The Problem of Action meets Habitual Processes

Does the fact that habitual processes and not only goal-directed processes influence instrumental actions pose a challenge to the Standard Solution to The Problem of Action? Might this fact even assist us, eventually, in developing a challenge to the Causal Theory of Action?

• youtube
• stream
• slides only

Notes

Why Focus on The Problem of Action?

From philosophy we want a framework that supports theorising about action in the behavioural and social sciences. Minimally, the framework should

• allow us to make all the important distinctions;
• enable us to formulate questions about how and why agents act; and
• support deriving predictions from hypotheses about the answers to these questions.

That, at least, is the framework we (well, mainly you^[1]) are attempting to construct in thinking through philosophical issues in behavioural sciences.

It seems reasonable to expect that any such framework must solve The Problem of Action. After all, the distinction between an action and event that merely happens to you looks fundamental. So while solving this problem is not sufficient for our aims, doing so does seem to be necessary.

Objection to the Standard Solution

The Problem of Action is to say what distinguishes your actions from things that merely happen to you (see Philosophical Theories of Action).

According to the Standard Solution to this Problem, actions are those events which stand in an appropriate causal relation to an intention (see Philosophical Theories of Action).

What counts as ‘appropriate’ here? This turns out to be a hard problem to answer. Davidson (1980, p. 79) noticed, in effect, that intentions can cause events which would not thereby count as intentional actions. We therefore cannot say simply that actions are events caused by intentions; they have to be caused ‘in the appropriate way’, whatever that is.

For our purposes (considering an objection to the Standard Solution), we need not fully specify what counts as ‘appropriate’.^[2] It is enough to notice that, for the causal relation to be appropriate, minimally:

• the action should not manifestly run counter to the agent’s intentions; and

• neither should whether the action occurs be independent of what the agent intends.

Objection to the Standard Solution: some actions are dominated by habitual processes and may therefore manifestly run counter to your intentions. For example, it is possible to continue seeking out a sweet chocolate drink instead of peppermint tea despite being sated on the drink and therefore currently preferring the peppermint tea (compare Schwabe & Wolf, 2010 discussed in Goal-Directed and Habitual: Some Evidence). Since it is irrational to intend to knowingly seek out a less preferred alternative at no greater cost than seeking a more preferred alterantive, it is possible for this action to occur counter to your intentions. Therefore not all actions do stand in an appropriate causal relation to an intention.

Other cases illustrating how habitual processes are insensitive to intentions and can therefore run counter to them about. (Wood & Rünger, 2016, p. 293) cite two:

1. ‘when students who frequently went to the sports stadium on campus were incidentally exposed to an image of the stadium, they raised their voices as they would habitually in that context, despite no change in their motivation to speak loudly (Neal, Wood, Labrecque, & Lally, 2012)’

2. ‘in a study conducted in a local cinema, participants with stronger habits to eat popcorn at the movies consumed more than those with weak habits, even when they disliked the popcorn because it was stale and unpalatable (Neal, Wood, Wu, & Kurlander, 2011).’

First Response to the Objection

In response to the above Objection, consider the possibility of insisting that in every case the agent really does have a guiding intention after all. Could there be a good reply based on this response?

Note that insisting on something contrary to what has been argued is not properly a reply to the Objection but merely a response.^[3] Your challenge is to evaluate whether or not this line of response might be turned into a way of overcoming or avoiding the objection.

Refined Version of the First Response to the Objection

A more refined version of this response might focus on the various causal roles intentions can have. As well as guiding actions directly, intentions may guide actions indirectly via habitual processes. To illustrate, you may act on an intention to get fitter by doing some exercises right after your alarm clock rings each day, and then giving yourself a little reward. Over time, if you are lucky, the habitual process may take over so that you no longer have to remember to exercise and just find yourself doing it. At this point, your actions are independent of your intentions in one sense. But the intention to exercise is still indirectly related to your actions through habitual processes. Perhaps, then, we should refine Davidson’s idea about what distinguishes actions from things that merely happen to you so that it can accommodate the ways that habitual processes can mediate between intentions and actions.

The refined response allows that actions can happen contrary to your current intentions. Perhaps you have a new partner who gets up later and would be disturbed by your exercise. Valuing the relationship over your work out, your sole concern is not to disturb them. Yet force of habit is too strong and, despite your clear intention to the contrary, you are dismayed to find yourself exercising vigorously one morning. This is a complication that the refined response should address. However, it also faces a deeper objection ...

Many habitual processes are established independently of intention. What is required is just that a stimulus is followed by an action which, in turn, is followed by a reward. These originating actions may not involve intention at all (as in utilisation behaviour), or else they may involve an intention but one that is not supposed to be habit forming. You may intentionally eat some chocolate after a meal relying on the idea that this will be a rare treat and yet, contrary to your intentions and policies, be landed with a habit. So while it is an important insight that habitual processes can mediate between intentions and actions, it also true that that habitual processes are autonomous in the sense that intentions are not necessary.

Given this objection, one might try to defend the idea that the actions habitual processes cause are actions when, and only when, they habitual processes involve intentions in an appropriate way.^[4] Is this position defensible?

Second Response to the Objection

In response to the above Objection, consider restricting both The Problem of Action and the Standard Solution to intentional action.

Could there be a good reply based on this response? It may avoid the Objection, given the further assumption that actions dominated by habitual processes are not intentional actions. But this appears to be a hollow victory. After all, what was supposed to be a bold revelation about action would, if we accepted the reply, turn out to be merely the claim that intentional actions are things that stand in an appropriate causal relation to an intention.

Further, unless we think that all actions are intentional actions (which would be hard to square with the above Objection), the original Problem of Action is still a good question. We still need to know what distinguishes actions of all kinds from things that merely happen to you.

Are there better replies to the Objection? Or can this response be developed in a much better way? If neither, should we revise or reject the Standard Solution?

We might minimally revise the Standard Solution by saying that actions are those events which stand in an appropriate relation to either a goal-directed process or a habitual process. This quite minor revision allows us to retain the Causal Theory of Action.

But can stimlus-action links and habitual processes really be relevant to solving The Problem of Action?

Further Replies to the Objection

An Anscombian Perspective

Kalis & Ometto (2021, p. 640ff) provide a critical overview of several philosophers’ attempts to reply to a variant of the above Objection. These authors propose their own response, which does involve rejecting the Standard Solution in favour of an Anscombe-inspired alternative.^[5]

If exploring further work by philosophers, be careful to check whether their understanding of habitual process matches yours. You can tell that this will be tricky from the fact that Kalis & Ometto (2021, p. 640ff) write about ‘habitual actions’, whereas, strictly speaking, no such things exist on the dual-process theory of instrumental action (as explained in The Minor Puzzle about Habitual Processes).

In fact, Kalis and Ometto do appear to have a view on which actions dominated by habitual processes are not actually (intentional) actions:

‘The concept of an intentional action, then, is essentially the concept of a kind of behavior that makes sense to the agent as her action. An intentional action is action that is, therefore, partially constituted by the agent’s point of view, or her own take on what she is doing. (Kalis & Ometto, 2021, p. 644)

Basic Actions?

A different line of response might be to appeal to so-called basic or primitive actions, that is, actions which you can perform without performing any other action (Davidson, 1971).^[6] In cases like popcorn eating where, supposedly, actions can run counter to any intention, consider that there is a distinction between the larger action (eating popcorn) and component actions like reaching for some popcorn, grasping it, transporting it to the mouth and eating it. Regardless of whether the larger action runs counter to any intention, might these component actions nevertheless be appropriately related to the agents intentions? If so, could we revise the Standard Solution to avoid the Objection above?

Discoveries about motor representation (see Motor Representation) complicate this line of response in two ways—they make it harder to characterise actions like reaching and grasping as basic actions, and they indicate that may be no need to postulate intentions concerning these actions specifically (as they are already well taken care of by motor representations).

Bonus Dangling Question: Alternative to the Causal Theory?

According to the Causal Theory of Action, an event is action ‘just in case it has a certain sort of psychological cause’ (Bach, 1978, p. 361). If we retain the Causal Theory and if we also accept that some actions are dominated by habitual processes and may therefore run counter to your intentions, then we will have to invoke not only beliefs, desires and intentions but also stimlus-action links in distinguishing actions from events that merely happen to you.

This may motivate considering alternatives to the Causal Theory.

Consider two questions:

1. What distinguishes instrumental actions from things which merely happen to an agent (and from noninstrumental actions, if there are any)? [This is ‘The Problem of Action’]

2. Which states cause instrumental actions?

Fully understanding action requires answering both questions (and more).^[7] But the Causal Theory of Action insists on answering the first question in a way that also involves answering, partially or wholly, the second. The idea is not simply that better understanding answers to the second question might guide us in working out the answer to the first question. On the Causal Theory of Action, any answer to the first question must already involve answering the second. There is no possibility, not even in principle, of answering the first question correctly but then discovering that everything we thought we knew about the second question is wrong.

Let us say that any answer to the first question which does not involve making commitments concerning which states, or structures of states, cause instrumental actions is mechanistically neutral (as opposed to a mechanistically committed answer, which the Causal Theory of Action requires).^[8]

If we reject the Causal Theory of Action, we will need a mechanistically neutral solution to The Problem of Action. What might that be?

Ask a Question

Your question will normally be answered in the question session of the next lecture.

More information about asking questions.

Glossary

References

Endnotes