Turning on the light when entering a dark room, getting off at the right bus stop when going to work, inserting the password as we turn on our laptop are just some among the actions that we execute out of habit. Such actions are typically executed without having in mind what one is doing, which allows us to devote our attention to other tasks while going through our daily routine. Recently, habitual actions gained a substantial attention among philosophers of action because they might pose a challenge to the Standard Theory of Action (STA).

The STA explains the intentionality of actions in terms of causation by the agent's mental states, typically intentions. However, whether habitual actions are all preceded and caused by intentions is at least controversial. Philosophers such as Pollard (2006) and Di Nucci (2011) argued that the STA cannot account for habitual actions. By contrast, endorsers of the STA tried to accommodate habitual actions within the framework of the STA by appealing to non-decisional intentions (Mele 1992, Roughley 2016, Fridland 2017). Actions slips might provide an argument for the claim that (at least some) habitual actions are not caused by intentions.

Generally attributed to absent-mindedness, action slips occur when one engages in a habitual sequence of actions or as a result of this and they involve the execution of habitual actions despite the agent's initial intention to act otherwise. An often quoted example is that of a mathematician entering his bedroom to change his clothes who undresses and goes to bed instead (James 1891, Roughley 2016). Yet, paradigmatic cases of action slips include driving straight home despite an intention to stop at the supermarket on the way, or taking the tram to our workplace on a Sunday morning, which occur more often in our everyday life. An argument from slips against the STA explanation for habitual actions might go as follows1:

(1) If an action φ is caused in a non-deviant way by an intention, then φ is intentional.

(2) Action slips are not intentional.

(3) Therefore, action slips are not caused in a non-deviant way by an intention.

(4) Action slips are habitual actions.

(5) Therefore, not all habitual actions are caused by intentions.

There are two replies to this argument that I can think of. A first reply consists in denying premise (2) by allowing for inconsistent intentions. Following the above example, an agent might form a decisional intention to stop at the supermarket, but she might also form an intention to turn left at some intersection. This latter intention is automatically generated and it causes the agent to step away from the supermarket and drive straight home. I will provide here two reasons to reject this reply.

Firstly, appealing to inconsistent intentions might require us to accept the claim that agents of slips are irrational. Many would accept that intentions involve commitments. Thus, if one intends to φ than one commits herself to φ or try to do so. But if one has an intention to φ and at the same time she also has an inconsistent intention she will commit to two incompatible plans and therefore be irrational. Agents of slips, however, seem distracted and prey of their habits rather than irrational.

Secondly, it is not clear why we need to postulate sui generis sorts of intentions if we could do without. Denying that (some) habitual actions are caused by intention would require an alternative explanation. Such an explanation could appeal to causation by habits through an association between a situation and the correspondent habitual action. When an agent is in the situation to which a habit is associated, such situation triggers in the agent the performance of the correspondent habitual action. Following the above example, when the agent drives on her usual way home, she would mindlessly perform the actions that she usually performs in doing so, including turning left at the intersection, and she would do so simply out of habit2. This would likely be the cheapest explanation.

A second (and separate) way of replying might be to say that action slips are caused by intentions, yet when an agent slips she fails to act in accordance with her intention (cf. Amaya 2011). This reply does not appeal to causation by automatic intentions, yet it denies that action slips are not caused by intentions. Mistakes can be doing in the act of doing something else intentionally (Davidson, 1971). Similarly it might be said that one who does not stop at the supermarket on her way home, despite having intended to do, still acts with an intention. Yet, she fails to act in accordance with it. To this latter reply it might be objected that even though the whole action is initiated by an intention, we lack an explanation of what causes the agent to slip.

Thus, the argument from slip might not be conclusive but replies to this argument are neither. However, if habitual actions are not caused by intentions it is not clear what grants for them to be actions. Possible candidates are (i) some form of control (cf. Di Nucci, 2011) and (ii) a relation to (implicit) goals (cf. Pollard, 2006), which I will briefly discuss in turn. The forms of control Di Nucci (2011) appeals to is that of guidance (Frankfurt, 1978), a passive form of control which allows the agent to intervene if something goes wrong. Pollard (2006) proposed that habitual actions have a sort of “intrinsic” intentionality due to their teleological structure, a goal or purpose that they have which need not be made explicit. Providing a criterion for action individuation adapt to habitual actions falls however outside the scope of this paper.