Link Search Menu Expand Document

This Is a Practical Problem

If the slides are not working, or you prefer them full screen, please try this link.


My conclusion so far is that nothing adequately anchors our understanding, as researchers, of action, belief, knowledge and the rest. Not our everyday expertise as mindreaders, not the philosophical accounts of these states, and not the attempts to operationalize Theory of Mind.

The truth is that, as researchers, you and I probably do not have a shared understanding of what intention is, what knowledge is, or what desire is. And even if we do, there are probably plenty of researchers who neither share our understanding nor have any reason to adopt our way of thinking about it.

The overall question is,

Which models of instrumental action and mental states are involved in the most sophisticated forms of everyday mindreading?

You can still say, if you like, that the most sophisticated forms of everyday mindreading involve models which involve mental states like belief, knowledge, desire, anger and joy. But this doesn’t get you very far because we do not know what these mental states are.

This is a practical problem. Let me illustrate.

I will start by talking you through a practical problem that I have encountered when trying to do some research on the development of action understanding. It has troubled me for several years, and I hope you will be able to emphathize.

Illustration 1: Intention

Deceptive Intentions

According to Scott et al,

‘infants in the 2nd year of life can understand deceptive intentions’ (Scott, Richman, & Baillargeon, 2015, p. 50)

Here my problem is very simple and quite modest. I am unsure whether intention matters in this context or whether they would be just as happy to say ‘actions influenced by some kind of decepitive motive’, leaving aside claims about whether the particular mental states are intentions, desires or something else.

Residing Within

Woodward suggests that:

‘infants understand intentions as existing independently of particular concrete actions and as residing within the individual. Each of these [...] is part of what it means to understand intention in psychological terms.’ (Woodward, 2009, p. 55)

When I read this I am first struck by ‘residing within the individual’. It’s so much the metaphor of a mental state having a residence (which suppose is supposed to be just a flourish); it is the idea that a mental state might have a location.

I guess there is room for disagreement on this, but personally I find it strange to think that mental state have locations. Also none of the evidence she cites bears on this as far as I can tell. For this reason, I am tempted to think that what Woodward needs to say here is just about intentions having subjects. Some intentions are mine and others are yours; and our intentions may differ.

Similarly, ‘existing independently of particular concrete actions’ triggers some metaphysical concerns about whether mental states might be like events in being individuated by their causal relations (Davidson, 1969). But I will spare you that.

I just do not understand what Woodward means when she writes that ‘infants understand intentions’.

Unfulfilled Intentions

Moses (2001, p. 74) defends the claim that ‘a child’s concept of intention could not fully emerge before the concept of belief.’[1] He holds this for the reason that:

‘an unfulfilled intention must be accompanied by at least one false belief.’ (Moses, 2001, p. 74)

And that claim is in turn justified on the grounds that:

‘Part of what distinguishes intentions from other motivational states, such as desires, is that intentions must be consistent with beliefs.’ (Moses, 2001, p. 74)

Using these considerations to interpret a body of evidence, he arrives at the view that:

‘children of age 3 and younger may not yet have differentiated their concept of intention from their concept of desire—that at this point in development they lack an understanding of the epistemic factors (and, possibly, of the causal factors) that distinguish intention from desire.’ (Moses, 2001, p. 78)

I am not sure what to make of this. Theoretically, I do not see why an intention could not go unfulfilled even though I lack any relevant false beliefs but just because I am very unlucky. But perhaps this is irrelevant and we should take Moses’ claim as specifying the notion of intention he has in mind. My problem, then, is not just whether Moses is right; I am unsure what would count as a terminological dispute about different notions of intention and what would count as substantive disagreement.

How do Woodward and Moses and Scott et al relate?

Things go even worse for me when I attempt to relate Moses to Woodward. As I understand Moses, his position is incompatible with Woodward’s in that he would not accept Woodward’s interpretation of the evidence from infants’ abilities concerning actions (because none of that evidence concerns whether infants are sensitive to how beliefs constrain intentions). But why are these positions incomaptible? I can see at least three posibilities:

  1. Woodward and Moses are working with different notions of intention. Despite using the same word, they are talking about different mental states. (And these mental states may even belong in incommensurable models of minds and actions.) So the disagreement is merely terminological.

  2. Woodward and Moses hold incompatible views about a single notion of intention. At most one of them is right. The other has based their interpretation of the evidence on an error about the features of this mental state.

  3. Woodward and Moses hold compatible views about a single notion of intention but disagree on what is required to ‘understand’ intention or to ‘differentite their concept of intention’.

I can find no good way to decide between these three possibilities.

I encounter the same difficulty if I attempt to relate Woodward to Scott et al. Woodward seems to have a higher standard than Scott et al for postulating that infants understand intentions. Again, I am unsure whether this is because they are saying compatible things about different notions or intention, whether they are saying substantially different things about a single notion of intention, or whether they have different views on what it takes to understand something.

Illustration 2: Knowledge

According to the ‘knowledge-first’ hypothesis:

‘Rather than representing what others know by first representing what they believe, people may have a separate set of processes that give rise to some comparatively simple representation of what others know.’ (Phillips et al., 2020; see also Nagel, 2013)

An immediate difficulty in understanding this hypothesis is that we need a shared understanding of what knowledge is.

Recognizing the difficulty, Phillips et al. (2020) propose to rely on four ‘signature features that are specific to knowledge’:

(i) it is factive

(ii) it is not just true belief

(iii) it allows for egocentric ignorance

(iv) it is not modality-specific.’ (Phillips et al., 2020)

A problem several commentators note is that these four features are not in fact specific to knowledge. It is easy to identify more than two mental states with these features.

This matters because the evidence that Phillips et al. (2020) and Nagel (2013) rely on concerns observations of mindreaders’ abilities to track knowledge.

Following Starmans’ commentary on Phillips et al. (2020), there is a dilemma:

On some notions of knowledge, the evidence could not support the hypothesis.

On other notions (e.g. encountering), the evidence might support the hypothesis but the hypotheses is trivial.

The lack of a shared understanding of knowledge prevents us from evaluating the ‘knowledge-first’ hypothesis that Phillips et al. (2020).[2]


instrumental action : An action is instrumental if it happens in order to bring about an outcome, as when you press a lever in order to obtain food. (In this case, obtaining food is the outcome, lever pressing is the action, and the action is instrumental because it occurs in order to bring it about that you obtain food.)
You may encounter variations on this definition of instrumental in the literature. For instance, Dickinson (2016, p. 177) characterises instrumental actions differently: in place of the teleological ‘in order to bring about an outcome’, he stipulates that an instrumental action is one that is ‘controlled by the contingency between’ the action and an outcome. And de Wit & Dickinson (2009, p. 464) stipulate that ‘instrumental actions are learned’.
model : A model is a way some part or aspect of the world could be.
tracking an attribute : For a process to track an attribute or thing is for the presence or absence of the attribute or thing to make a difference to how the process unfolds, where this is not an accident. (And for a system or device to track an attribute is for some process in that system or device to track it.)
Tracking an attribute or thing is contrasted with computing it. Unlike tracking, computing typically requires that the attribute be represented.


Davidson, D. (1969). The individuation of events. In Essays on actions and events (pp. 163–180). Oxford: Oxford University Press.
de Wit, S., & Dickinson, A. (2009). Associative theories of goal-directed behaviour: A case for animalhuman translational models. Psychological Research PRPF, 73(4), 463–476.
Dickinson, A. (2016). Instrumental conditioning revisited: Updating dual-process theory. In J. B. Trobalon & V. D. Chamizo (Eds.), Associative learning and cognition (Vol. 51, pp. 177–195). Edicions Universitat Barcelona.
Hyman, J. (1999). How knowledge works. Philosophical Quarterly, 49(197), 433–451.
Moses, L. J. (2001). Some Thoughts on Ascribing Complex Intentional Concepts to Young Children. In B. Malle, L. J. Moses, & D. Baldwin (Eds.), Intentions and Intentionality: Foundations of Social Cognition (pp. 69–83). MIT Press.
Nagel, J. (2013). Knowledge as a mental state. Oxford Studies in Epistemology, 4, 273.
Phillips, J., Buckwalter, W., Cushman, F., Friedman, O., Martin, A., Turri, J., … Knobe, J. (2020). Knowledge before Belief. Behavioral and Brain Sciences, X, 1–37.
Scott, R. M., Richman, J. C., & Baillargeon, R. (2015). Infants understand deceptive intentions to implant false beliefs about identity: New evidence for early mentalistic reasoning. Cognitive Psychology, 82, 32–56.
Woodward, A. L. (2009). Infants’ Grasp of Others’ Intentions. Current Directions in Psychological Science, 18(1), 53–57.


  1. This claim is also defended by Davidson on different grounds. ↩︎

  2. Because Nagel offers a different way of characterising knowledge, her view does not face exactly this problem. Instead the problem facing her view is that the evidence concerning abilities to track knowledge could not support the hypothesis. This is because there are closely related hypotheses involving notions weaker than knowledge (such as encountering, facts or not being ignorant) which are equally well supported by the evidence. ↩︎