description (in accordance with identity (3b)), (3c) is a false belief-ascription and would be recognized as false by reflective adults:

(3a) Vicky believes that the toy that she placed in the box on the left is the toy that she wants.
(3b) The toy that Vicky placed in the box on the left is the toy that is now in the box on the right.
(3c) Vicky believes that the toy that is now in the box on the right is the toy that she wants.

In a nutshell, the fact that the replacement of one description in (3a) by a co-referential description turns (3a), which is a true belief-ascription, into (3c), which is false, shows the intensionality of belief-ascription (3a), which in turn reflects the aspectuality Vicky’s beliefs. Hence the two sides of Frege’s puzzle are being re-instantiated.

The fact that Vicky’s predicament in the false-belief condition meets all the conditions for being a bona fide instance of Frege’s puzzle is not sufficient to show that toddlers do understand the aspectuality of Sally’s false belief. Nor do toddlers need reason explicitly in accordance with (3a)-(3c).

In both the true- and the false-belief conditions, while Vicky’s utterance is the same: ‘Do you remember what I put in here? There’s a sefo in here. There’s a sefo in this box’, there is a relevant dissociation between the two parts of her utterance.

In the false belief-condition, toddlers understand the aspectuality of Vicky’s beliefs if (and only if) they understand that she just expressed a true belief about where she last placed the sefo (i.e. the toy that she now wants) and a false belief about its present location. In other words, they understand aspectuality if they understand that Vicky is correctly thinking of the sefo under the description ‘the toy that I earlier placed in the box on the left’ and incorrectly thinking of it under the description ‘the toy that is now in the box on the left’, while they know that the description ‘the toy that is now in the box on the right’ correctly applies to the sefo (i.e. the toy that Vicky wants).

One bold deflationary hypothesis is that in the false-belief condition, the toddlers show no understanding of the aspectuality of Vicky’s beliefs. Rather, they selectively attend to the part of Vicky’s utterance that expresses her true belief about where she earlier placed the sefo (that she now wants), and they just ignore the part of her utterance that expresses her false belief about the sefo’s present location.

But in the true-belief condition also, there is a relevant, though different, dissociation between the two parts of Vicky’s utterance. When Vicky says ‘There’s a sefo in here. There’s a sefo in this box. Shall we play with the sefo?’ most competent adults would take her to be referring to the toy that she wants under the label ‘a sefo’. But when she reminds toddlers of what she earlier placed in the box she is pointing at (‘Do you remember what I put in here?’), most competent adults would assume that she is talking about the other toy, not about the sefo.

One further bold deflationary hypothesis is that in the true-belief condition, the toddlers selectively attend to the part of Vicky’s utterance in which she is talking about the sefo and they just ignore the part of her utterance in which she is reminding them of the object that she earlier placed in the box she is pointing at (which is not the sefo).

According to these two deflationary hypotheses, which are not particularly compelling, unlike adults, toddlers would selectively disregard a distinctive chunk of Vicky’s utterance in both the true- and the false-belief conditions. It is at least equally plausible that, like adults, toddlers take into account Vicky’s full utterance in both conditions. If they do, then they understand the aspectuality of Vicky’s relevant belief in the false-belief condition.

Although the Sefo study was initially designed to investigate toddlers’ capacity to attribute to others false beliefs about an object’s location, it turns out to provide all the ingredients necessary for investigating toddlers’ understanding of the aspectuality of belief. Further experimental work is needed to know whether toddlers understand the aspectuality of Vicky’s beliefs in the false-belief condition or whether the deflationary hypothesis stands. On the one hand, the issue can be addressed experimentally. On the other hand, if the deflationary hypothesis turns out to be correct, then the Sefo study fails, not only to show that toddlers understand aspectuality, but also that they can attribute to others false beliefs about an object’s location.[1]


References

Apperly, I. A., & Robinson, E. J. (1998) Children’s mental representation of referential relations. Cognition, 67, 287–309.

Apperly, I., & Robinson, E. (2003) When can children handle referential opacity? Evidence for systematic variation in 5- and 6-year-old children’s reasoning about beliefs and belief reports. Journal of Experimental Child Psychology, 85, 297–311.

Buttelmann, D., Carpenter, M., & Tomasello, M. (2009) Eighteen-month-old infants show false belief understanding in an active helping paradigm. Cognition, 112, 337–342.

Buttelmann, F., Suhrke, J., & Buttelmann, D. (2015) What you get is what you believe: Eighteen-month-olds demonstrate belief understanding in an unexpected-identity task. Journal of Experimental Child Psychology, 131, 94–103.

Fizke, E., Butterfill, S., van de Loo, L., Reindl, E., & Rakoczy, H. (2017) Are there sig- nature limits in early theory of mind? Journal of Experimental Child Psychology, 162, 209–224.

Rakoczy, H. (2017) In defense of a developmental dogma: Children acquire propositional attitude folk psychology around age 4. Synthese, 194, 689–707.

Rakoczy, H., Fizke, E., Bergfeld, D., & Schwarz, I. (2015) Explicit theory of mind is even more unified than previously assumed:Belief ascription and understanding aspectuality emerge together in development. Child Development, 86, 486–502.

Scott, R. M., & Baillargeon, R. (2009) Which penguin is this? Attributing false beliefs about object identity at 18 months. Child Development, 80, 1172–1196.

Southgate, V., Senju, A., Csibra, G. (2007) Action anticipation through attribution of false belief by 2-year-olds. Psychological Science, 18:587–592.

Sprung, M., Perner, J., & Mitchell, P. (2007) Opacity and discourse referents: Object identity and object properties. Mind & Language, 22, 215–245.

***

[1] Thanks Dan for many conversations and comments on this post.

4 Comments

  • comment-avatar
    Gergely Csibra 16 May 2018 (16:28)

    Ascribing second-order beliefs and interpreting verbal utterances by toddlers
    Pierre concludes from his insightful analysis that the Southgate et al. (2010) study may have provided a proof of understanding the aspectuality of beliefs by 18-month-old toddlers. While I would love to see that conclusion justified, I think that Pierre’s analysis suffers from some shortcomings. In particular, I think that (A) the complex mental state ascriptions that he proposes were involved in the task are unnecessary to solve it, and (B) he overestimates 18-month-olds linguistic skills.

    (A)

    Pierre proposes that, in the false belief condition of the task, “reflective adults” would accept both (3a) and (3b) and would therefore draw (3c) unless they take into account the intensionality of belief ascriptions:

    (3a) Vicky believes that the toy that she placed in the box on the left is the toy that she wants.
    (3b) The toy that Vicky placed in the box on the left is the toy that is now in the box on the right.
    (3c) Vicky believes that the toy that is now in the box on the right is the toy that she wants.

    Compared to these ascriptions and inferences, Southgate et al. (2010) suggested that successful toddlers in the false belief version of the task relied on premises (4a) and (4b), and drew the valid conclusion (4c) without having to deal with the aspectuality of Vicky’s beliefs:

    (4a) Vicky wants the toy that she believes is in the box on the left.
    (4b) The toy that Vicky believes is in the box on the left is now in the box on the right.
    (4c) Vicky wants the toy that is now in the box on the right.

    (3a) looks very different from (4a) at first sight, but it turns out that it is simply a reformulation of (4a) in the form of “Vicky believes that (4a).” Let my show you why.

    (3a) and (3b) identify the toy in question with reference to a past event (“the toy that she placed in the box on the left”), while (4a) and (4b) identify the same toy by Vicky’s belief content. The result is the same; after all, Vicky’s belief about the content of that box is generated by the past event when she placed it there. However, in the context when such a premise might be formulated (i.e., when she makes a request), past events are relevant only via justifying her current state of mind, i.e., her beliefs formed about the locations of toys. If so, then referring to her belief rather than to the cause of her belief is more relevant for the purpose the premise is formulated. Let’s then replace the reference to the past event with a reference to her belief in (3a) and (3b) to obtain (5a) and (5b):

    (5a≈3a) Vicky believes that the toy that she believes is in the box on the left is the toy that she wants.
    (5b≈3b) The toy that Vicky believes is in the box on the left is the toy that is now in the box on the right.
    (5c=3c) Vicky believes that the toy that is now in the box on the right is the toy that she wants.

    The content of Vicky’s belief in (5a), “the toy that she believes is in the box on the left is the toy that she wants,” formulates a relation between her belief about the toy in a location and about the toy of her desire, which can also be put into the simpler form of “she wants the toy that she believes is in the box on the left.” While these two statements may be linguistically and pragmatically different from each other (the first one may suggest a potential contrast against which the statement is formulated), their truth conditions are exactly the same, so they describe the same state of affairs. If so, they are equivalent as belief contents, and we can replace the first one with the second one:

    (6a≈5a≈3a) Vicky believes that she wants the toy that she believes is in the box on the left.
    (5b≈3b) The toy that Vicky believes is in the box on the left is the toy that is now in the box on the right.
    (6c≈5c=3c) Vicky believes that she wants the toy that is now in the box on the right.

    Finally, (6b) and (5b), just like (3b) from which they are derived, are formulated as identity statements. Such statements must have originated from the observation of the relocation of the toy, about which a simpler statement can be made in the form of (4b).

    (6a≈5a≈3a) Vicky believes that she wants the toy that she believes is in the box on the left.
    (6b=4b≈5b≈3b) The toy that Vicky believes is in the box on the left is now in the box on the right.
    (6c≈5c=3c) Vicky believes that she wants the toy that is now in the box on the right.

    With these clarifications it becomes clear that the premises in (3) are derived from the premises in (4) with three modifications: (4a) is embedded as the content of a belief ascription to make (3a); (4b) is reformulated to an identity statement to create (3b); and both (4a) and (4b) refer to the source, rather to the content, of the relevant embedded belief, thereby disguising its presence. If this is correct, then “reflective adults” and toddlers do indeed understand the aspectuality of beliefs in the Southgate et al. (2010) situation IFF they attribute agents beliefs about their own mental states, i.e., IFF they reformulate their representation of the relevant states of affairs, including their first-order belief ascriptions, into second-order belief ascriptions.

    This is not a surprise. Just like (2a) [“Sally believes that Cicero was a Roman orator.”] becomes intensional by embedding (1a) [“Cicero was a Roman orator.”] as the content of a belief, (3a) is intensional because it embeds (4a) as the content of a belief attributed to Vicky.

    But did toddlers in the Southgate et al. (2010) study have to make this extra step of producing second-order attributions? The fact that they succeeded in the task is not evidence for this as they could just rely on the inference expressed in (4), without upgrading (4a) to (3a).

    (B)

    Pierre further argues that, if the toddlers in the task carefully interpreted Vicky’s utterances, they must have realized the tension between her assertions. When she referred to the past (“Do you remember what’s in here?”) and when she labeled the toy (“There’s a sefo here”) she identified either the same toy under two different descriptions (in the false-belief condition), or referred to two different toys (in the true-belief condition). Thus, argues Pierre, successfully disambiguating her referential intent must have involved dealing with aspectuality of Vicky’s beliefs, for example, by attributing her second-order beliefs as in (3a).

    However, I think that we have good reasons to believe that they did not do so. First, Experiments 2 & 3 of Southgate et al. (2010) showed that neither labeling the toy (by ‘sefo’) nor explicit reference to the past (‘remember’) is necessary for succeeding in the task. In fact, toddlers behaved remarkably similarly in all experiments, whatever utterances they heard. This suggests that they could have just completely ignored Vicky’s speech, and interpreted only her pointing as a request (as if they had heard, “give me that one”). Even if this was the case, in order to succeed in the task, they must have taken into account Vicky’s (first-order) beliefs about the location of the toys, for example by making the inference expressed in (4).

    Second, there is no evidence that 18-month-old infants would understand the word “remember” (or “know,” which was used in Experiment 3). In fact, there is no evidence of not just the ability of entertaining the concept the word “remember” appeals to (i.e., episodic memory), but even of the ability to recall specific past episodes before about 3 years of age. Without such abilities, toddlers could not have detected the tension in Vicky’s utterances that is crucial for Pierre’s argument.

    In sum, while Pierre is right that there is an interpretation of the Southgate et al. (2010) task that requires understanding the aspectuality of beliefs, if toddlers failed to carefully analyze Vicky’s verbal utterances to make them free of contradictions, their success provided no evidence that they did ascribe aspectual beliefs to anyone. Of course, such a conclusion does not mean that other tasks could not reveal such an ability at this age, but it does not mean either that the toddlers’ success in this task would not be evidence of attributing (first order) false beliefs about object location in 18 months of age.

  • comment-avatar
    Dan Sperber 24 May 2018 (08:39)

    The Sefo task may still involve aspectuality
    In his post, Pierre suggested that children’s performance in the Sefo task could be analysed in two way: according to a richer analysis, children who succeeded in the task needed some grasp of aspectuality. According to a possible deflationary analysis, grasp of aspectuality was not involved. He suggested that further experiments based on the Sefo task might provide more direct evidence of 18-month-olds’ grasp of aspectuality. In his compelling comment, Gergo argues that Pierre’s richer analysis is not really justified. He provides a detailed simple analysis that, as claimed by Southgate, Chevallier, and Csibra, gives evidence of simple false belief attribution – but not of grasp of aspectuality. I am convinced by Gergo’s re-analysis, but, at the same time, I suspect that Pierre may have had an important and valid hunch, well- worth exploring further.
    Experimental studies of grasp of aspectuality in children by Apperly, Rakosczy and others have used as stimuli special objects having a double identity (e.g. a dice that is at the same time an eraser). But of course ordinary objects can be referred to under two different descriptions (e.g. a person can be both “the teacher” and “Anna’s mother”; a cup can be both “the cup on the table” and “John’s favourite cup”). So can ordinary actions (e.g. “hiding a toy” and “putting the toy behind a curtain”; “moving an object” and “taking an object”). So, children’s grasp of aspectuality can be investigated with ordinary objects (or ordinary actions) the two relevant aspects of which, unlike dice-erasers, do not challenge children’s ontological expectations.
    (In fact dire-erasers, bunny-carrots and other transforming toys are produced precisely because they are tricky, ontologically challenging, not the normal kind of object you interact with; this may make them tricky stimuli for uncovering ordinary cognitive dispositions.)
    Moreover – and this is Pierre’s insight –, aspectuality may already be involved in some past experiments on false beliefs without this having been investigated per se. This is likely to be the case in false belief studies where the false-believer communicates with the participant, as in the Sefo task. The two relevant aspects here may not be aspects of the object communicated about, but aspects of the act of communication.
    As Gergo points out, whatever Vicky says to the infants in the Sefo task, or even whether she communicates verbally or not doesn’t really affect children’s behaviour. What matters however is that she is communicating a request. This request properly interpreted has two aspects, one of which is based on a false belief and should be ignored, and the other, based on a genuine desire and which is to be satisfied. Children correctly understand in all cases that, by pointing at the box on the left (with or without words), Vicky is expressing her desire be given a specific toy. Presumably, they also understand her to be communicating (or at least expecting) that they should give her the toy in the box she is pointing at. To act appropriately, it is not enough that the children should interpret this second aspect of her request by attributing to Vicky a false belief about the location of the toy she wants. They must use this attribution to ignore the location aspect of her request and to fulfil its object aspect.
    So, children who respond appropriately to Vicky’s request may do so because they understand her request as having two aspects which she sees as congruent but which they understand are not. They may also, to make sense of her incongruent request, see her as thinking about the toy she wants under two incompatible aspects: the toys which she placed in the box and the toy which is now in the box. Again this goes beyond attributing a false belief ( i.e. a mismatch between a belief and reality) and involves attributing an incongruent two-aspects mental representation. Pierre’s unpacking of children understanding this duality of aspects may have been unjustifiably complex, as Gergo showed, but still, children may have the ability and in fact demonstrate it in the Sefo task. To investigate this possibility, analysis may not be enough and properly targeted experimental work may be necessary.

  • comment-avatar
    Pierre Jacob 25 May 2018 (15:01)

    The many aspects of aspectuality
    I’d like to add two things to the (to me) very instructive discussion of my initial attempt at arguing that the Sefo study provides basic ingredients demonstrating that toddlers understand the aspectuality of Vicky’s belief in the false-belief condition: an acknowledgment and a question.
    The acknowledgement first. I’m convinced by Gergo’s detailed reply that I am overestimating the toddlers’ understanding of Vicky’s speech.
    My question next. As I understand it, Dan argues that even if Gergo is right that the findings of the Sefo study do not show that toddlers understand the aspectuality of Vicky’s belief, the toddlers could not give Vicky the toy in the non-designated box unless they understood the aspectuality of Vicky’s ostensive act of pointing to the box that contains the toy that she falsely believes to be the one she wants. On the one hand, by pointing to one of the two boxes, Vicky intends to refer to a toy (based on a false belief about its location). On the other hand, she also expresses her desire to be given the toy which she falsely believes to be in the box she is pointing to. Assuming both that this is evidence of the aspectuality of Vicky’s act of pointing and that toddlers must understand it, the question is: what contribution is this understanding of the aspectuality of an agent’s communicative act likely to make to the understanding of the aspectuality of an agent’s belief?

  • comment-avatar
    Dan Sperber 25 May 2018 (15:25)

    Isn’t the aspectuality of intentions enough?
    Pierre asks: “what contribution is … understanding of the aspectuality of an agent’s communicative act likely to make to the understanding of the aspectuality of an agent’s belief?”
    Well, understanding the aspectuality of an agent’s communicative act is understanding the aspectuality of her communicative intention, i.e., of a mental state. Assuming that there is a sound argument showing that true mindreading must involve the ability to understand the aspectuality of mental states, this should do it. Is there a further sound argument showing that only the aspectuality of beliefs counts, and not that of intentions? Let’s hear the argument.

    In any case, I was speculating that, in the Sefo task, children may understand the aspectuality of the experimenter’s communicative intention, or that of her belief about the toy, or both. The interesting challenge, it seems to me, is to devise experiments to test these possibilities in a manner that does not disorient the cognitive abilities of the participants (contrary to what, I believe, experiments with ontologically challenging stimuli do).