Philosophical Studies

, Volume 173, Issue 2, pp 547–571 | Cite as

Mind-wandering is unguided attention: accounting for the “purposeful” wanderer



Although mind-wandering occupies up to half of our waking thoughts, it is seldom discussed in philosophy. My paper brings these neglected thoughts into focus. I propose that mind-wandering is unguided attention. Guidance in my sense concerns how attention is monitored and regulated as it unfolds over time. Roughly speaking, someone’s attention is guided if she would feel pulled back, were she distracted from her current focus. Because our wandering thoughts drift unchecked from topic to topic, they are unguided. One motivation for my theory is what I call the “Puzzle of the Purposeful Wanderer”. On the one hand, mind-wandering seems essentially purposeless; almost by definition, it contrasts with goal-directed cognition. On the other hand, empirical evidence suggests that our minds frequently wander to our goals. My solution to the puzzle is this: mind-wandering is purposeless in one way—it is unguided—but purposeful in another—it is frequently caused, and thus motivated, by our goals. Another motivation for my theory is to distinguish mind-wandering from two antithetical forms of cognition: absorption (e.g. engrossment in an intellectual idea) and rumination (e.g. fixation on one’s distress). Surprisingly, previous theories cannot capture these distinctions. I can: on my view, absorption and rumination are guided, whereas mind-wandering is not. My paper has four parts. Section 1 spells out the puzzle. Sections 2 and 3 explicate two extant views of mind-wandering—the first held by most cognitive scientists, the second by Thomas Metzinger. Section 4 uses the limitations of these theories to motivate my own: mind-wandering is unguided attention.


Mind-wandering Daydreaming Attention Guidance Mental action Task-unrelated thought 

I am reading a dry article or walking a familiar path to campus. While my eyes scan the page or my feet navigate the sidewalk, my mind is elsewhere. I compose a few incomplete lines of a lecture. Then? Remember, with a touch of guilt, my brash style as an undergraduate, “Was I one of the problem students teachers commiserate about?” Then? Muse that it’s sunny in California today, wondering whether my boots will survive the winter. Reminisce about that snowy February they shut down elementary school…

Mind-wandering—the mental activity illustrated in this vignette—occupies up to half of our waking thoughts (Killingsworth and Gilbert 2010; Kane et al. 2007). Mind-wandering research is also one of the fastest-growing areas of cognitive science (e.g. Smallwood and Schooler 2006; Callard et al. 2013) and scientists have stressed that the field requires firmer philosophical foundations (e.g. Christoff 2012; Callard et al. 2013; Smallwood 2013). Despite this, even passing mentions of mind-wandering are rare in contemporary philosophy. Velleman (2008, p. 174) makes an offhand remark about the practical benefits of a wandering mind. Mole (2011, p. 57) and Watzl (2011b) argue briefly over whether mind-wandering is a counter-example to Mole’s theory of attention.1 One welcome exception is Thomas Metzinger, who recently developed a groundbreaking philosophical account of mind-wandering (2013a).

This paper proposes a novel theory: mind-wandering is unguided attention. Whether attention is guided concerns how it is monitored and regulated as it unfolds over time. Roughly speaking, a person’s attention is guided when she would feel pulled back, were she distracted from her current focus. Attentional guidance is conscious and involves the agent: the guided agent experiences distractions as (in some sense) calling for correction. Yet attentional guidance is minimal in other ways: for example, the agent’s attention can be drawn to something that she genuinely wants to ignore. Mind-wandering isn’t guided in even my minimal sense. When our minds wander from one topic (e.g. lecture) to the next (e.g. the winter), we never feel pulled back. Rather, we drift on unchecked.

There are two principle motivations for my theory. One is to solve what I call the “Puzzle of the Purposeful Wanderer”. From the viewpoint of common sense psychology, mind-wandering seems essentially purposeless; almost by definition, it contrasts with goal-directed forms of cognition like reasoning, planning, or watching. Yet empirical evidence suggests that our minds frequently wander to personal goals. These results have led some psychologists to conclude that mind-wandering is (often) goal-directed after all (e.g. Callard et al. 2011).

I side with the folk, securing the core intuition that mind-wandering is purposeless.2 My solution to the puzzle draws on a distinction between two features of purposeful thought: guidance and motivation. Whereas guidance concerns how thoughts unfold over time, motivation concerns their causal antecedents. Motivated thoughts are caused by the agent’s beliefs and desires/goals. My solution to the puzzle is this: mind-wandering is never guided—which is why it is purposeless—but frequently motivated—which is why our minds often wander to our goals.

The other motivation is to distinguish mind-wandering from two antithetical forms of cognition: rumination (e.g. being fixated on one’s distress) and absorption (e.g. being engrossed in an intellectual idea). Both rumination and absorption seem stable in a way that contrasts with the drifting character of mind-wandering. Surprisingly, current theories cannot capture these distinctions. I can: when someone is ruminating on her distress or absorbed in an argument, her attention does not drift. Instead, it is drawn to her distress or held on that argument. I argue that “drawing” and “holding” are forms of guidance. Rumination and absorption are thus forms of guided attention; mind-wandering is not.

My theory improves upon its two main competitors. Most cognitive scientists define mind-wandering as thought that is unrelated to one’s current task (see Christoff 2012 for a review). I will argue that this definition cannot overcome the Puzzle of the Purposeful Wanderer. Metzinger’s (2013a) theory of mind-wandering—the other competitor—fares better against this puzzle, though he concedes more of the folk picture than I do. However, Metzinger cannot distinguish mind-wandering from absorption or rumination. And unlike me, he cannot allow for cases where we watch our mind wander.

This paper has four parts. Section 1 spells out the puzzle. Sections 2 and 3 explicate two competing theories—the first held by most cognitive scientists, the second by Metzinger. Section 4 uses the limitations of these theories to motivate my own: mind-wandering is unguided attention.

1 The Puzzle of the Purposeful Wanderer

The Puzzle of the Purposeful Wanderer reflects an apparent conflict between the folk and scientific conceptions of mind-wandering. According to our common sense picture, it seems essential that mind-wandering lacks purpose; almost by definition, it contrasts with goal-directed forms of cognition like planning a trip or solving a crossword. Consider the term ‘mind-wandering’ itself. Wandering is purposeless movement. At the risk of violating a rule we lay down for undergraduates, let me cite the Oxford English Dictionary: “to wander” means “[t]o move hither and thither without fixed course or certain aim” or “to be (in motion) without control or direction” (2008). To say that someone’s mind is wandering, then, implies that her thinking is purposeless; it is not developing toward a goal or end-point. To this extent, mind-wandering is unlike solving a crossword.

Relatedly, mind-wandering seems non-agentive—something that happens to us, not something we do. Consider that when we refer to mind-wandering, we use what I call “non-agentive” stylistic constructions, in which the grammatical subject is a not a human agent. For example, we would say “my mind was wandering” or “Luke’s mind wandered” rather than “I was mind-wandering” or “Luke mind-wandered”. In fact, we frequently use constructions in which a human agent and his wandering mind are distinct grammatical subjects: e.g. “he let his mind wander” or “letting his mind wander, Juan Pablo lay amongst the petunias”. Interestingly, as far as I am aware, contemporary psychologists are the only English-speaking population to use “mind-wandering” as a verb: e.g. “people of lower WMC [working memory capacity] mind-wandered more than people of higher WMC” (Kane et al. 2007). I believe this neologism should be avoided, since it obscures an important feature of the folk concept: when our minds wander, we don’t feel responsible for our thoughts; our minds are what wanders, not us. Contrast talk about mind-wandering with talk of goal-directed mental agency like watching, listening, planning, or reasoning. Rather than say, “Julie’s mind was watching” or “my mind reasoned”, we would say “Julie was watching” or “I reasoned”.

Thomas Hobbes’ chapter on the “Trayne of Thoughts” predicts much of our common sense picture of mind-wandering:
Trayne of Thoughts Unguided

By…Trayne of Thoughts, I understand that succession of one thought after another…This…is of two sorts.

The first is Unguided, without Designe, and inconstant; Wherein there is no Passionate Thought, to govern and direct those that follow, to it self, as the end and scope of some desire, or other passion: In which case the thoughts are said to wander, and seem impertinent one to another, as in a Dream. Such are Commonly the thoughts of men, that are not only without company, but also without care of any thing; though even then their Thoughts are as busie as at other times, but without harmony. …

Trayne of Thoughts Regulated

The second is more constant; as being regulated by some desire, and designe. For the impression made by such things as wee desire, or feare, is strong and permanent, or, (if it cease for a time,) of quick return: so strong it is sometimes, as to hinder and break our sleep. From Desire, ariseth the Thought of some means we have seen produce the like of that which we ayme at; and from the thought of that, the thought of means to that mean; and so continually, till we come to some beginning within our own power. And because the End, by the greatnesse of the impression, comes often to mind, in case our thoughts begin to wander, they are quickly again reduced into the way (Hobbes 1651/1928, 20–21; original emphasis).

Hobbes’ rich discussion of mind-wandering (i.e. unguided trains of thoughts that “are said to wander”) contains many ideas to which I will return. For now, note how closely Hobbes’ characterization mirrors our current folk picture. According to Hobbes, mind-wandering is “unguided, without design”; in other words, it is purposeless. And he defines mind-wandering in opposition to trains of thought that are “regulated by some desire, and design”; in other words, it contrasts with goal-directed thinking.

The intuitive contrast between mind-wandering and goal-directed thinking is also reflected in Branch’s Thoughts on Dreaming (1738), which may be the first Western treatise on daydreaming:

We very often find ourselves, whilst Awake, and in an indolent and musing condition [i.e. daydreaming or mind-wandering], in new and very busy [imagined] Scenes; and are no more conscious of the Design of the Soul to form them [the scenes], either by instantaneous act or a continued Series of Production, than in sleep.…[In contrast] when the imagination is exercised for any particular purpose, the Soul confines it, and rectifies or selects the Forms it brings before it; those which are approved are…preserved in Memory or…committed to writing; the others are rejected and discarded…; and then the imagination is sent in search of more; and so on. This is certainly a work of Fatigue. But, on the contrary, when we control not the imagination, but let it fly…and pursue its own Game, this costs us no pains (ibid., pp. 65–66).

Like Hobbes, Branch defines mind-wandering in contrast to goal-directed forms of cognition. Specifically, he contrasts effortful, goal-directed imagination with effortlessly letting one’s mind wander. Furthermore, Branch isolates two phenomenological differences that remain in our present-day folk picture. Suppose you deploy your imagination for some purpose: e.g. to anticipate ways a difficult conversation could unfold. Branch contends that your imagination will be subject to various kinds of control: you can call thoughts to mind, strengthen, correct, or redirect thoughts, etc. When you let your mind wander, such control is absent. And your wandering imagination is not deployed for a purpose; rather, your imagination “fl[ies]…and pursue[s] its own Game”.

On the face of it, the common sense view conflicts with empirical evidence that our minds frequently wander to unfulfilled goals (Klinger 1971). Baird et al. (2011) interrupted (i.e. “probed”) subjects at random while they performed an easy and monotonous task, and asked them to “describe anything in your stream of consciousness in the moments prior to the probe” (ibid.). Independent judges rated the thoughts in terms of their task-relatedness (on-task/off-task), temporal focus (about the past, present, or future), and goal-directedness. Thoughts were “classified as goal-directed if they included an indication of a specific goal (defined as an objective or desired result that an individual endeavours to achieve)” (ibid., p. 1606). Baird et al’s results are striking. Roughly half of the subjects’ off-task thoughts were about the future. And of those, 55 % were related to a subject’s goals.

Morsella et al. (2010) found similar results with an experimentally induced goal. Subjects in the experimental condition were told that, later in the study, they would be asked to recall the names of the fifty U.S. states. Before the quiz, they performed a “concentration exercise”: they followed taped instructions on how to focus attention on their breath. During this exercise, they were instructed to record their wandering thoughts: “every time you catch yourself being distracted…jot that thought down in the space provided…and…then bring back your attention to the exercise” (Morsella et al. 2010, p. 644). Independent judges found that, when subjects believed they were going to take the quiz, approximately 70% of their wandering thoughts were about geography (especially state names). In contrast, subjects in control conditions wandered to geography less than 10% of the time. Importantly, subjects in one control condition were told that they would later be asked to count the letters in U.S. state names. Like the experimental condition, this task is about the names of U.S. States. But unlike the experimental condition (which was a quiz), advance preparation is unlikely to make one count letters faster. One interpretation of this result is that someone’s mind will wander to a goal more frequently if she believes (on some level) that doing so will help to accomplish that goal.

We can now bring the Puzzle of the Purposeful Wanderer into view. On the one hand, a core intuition about mind-wandering is that it is purposeless: it is not directed by our goals. This intuition is reflected in the contrast we have drawn between mind-wandering and goal-directed cognition since at least the work of Hobbes and Branch.3 On the other hand, empirical research suggests that our wandering thoughts frequently advance our goals: e.g. to prepare for an upcoming geography quiz. Perhaps, then, the dichotomy between mind-wandering and goal-directed cognition (or at least the usual way of drawing it) is ill-founded.

Another side of the puzzle concerns agency. Mind-wandering seems non-agentive—something that happens to us, not something we do. But empirical evidence suggests that our wandering thoughts (often) have the right kind of causal antecedents to count as actions in a relatively strong sense. Davidson famously holds that actions—in contrast to mere behaviours—are caused (in a non-deviant way) by beliefs and desires (e.g. Davidson 1980, Chapter 4). For example, Winnie’s action of reaching into the honey pot is caused by his belief that there’s honey in the pot and his desire for honey. Because Morsella et al. (2010) used an experimental intervention to give subjects a goal, their results support causal inferences. Specifically, we can infer that having a goal and believing that thinking is a means to achieve the goal causes one’s mind to wander to that goal. This is precisely the sort of belief/desire cause characteristic of Davidsonian actions, there being no obvious sense in which this causal relation is deviant or accidental. Perhaps, then, mind-wandering is (often) a full-blooded form of agency after all.4

The Puzzle of the Purposeful Wanderer is not an idle philosophical curiosity. Indeed, I will argue that the puzzle dethrones the most prevalent view of mind-wandering in cognitive science.

2 What is mind-wandering? The view from cognitive science

2.1 Mind-wandering as task-unrelated thought

Many cognitive scientists define mind-wandering as task-unrelated thought: a person’s mind wanders when her thoughts drift away from what she is currently doing (e.g. walking to campus or reading an article). More precisely, an agent’s mental state is mind-wandering if and only if that mental state is not related to any task she is currently performing. As an operational definition, this account is widely accepted (e.g. see Christoff 2012; McVay and Kane 2009). Consider the method of ‘thought sampling’ (see Smallwood and Schooler 2006 for a methodological review). While a subject performs a laboratory task or goes about her everyday life, she might be interrupted at random intervals and asked whether her “…mind had wandered to something other than what [she] was doing” (Kane et al. 2007). Here, any thought that is unrelated to what one is currently doing (i.e. task-unrelated) is classified as mind-wandering. Ultimately, I will reject this definition; but first I need to clarify it in several ways.

Obviously, task-unrelatedness is not a sufficient condition for mind-wandering. Most non-occurrent mental states (e.g. one’s standing belief that Paris is the capital of France) are unrelated to one’s current task(s), but don’t count as mind-wandering. Psychologists avoid this problem by adding further restrictions. To begin with, mind-wandering is conscious. This is intuitive: when a person’s mind wanders, the lights don’t go out. Rather, she experiences a stream of memories, imaginings, inner speech, etc. Moreover, the accessibility of mind-wandering is what makes it possible to study using thought sampling, where a subject verbally reports whether his thoughts are on-task or off-task (Smallwood and Schooler 2006). Thus, one could argue that mind-wandering is at least “access conscious”, i.e. available for use in reasoning and the rational control of action and speech (Block 1996; cf. Smallwood 2010 for a similar argument). Two further restrictions are prevalent, although I am committed to neither. Many assume that perception cannot count as mind-wandering (Schooler et al. 2011; though see Christoff 2012). This restriction, too, has intuitive pull: when an individual’s mind wanders, she seems to become less aware of her perceptible environment. Think of the stereotypical mind-wandering professor who walks headlong into a pole, for example. Others hold that our minds wander only when we are fully awake, even if there are similarities between mind-wandering and dreaming or hypnagogic thought (the state of awareness immediately before sleep; Christoff et al. 2011; Metzinger 2013b).

One might also question why mind-wandering must be unrelated to every task one is currently performing. Alternatively, we could say that mind-wandering is unrelated to some task that one is currently performing. The latter definition has two problems. First, it rules out the intuitive possibility that someone’s mind can wander while he is at rest (i.e. in the absence of a task). Consider someone who isn’t performing a task: e.g. Juan Pablo as he lies languidly amongst the petunias. Surely, Juan Pablo’s mind may wander. Second, this definition classifies most multi-tasking thoughts as mind-wandering. Consider Elizabeth, who purposefully solves a math problem while knitting. Elizabeth’s goal-directed mathematical reasoning will count as mind-wandering, since these thoughts are unrelated to knitting.

Finally, proponents of this definition should say what it means to be related to a task. Although any account of task-relatedness will likely be fraught with difficulties, we can set those aside for two reasons. First, the methodological success of thought sampling suggests that people have an intuitive grip on whether or not thoughts are task-related. Second, I will provide a precise and clear account of task/goal-directed attention (Sect. 4.1), which one could leverage to define task-relatedness.

It is initially plausible that mind-wandering is task-unrelated thought. Indeed, this definition seems to explain the common sense intuition that mind-wandering is less purposeful than goal directed thought: goal-directed thought is related to some task that we are currently working on (e.g. solving a crossword); mind-wandering is not. Yet the fact that our minds frequently wander to goals raises decisive problems for this definition.

2.2 Task-unrelated thought and the puzzle

The Puzzle of the Purposeful Wanderer creates problems for those who define mind-wandering as task-unrelated thought. Here is a first pass at the problem: our minds frequently wander to goals like writing a chapter or taking a geography quiz (Sect. 1). But our ongoing tasks arguably include working toward such goals. Thus, mind-wandering is frequently related to ongoing tasks.

The notion of an ongoing task requires elucidation. Ongoing tasks are temporally gappy: we begin them at one time, take breaks from them, and continue to perform them at later times. In contrast, some tasks (arguably) must be performed in one continuous stream of behaviour and/or thoughts (e.g. swinging a bat or tracking a ball through the air). Long-term projects (e.g. writing a book) are paradigm examples of ongoing tasks, in my sense. Moreover, many ordinary actions (e.g. baking a cake) may also qualify as performing ongoing tasks (Tenenbaum 2010, pp. 131–132; Thompson 2008, Part II). While Martha bakes a cake, for example, she might take a break while her oven heats up to check the latest stock prices. Here’s the key move, for my argument: even though ongoing tasks contain gaps, there’s a sense in which we perform such tasks from the moment we begin them to the moment we complete or abandon them. When you ask Martha what she’s been doing over the past month, she might coherently answer, “writing a cookbook”. But if one of Martha’s tasks is to write her cookbook, then every time her mind wanders to her cookbook (which it frequently will), her mind-wandering will count as task-related.

In reply, one could distinguish between ongoing and occurrent tasks. Ordinary language permits Martha to say she’s been writing a cookbook for the past month. Arguably, however, she’s not occurrently writing the cookbook that whole time. Sometimes her occurrent task is to manage her stock portfolio. On this basis, we could clarify our definition: mind-wandering is thought that is unrelated to any of the subject’s occurrent tasks. When Martha’s occurrent task is to manage her stock portfolio, if her mind drifts to her cookbook, she thereby counts as mind-wandering (in the task-unrelated sense).

Yet even this modified definition is vulnerable to two objections, both powerful. To begin with, one could object that when Martha’s mind wanders to her cookbook, her task switches from managing her stock portfolio to planning her cookbook. Thus, although Martha’s wandering thoughts are unrelated to her former task (managing her stocks), they are obviously related to her new task (planning her cookbook).5 Admittedly, there is room to resist: in particular, I argue that mind-wandering isn’t guided in the right way to count as performing a task like planning a cookbook (Sect. 4). But we cannot make this move with resources internal to the view that mind-wandering is task-unrelated thought. We instead have to invoke a substantial assumption: mind-wandering is unguided. Why not then define mind-wandering as unguided thinking?

Furthermore, the assumption that mind-wandering is unguided does not help against the second objection. Wandering thoughts are sometimes related to the (occurrent) task from which one’s mind wanders. Suppose that my mind is wandering as I walk to campus. At some point, I recall that there’s a disguised rock in the grass through which I’m walking. Rather than stop my mind from wandering, this recollection changes the course of my wandering thoughts: I muse that it’s lucky I remembered the rock, recall the many times I’ve tripped over it, imagine myself sprawled face-first on the grass, etc. Fortified by my wandering thoughts, I successfully remember the rock when I reach it, and hop over it unscathed. Here, my wandering thoughts seem related to the occurrent task I perform as my mind wanders: namely, walking to campus. Thus, mind-wandering is sometimes related to one’s occurrent task.

We should search for a new definition of mind-wandering. One might worry that doing so undermines extant empirical results on the wandering mind. Most thought sampling studies, for instance, operationally define mind-wandering as task-unrelated thought (Sect. 2.1). I think this worry is unfounded. For starters, the pragmatic context is sufficient to clarify thought sampling questions. Most people who are asked whether “[a]t the time of the beep, my mind had wandered to something other than what I was doing” (Kane et al. 2007) will ignore the philosophical worries discussed here. The average subject won’t respond, “well, my mind wandered away from reading…but at that point, my task also switched to planning my essay…so at the time of the beep, I guess I was on-task after all”. Furthermore, the definition I will develop implies implies that mind-wandering is unguided (see Sect. 4.1), which undermines the task-switching objection (see above). So although we cannot define mind-wandering as task-unrelated thought, many (but not all) of our wandering thoughts are task-unrelated after all.

3 What is mind-wandering? Metzinger’s view

Metzinger proposes an alternative theory of mind-wandering that avoids the central difficulties facing the standard definition in cognitive science. Metzinger argues that when someone’s mind wanders, she lacks meta-awareness of and thus veto control over her ongoing thoughts. His theory solves the Puzzle of the Purposeful Wanderer, coheres with (some of) the intuitive picture of mind-wandering, and enjoys empirical support (Sect. 3.1). That said, Metzinger’s theory is too broad in one way (Sect. 3.2) and too narrow in another (Sect. 3.3). The first problem—originally noted by Ribot (1890)—also affects those who define mind-wandering as task-unrelated thought.

3.1 Metzinger: mind-wandering lacks meta-awareness and (thus) veto control

Metzinger carves out a middle ground between the common sense and scientific pictures. Although he concedes that mind-wandering can be goal-directed (2013a, p. 13), Metzinger agrees with the common sense distinction between mind-wandering and forms of mental agency like reasoning and watching. In fact, he argues that mind-wandering is philosophically significant precisely because it puts pressure on the “Myth of Cognitive Agency”, according to which “the paradigmatic case of conscious cognition is one of autonomous, self-controlled rational thought. Hard-thinking philosophers, in particular, have perpetuated this myth like a phenomenologically self-fulfilling prophecy” (ibid., p. 7).

Metzinger argues that, when someone’s mind wanders, she lacks mental agency6 over her thoughts. Mental agency, in Metzinger’s sense, has at least two necessary conditions: goal-directedness and veto control. Metzinger focuses on the latter, which he thinks is lacking when our minds wander. Veto control is a notion from cognitive science: the (person-level) functional ability to “withhold a…[behaviour]7 whose preparation and path towards execution has already begun” (Filevich et al. 2012, p. 1108). Here is an example of exercising veto control:

You’ve Got Mail…Not!: You are posting a letter, and are just about to release your grip on it and let it fall into the post box, when you suddenly get the feeling that you should check whether you put a stamp on the envelope. You tighten your grip and inspect the letter (ibid., p. 1108).

Note that you would have possessed veto control even if you had released the letter, because veto control requires only that you are able—and know that you are able— to suspend the relevant behaviour (Metzinger 2013a, p. 4).

Contrast You’ve Got Mail with a case where you lack veto control over an ongoing behaviour:

Dropping the (Bowling) Ball: You are about to bowl a ball down the lane when you lose your grip. To your horror and your friends’ amusement, there’s nothing you can do: the ball slips from your fingers and crashes behind you.

Unlike the case in which you release a letter into a mailbox, you lack veto control over dropping the bowling ball. Once the ball begins to slip, you are unable to stop yourself from dropping it behind you.

Metzinger argues that when our minds wander, we lack veto control, and thus lack mental agency, over our thoughts. Goal-directedness is insufficient for mental agency, which also requires (at least) veto control. So Metzinger can concede that mind-wandering is often goal-directed, but still draw a principled distinction between mind-wandering and forms of mental agency like reasoning, listening, or planning. When I am solving the Times Crossword, Metzinger would say, I can stop myself from doing so. In contrast, when my mind wanders to my upcoming geography quiz, I cannot voluntarily stop my thoughts. So even though the folk are wrong to think that mind-wandering is purposeless, they are right to contrast mind-wandering with examples of linear thinking like solving the Times Crossword.

To support his view, Metzinger appeals to (admittedly inconclusive) evidence that mind-wandering often unfolds without meta-awareness (Schooler et al. 2011). Meta-awareness is a higher-order mental state that is about one’s ongoing, or very recently past, mental states or processes. One example is the tip of the tongue phenomenon (“I’m sure I remember her name…”), whose object is memory. Another is the sudden realization that one’s mind was wandering: for instance, Rosewood might realize, “Shoot! My mind was wandering…I haven’t listened to a word Taggart said!”.

My reconstruction of Metzinger’s argument has two premises. First, meta-awareness is necessary for veto control over a mental state or process (Metzinger 2013a, p. 3): a person cannot know that he can terminate something he’s unaware of. Suppose Taggart notices that Rosewood wasn’t paying attention and asks, “Rosewood! Why didn’t you stop your mind from wandering earlier?!” Rosewood might respond, “I didn’t know my mind was wandering until just now!” Second, whenever a person’s mind is wandering, she lacks meta-awareness of her wandering thoughts. From these premises, it follows that people lack veto control over their wandering thoughts.8

Premise two is the crucial one, for our purposes: why think that mind-wandering unfolds without meta-awareness? Cases like Rosewood’s are suggestive: Rosewood’s mind wanders for some time before he recognizes this fact. According to Metzinger, Rosewood regains meta-awareness only after his mind has stopped wandering; only then can he look back, aghast, at how his thoughts drifted away from Taggart.

In addition, there is empirical evidence that our minds sometimes wander without meta-awareness. For example, Schooler et al. (2004) examined the relationship between mind-wandering and meta-awareness while subjects read War and Peace. Subjects in the experimental condition were a) asked to report that their mind had wandered whenever they became aware of this fact and b) randomly probed and asked whether their immediately preceding thoughts had wandered away from the passage. Crucially, 13% of the total probes caught a subject’s mind wandering before she reported this on her own. Schooler et al. (ibid.) took this as evidence that mind-wandering frequently unfolds without meta-awareness.

Here, I should forestall an obvious objection to Metzinger. My description of Rosewood is perhaps too charitable to Metzinger. Here’s an alternative description of the case: Rosewood first becomes aware that his mind is wandering. Then, exerting veto control, he brings his attention back to Taggart. So for a brief time, Rosewood’s mind wanders with meta-awareness and veto control. Surely mind-wandering sometimes ends like this (with a brief period of meta-awareness and veto control), which is enough to refute Metzinger’s view. Metzinger considers cases like this, and responds that they involve the mere “illusion of control” (Metzinger 2013a, p. 11; cf. Schooler et al. 2011). After someone’s mind stops wandering, she might retrospectively confabulate that she brought her attention back to the task. In reality, an unconscious and sub-personal process stopped her mind from wandering; she neither had, nor exercised, veto control during the mind-wandering episode itself.

Metzinger provides a plausible definition of mind-wandering. Here’s a first approximation:

Mind-wandering is conscious, waking, non-perceptual9 thinking that unfolds without meta-awareness and thus without veto control.

Nonetheless, this definition is too broad in one way (Sect. 3.2) and too narrow in another (Sect. 3.3).

3.2 Too broad: mind-wandering versus absorption and rumination

Mind-wandering intuitively contrasts with two mental activities—absorption and depressive rumination—even though both satisfy Metzinger’s definition. In The Psychology of Attention, Ribot (1890) distinguishes between two forms of distraction: dispersion and absorption.

“[D]istraction”…in our language (the French) has an equivocal sense. We call “distracted” people whose intelligence is unable to fix itself with any degree of persistence, and who pass incessantly from one idea to another….It is a perpetual state of mobility and dispersion….But the term “distraction” is also applied to cases entirely different from this…[to] people who, wholly absorbed by some idea, are also really “distracted” in regard to what takes place around them….Such people appear incapable of attention for the very reason that they are very attentive. Many scholars have been noted for their “distraction” [in this second sense]….While those whose distraction amounts to dispersion are characterized by the incessant transition from one idea to another…those whose distraction amounts to absorption are distinguished precisely by the impossibility or the great difficulty of a transfer of this (ibid., pp. 78–79).

Paradigm examples of mind-wandering seem to involve dispersion in Ribot’s sense: “a perpetual state of mobility and dispersion…characterized by the incessant transition from one idea to another”. I think Ribot is right to contrast the phenomenologies of dispersion (i.e. mind-wandering) and intellectual absorption. Intellectual absorption doesn’t involve mobility and dispersion; rather, it involves being “wholly absorbed by some idea…distinguished precisely by the impossibility or the great difficulty of a transfer…from one idea to another”. However, intellectual absorption can satisfy Metzinger’s definition of mind-wandering. People who are absorbed are conscious and awake. Their experiences are non-perceptual, in that they “…are really ‘distracted’ in regard to what takes place around them”. And paradigm cases of absorption unfold at least without veto control and, possibly, without meta-awareness. I was once so absorbed in an argument I was constructing that I failed to notice a tornado outside. It seems to me that, in such a state, I would have been unable to immediately suspend my ongoing train of thoughts (e.g. if someone told me to look at the tornado). Before doing so, it seems that I would need a moment to become less absorbed, regaining some awareness of my surroundings (and possibly, some meta-awareness that I was absorbed). Only then would I regain veto control.10

Depressive rumination presents a structurally analogous counter-example to Metzinger’s view. Rumination is a hallmark of major depressive disorder, defined as “a mode of responding to distress that involves repetitively and passively focusing on symptoms of distress and on the possible causes and consequences of these symptoms….[P]eople who are ruminating remain fixated on the problems and on their feelings about them” (Nolen-Hoeksema et al. 2008). Less extreme analogues of rumination are common: consider that for days after I received my first nasty teaching evaluation, my attention was obsessively drawn back to the negative comments and their causes and consequences.

Absorption and rumination present the same sort of problem for Metzinger. Absorption and rumination seem antithetical to mind-wandering: attention remains fixed (albeit involuntarily in rumination) on a single topic, rather than drifting from one topic to the next. One hasn’t wandered—“moved hither and thither”—if one has stayed on a single spot. Yet like mind-wandering, we often lack veto control over absorbed or ruminative thoughts. And rumination (at least) can likely unfold without meta-awareness: while I’m reading a book after receiving my teaching evaluations, I might begin to ruminate about those darned comments before I realize that I am doing so. Thus, Metzinger’s analysis is not sufficiently fine-grained to reflect the differences between mind-wandering and its antitheses.

The definition of mind-wandering as task-unrelated thought faces a similar problem. Rumination is frequently unrelated to the task one is currently performing. For instance, someone who is reading a book might find his thoughts continually drawn to the symptoms, causes, and consequences of his distress. For this reason, many psychologists classify rumination as a type of mind-wandering (see Smallwood and Schooler 2006 for a review). But again, this seems to conflate mind-wandering and its antithesis.

Here is a diagnosis of the problem: mind-wandering differs from both absorption and rumination in terms of its dynamics—in particular, how attention evolves over time. Mind-wandering is “…characterized by the incessant transition from one idea to another” (Ribot 1890, p. 79). Here, attention is unstable: the focus of attention drifts from one topic to the next. Absorption and rumination are characterized by irresistible perseveration on a single idea. Here, attention is stable: it remains fixed on the same topic over time. Hobbes similarly contrasts mind-wandering with regulated trains of thoughts on the grounds that the latter are more stable or “more constant….For the impression made by such things as wee desire, or feare, is strong and permanent, or, (if it cease for a time,) of quick return” (Hobbes 1651/1928, p. 21). Metzinger’s account and the orthodox definition overgeneralize because they abstract from a central characteristic of mind-wandering that any theory should explain: its instability. Stable trains of thoughts can proceed without veto control and be task-unrelated, just as mind-wandering can.

We have seen that the absence of veto control is insufficient for mind-wandering. I will now argue that it is unnecessary.

3.3 Too narrow: watching the mind wander

Metzinger holds that when our minds wander, we always lack meta-awareness of and thus veto control over our wandering thoughts. The empirical evidence we’ve surveyed shows that mind-wandering sometimes unfolds without meta-awareness. However, there are anecdotal counter-examples to the universal claim that it always does so:

Watching the Mind Wander: While I am reading an article, my mind wanders: I compose lines for lecture, muse that it’s sunny in California, etc. All the while, I watch my thoughts drift. I don’t actively control my thoughts, directing them toward one topic or another. Yet I am aware that I could do so; at any moment, for example, I could intervene and bring myself back on task.

Watching the Mind Wander is recognizable as a case of mind-wandering. Yet I have meta-awareness of and veto control over my ongoing thoughts. What’s more, the case is recognizably non-agentive: when I watch my mind wander, I don’t “actively control my thoughts”. Metzinger agrees that meta-aware mind-wandering (if it existed) would be passive, rather than active: one would be “mind wandering while being passively aware of this fact” (2013a, footnote 4). But then even with meta-awareness and veto control, mind-wandering is not a form of mental agency: unlike solving a crossword, mind-wandering is still something that happens to us, not something we do. By Metzinger’s own lights, therefore, his definition cannot be the whole story about why mind-wandering is non-agentive.

A closer inspection of the empirical literature suggests that such cases are common. Various studies use self-report to assess meta-awareness: e.g. subjects who catch their minds wandering are asked whether they were previously aware of this fact. For example, Smallwood et al. gave subjects the following instructions to distinguish aware and unaware mind-wandering (which they called “tuning out” and “zoning out”, respectively):

Tuning Out: Sometimes when your mind wanders, you are aware that your mind has drifted, but for whatever reason you still continue to read. This is what we refer to as “tuning out”—i.e., when your mind wanders and you know it all along.

Zoning Out: Other times when your mind wanders, you don’t realize that your thoughts have drifted away from the text until you catch yourself. This is what we refer to as “zoning out”—i.e., when your mind wanders, but you don’t realize this until you catch it (Smallwood et al. 2007, p. 533).

Across all conditions, Smallwood et al. (2007) found that tuning out was as frequent or more frequent than zoning out: if anything, therefore, it is probably more common for one’s mind to wander with meta-awareness than without (cf. Smallwood et al. 2008; Schooler et al. 2004).

In response, Metzinger could question whether self-reports of “tuning out” (i.e. mind-wandering with meta-awareness) are reliable. Metzinger, recall, believes that after our minds stop wandering, we often retrospectively confabulate that we had meta-awareness and veto control before this point (Sect. 3.1). But this response squares poorly with evidence that tuning and zoning out have different behavioural and neural profiles (see Schooler et al. 2011 for a review). One study examined subjects who reported that their minds wandered away from key passages in a Sherlock Holmes novel (in this case, passages providing a clue about the villain’s identity). Those whose minds wandered with meta-awareness (as opposed to without) were better able to deduce the villain’s identity and answer ancillary questions in a subsequent task (Smallwood et al. 2008). Metzinger owes us some account of the difference between the empirical cases in question. And it’s not clear that any such account will avoid being ad hoc.11

My diagnosis of the problem again concerns the dynamics of mind-wandering. Mind-wandering is “…characterized by the incessant transition from one idea to another” (Ribot 1890, p. 79). And such transitions are evident in Watching the Mind Wander: my thoughts drift from lecture, to California, to that snowy winter from my childhood. Watching the Mind Wander is recognizable as a case of mind-wandering due to its unstable dynamics. Again, Metzinger’s definition founders because he cannot capture the dynamics of the wandering mind. My account will explain these dynamics.12

4 My theory: mind-wandering is unguided attention

I define mind-wandering as unguided attention.13 When the mind wanders, the focus of attention drifts unguided from one topic to the next. Contrast this with attention fixed on ruminative thoughts or absorbed in a philosophical argument. Here, attention doesn’t drift without purpose or guidance; rather, it is drawn to the ruminative thoughts or held on the argument. I will argue that “drawing” and “holding” are forms of guidance; thus, rumination and absorption are forms of guided attention.

My view rests on a distinction between what I call “motivation” and “guidance”. Motivated (bodily or mental) behaviours are actions in a broadly Davidsonian (1963) sense: their causal antecedents include the agent’s beliefs and desires/goals. Frankfurt’s seminal critique of the belief/desire picture of action is that it concerns only how behaviours are initiated, not how they unfold over time (1978, pp. 157–158). But whether a behaviour is purposeful or active depends also on how it is guided while it unfolds: “Behaviour is purposive when its course is subject to adjustments which compensate for the effects of forces which would otherwise interfere with the course of the behaviour….This is merely another way of saying that their course is guided” (ibid., pp. 159–160).14 Guidance in Frankfurt’s sense involves the online monitoring and regulation of behaviour. Although our wandering thoughts are frequently motivated, they are not monitored or regulated in the right way to count as guided.

From here, we can see the contours of how to solve our puzzles. Mind-wandering is unguided and thus purposeless and non-agentive (just as the folk indicate). And yet it can also be motivated (as suggested by empirical evidence). Furthermore, there is a core distinction between rumination/absorption and mind-wandering: the former two are guided; the latter is not.

Yet our analysis cannot stop with Frankfurt’s minimal notion of guidance. For one, guidance in Frankfurt’s sense cannot capture the phenomenology of rumination or absorption. When one’s attention is drawn to ruminative thoughts or held on a philosophical argument, the experience of being drawn to or held on something is part of one’s phenomenology. But Frankfurtian guidance has no essential tie to phenomenology, since unconscious processes can be guided. For example, “[t]he dilation of the pupils of a person’s eyes when the light fades…is a purposive movement” because “there are mechanisms which guide its course” (Frankfurt 1978, p. 159). Relatedly, Frankfurtian guidance need not involve the agent. But attentional guidance is agent-involving: the agent is drawn to her ruminative thoughts or held on an argument.

I need a stronger account of guidance than Frankfurt’s. Yet this account must be weak enough to allow that rumination is guided, even though it is frequently involuntary (“I don’t mean to have these thoughts… I just can’t stop them”) and in conflict with the agent’s avowed goals (e.g. “I can’t pull my attention out of this loop and back to my essay”). I will argue that the relevant notion of guidance is exemplified by habitual behaviours like striking a pool ball.

4.1 Attentional guidance as habitual guidance

My analysis of attentional guidance is drawn from inspecting habitual behaviours. Such behaviours are guided in a thin way that nonetheless involves the agent (Brownstein and Madva 2012; Railton 2009): an agent’s behaviour is habitually guided by a norm if and only if the behaviour is the manifestation of a reliable disposition to act in accordance with the norm, such that the agent is disposed to notice, feel discomfited by, and thereby correct behaviour that seems (to her) to violate the norm.15 One example of habitual guidance is our tendency to notice when we are standing (what seems to us) too close to a conversational partner. Such violations of conversational norms cause discomfort, and thereby dispose us to take a step back until the discomfort stops (Brownstein and Madva 2012).

Habitual guidance is stronger than reliability (this will be crucial when I discuss mind-wandering). When the agent is habitually guided, she is disposed to experience errors or violations “…as calling for correction” (Railton 2009, p. 9). In this sense, habitual guidance involves the agent: to her, violations seem to call for correction. Contrast this with Railton’s reliable disposition to leave the kitchen cupboards ajar: “…if I find myself in a kitchen with self-closing cabinets, I make no effort to prevail against them” (ibid., p. 9). What distinguishes habitual guidance from mere reliability is a counter-factual condition concerning how the agent would respond, were she to violate the norm in question.

From the notion of habitual guidance, I derive a notion of attentional guidance: An agent \(\mathcal {A}\) is guided to focus her attention on some information \(\iota\) if and only if she has two dispositions:
  1. 1.

    \(\mathcal {A}\) is reliably disposed to focus her attention on \(\iota\) and

  2. 2.

    If \(\mathcal {A}\)’s attention isn’t focused on \(\iota\), she notices, feels discomfited by, and is thereby disposed to correct this fact.

Rumination is guided in this minimal sense. Rumination, recall, involves “repetitively and passively focusing on symptoms of distress and on the possible causes and consequences of these symptoms” (Nolen-Hoeksema et al. 2008). If a depressed person breaks away from his ruminative thoughts, I hypothesize, he will feel pulled back to those thoughts. This is precisely the sort of “directed discontent” (Brownstein and Madva 2012, p. 421; Rietveld 2008, Section 3) that is characteristic of how habitually guided agents respond to errors. Likewise, rumination satisfies the counter-factual condition in my definition of attentional guidance (condition 2).

Despite this, attentional guidance might be out of synch with the agent’s expressed or avowed goals. For example, someone who is depressed might sincerely say, “I don’t want to think about my distress. I just get pulled back to these thoughts.” This is no counter-example to my account: in general, habitual guidance does not require reflective endorsement. For example, many avowed egalitarians have discriminatory micro-behaviours (e.g. standing farther from black than white conversants). Regrettably, however, discriminatory behaviour can still be habitually guided: e.g. one might feel uncomfortable and step back sooner when one is talking to a black rather than a white speaker (Brownstein and Madva 2012). Although this behaviour is not directed by goals the agents reflectively endorse—i.e. avowed egalitarians don’t discriminate on purpose—it is still habitually guided. Similarly, depressive rumination is still habitually guided, even in cases where it is out of line with the agent’s goals.

Absorption is guided in a strictly stronger sense: it is goal-directed.16 Whereas attentional guidance merely involves someone who is ruminating, attentional guidance issues from someone who is goal-directed. Whereas a depressed person is guided to attend to his distress, a goal-directed person guides his attention to relevant information. I can remain somewhat neutral on how to distinguish guidance that merely involves, rather than issues from an agent. Perhaps the distinction concerns whether the guided behaviour is motivated by the agent’s goals in the Davidsonian sense I specified above (Sect. 4). Or perhaps the distinction concerns whether the guided behaviour is motivated by the agent’s higher-order goals (Frankfurt 1971).

What I am committed to is this: goals set standards for what counts as relevant to, and thus what guides, goal-directed attention. An agent \(\mathcal {A}\)’s attention is directed by one of her goals, \(\tau\), if and only if \(\mathcal {A}\) is guided to focus her attention on information that she takes to be relevant to \(\tau\); that is, \(\mathcal {A}\) has two dispositions
  1. 1.

    \(\mathcal {A}\) is reliably disposed to focus her attention on information that she takes to be relevant to \(\tau\) and

  2. 2.

    If \(\mathcal {A}\)’s attention is focused on information that she takes to be irrelevant to \(\tau\), \(\mathcal {A}\) notices, feel discomfited by, and is thereby disposed to correct this fact.


Consider the case where I was so absorbed17 in my argument that I failed to notice a tornado outside. Plausibly, my absorption would affect my attention in two ways. First, I would be disposed to focus my attention on the argument. Second, whenever my attention fell upon seemingly irrelevant information (e.g. a student dropping his books), I would notice that I was distracted and, after a moment of frustration, my attention would return to the argument. Note that when I construct my argument, I implicitly (though not always reflectively) distinguish between information I take to be goal-relevant (i.e. that pertains to my argument) and goal-irrelevant distractions. This is a crucial feature of goal-directed attention in my sense: to pursue a goal, one must have a conception of what counts as relevant and irrelevant to that goal. Conceptions of a goal often have hierarchical structure. For example, one’s conception of an algebra problem usually specifies various sub-goals (e.g. solve the sums inside the brackets, multiply these sums together) and an order in which these sub-goals should be performed (i.e. PEDMAS/BEDMAS). Goal-directed attention inherits this hierarchical structure. For example, an algebra problem solver is guided to attend to information that seems relevant to her current sub-goal, not just her overarching goal.18

In contrast, although rumination is a form of guided attention, it needn’t also be goal-directed. When someone ruminates, her attention needn’t be guided toward information that seems relevant to any of her goals. Indeed, that agent’s attention might be drawn to information (e.g. a snarky teaching evaluation) that she genuinely wants to ignore.

Mind-wandering is neither habitually guided nor goal-directed. Rather, it is a form of unguided attention: an agent \(\mathcal {A}\)’s attention is unguided if and only if \(\mathcal {A}\) is not habitually guided to focus her attention on any information. In particular, she does not satisfy the counter-factual condition for attentional guidance:
  • There is no information \(\iota\) such that, if \(\mathcal {A}\)’s attention isn’t focused on \(\iota\), she will notice, feel discomfited by, and thereby be disposed to correct this fact.

One moment, the wanderer might think about his lecture; the next moment, he might reminisce about the winter. Crucially, whenever the focus of his wandering attention shifts from one topic (e.g. lecture) to the next (e.g. the winter), he feels no discomfort drawing him back. Rather, he drifts between topics unchecked. We can also characterize my view in terms of the notion of distraction. Someone whose attention is guided would experience interruptions to her current focus as distractions. In contrast, someone whose mind is wandering would not feel distracted if her attention were to shift. Her attention would simply wander onward.

My definition of mind-wandering as unguided attention solves all three problems with Metzinger’s view. To begin with, I do not concede the core intuition that mind-wandering is purposeless. Mind-wandering isn’t guided in the right way to count as purposeful or goal-directed. Yet because mind-wandering is frequently motivated, our minds reliably wander to goals. For example, someone who is about to take a geography quiz will reliably wander to geography. This is no counter-example to my account, since we have seen that reliability is in general insufficient for guidance. Recall that Railton reliably leaves his cupboard doors open, even though he is not guided by the goal of leaving his cupboard doors open. He isn’t guided because he doesn’t experience errors as calling for correction. Likewise, the empirical fact that our minds reliably wander to goal-relevant information doesn’t imply that goals guide our wandering thoughts. Mind-wandering isn’t guided, because when our attention shifts away from a goal, we feel no discomfort drawing us back. We don’t experience these shifts as distractions that call for correction.

Furthermore, unlike Metzinger, I can explain how mind-wandering differs from rumination and absorption: the latter two are forms of guided attention whereas mind-wandering is not. I can also explain why mind-wandering is less stable than rumination or absorption. Following Williamson (2000, p. 123), I assume that stability is a modal notion: i.e., stability concerns not whether something will remain the same over time, but rather whether something would remain the same under slightly different conditions. To clarify the modal notion of stability, imagine two balls, one balanced on top of a (smooth) hill and the other settled on the bottom of a basin.19 Suppose that neither will move, since both are in an equilibrium state. The latter is nevertheless more stable, in that it will remain even if it is perturbed (e.g. by a small breeze). Hobbes also characterizes the distinction between mind-wandering and regulated trains of thoughts in terms of counterfactual stability. Regulated trains of thoughts are “more constant.…For the impression made by such things as wee desire, or feare, is strong and permanent, or, (if it cease for a time,) of quick return” (Hobbes 1651/1928, 21, my emphasis). The unguided character of mind-wandering explains its relative instability. When someone is absorbed or ruminating and her attention is perturbed, she will feel drawn back to her original topic. In contrast, perturbations to a wandering attention go unchecked. Given this, absorbed or ruminative attention is more likely to return after perturbations, and is thus more stable in Williamson’s counterfactual sense.

Finally, I allow that one can watch one’s mind wander. Such cases are counter-examples to Metzinger’s view that mind-wandering always lacks meta-awareness. Worse still, Metzinger cannot explain why meta-aware mind-wandering is non-agentive, since it unfolds with meta-awareness and thus with veto control. On my view, when someone watches his mind wander, he has meta-awareness of and veto control over a stream of unguided attention. I therefore have an easy explanation of why such cases are non-agentive: they are not guided.

Two aspects of my definition of mind-wandering deserve clarification here.20 First, given that my discussion is pitched at the level of phenomenology and agency, my definition does not rule out the possibility that mind-wandering is guided by unconscious mechanisms. Such mechanisms would monitor and regulate the course of our wandering thoughts, but post no error signals to consciousness. Proponents of such a view could maintain that mind-wandering still feels unguided to the agent, even though it is in fact guided. For present purposes, I am neutral on the prospects of this view. Note that even if mind-wandering is guided unconsciously, my view still elucidates how mind-wandering is non-agentive; for it is not subject to a type of conscious guidance that involves the agent. However, it’s unclear whether the instability of our wandering thoughts could be explained by a view that takes mind-wandering to be guided unconsciously. In contrast, views that take mind-wandering to be unguided simpliciter can explain this fact easily: mind-wandering is unstable because it is unguided.

Second, I draw a sharp distinction between mind-wandering and cases like rumination and goal-directed attention: the latter are guided; mind-wandering is not. I therefore diverge from theorists who draw gradual distinctions between mind wandering and other phenomena (e.g. Sutton 2010, p. 5; Fox et al. 2014). One might worry that because of my sharp definition, I cannot explain why certain cases of goal-directed attention (e.g. algebraic reasoning) seem less similar to mind-wandering than others (e.g. brainstorming). Yet recall that on my view, conceptions of a goal can have hierarchical structure. For example, algebra problem solvers usually know what sub-goals to perform, and in what order. In contrast, a brainstormer may have only the broadest conception of her overarching goal. Certain cases of goal-directed attention (e.g. algebraic reasoning) may thus be guided in a more structured way than others (e.g. brainstorming), and therefore seem more dissimilar to mind-wandering.

5 Conclusion

Philosophers of mind, in our neglect of mind-wandering, have ignored between one third and one half of our waking thoughts. We will learn much about the mind, if we bring these neglected thoughts back into focus. Mind-wandering, for example, occupies an intriguing middle ground between fully fledged mental agency (reasoning, planning, etc.) and unconscious, automatic cognitive processes. Indeed, two of the central features of agency—motivation and guidance—come apart when the mind wanders. And in accounting for mind-wandering, we can discover how guidance involves the dynamics and phenomenology of thinking. The wandering mind is also at the centre of a web of mental activities: rumination, absorption, goal-directed thinking, etc. And it may prove to be a test case for the philosophy of attention. Mind-wandering, then, is not merely a phenomenon that calls for explanation; it is also a new lens on the mind.21


  1. 1.

    Historical, interdisciplinary, and cross-cultural philosophers have done better: Sutton (2010) discusses early modern theories of fantasy. Regis’ Daydreams and the Functions of Fantasy (2013) lies at the intersection of English, Cognitive Science, and Philosophy (for a brief discussion, see footnotes 3 and 8). And Thompson (2015) considers mind-wandering in the context of the cognitive science of meditation. But these are still isolated discussions, not sustained debates in the literature. Dorsch’s forthcoming paper (accepted after my paper was under review) presents an intriguing philosophical theory of mind-wandering. For a brief comparison of Dorsch’s view and my own, see footnotes 4 and 18.

  2. 2.

    When I say that mind-wandering is purposeless, I do not mean that mind-wandering serves no purpose—i.e. has no function—in cognitive life.

  3. 3.

    Regis (2013) also contrasts mind-wandering with goal-directed forms of cognition such as planning and problem-solving. Against those who claim that mind-wandering (or “fantasizing”) is frequently goal-directed, Regis responds that “[a] subject may have a…[fantasy] and subsequently make a plan or use [the fantasy] in other reflections, but this does not mean that the fantasy…originally developed as a plan” (ibid., p. 24). But this response underestimates the challenge posed by empirical research. The challenge is not simply that our minds wander to thoughts that subsequently support planning or problem solving. Rather, it is that having a goal causes one’s mind to wander to goal-relevant information. Regis must explain why this causal relation doesn’t imply that our wandering thoughts “originally developed” in a goal-directed way.

  4. 4.

    The puzzle puts pressure on those who define mind-wandering as associational, and thus emphasize its unintentional character. For example, Dorsch argues that the contents of our wandering thoughts are never caused by an “underlying intention to form or have mental representations with a certain content. Rather, what is operative and responsible for the constituents and progress of the sequence is the force of association” (Dorsch 2014, Sect. 2.2; cf. footnotes 1 and 18). Empirical evidence that intentions—in a Davidsonian sense—often cause our wandering thoughts to have certain contents poses a prima facie problem for Dorsch.

  5. 5.

    Alternatively, one might propose that Martha begins to split her attention between two tasks when her mind wanders: managing her stocks and planing her cookbook. On this analysis, again, Martha’s wandering thoughts are related to one of her occurrent tasks. I thank Dominic Alford Duguid and Cliff Saron for helping me to formulate this objection.

  6. 6.

    Metzinger uses the term “mental autonomy” rather than “mental agency”. I use the latter term to keep my terminology and his consistent.

  7. 7.

    Filevich et al. originally define veto control is the ability to “withhold an action”. I have changed the definition, because veto control is (arguably) necessary for action. In that case, Filevich’s definition (trivially) implies that one never lacks veto control. My definition also departs from Metzinger’s in one respect. Metzigner’s notion of veto control is all or nothing: either “all currently ongoing processes can in principle be suspended or terminated” (Metzinger 2013a, p. 4) or a person lacks veto control entirely. This condition may be too demanding. One can arguably possess veto control over one cognitive process (e.g. attending to the drums in a jazz band), even though one lacks veto control over other concurrent cognitive processes (e.g. a stream of wandering thoughts). Filevich and I define a restricted notion of veto control over one ongoing behaviour, which is flexible enough to allow for cases like the one above.

  8. 8.

    Metzinger and Regis independently develop similar arguments. Regis argues that daydreams contrast with goal-directed cognition because daydreams proceed without meta-awareness (or “self-reflexivity”). When we daydream, we do not “…have ideas about the ideas we are having” (Regis 2013, p. 10). Against Metzinger, I will object that mind-wandering can proceed with meta-awareness (Sect. 3.3). This objection also applies to Regis, assuming that mind-wandering is a special case of daydreaming.

  9. 9.

    I discuss these restrictions in Sect. 2.1.

  10. 10.

    Ribot’s absorption is similar to at least two classic psychological constructs. One is the “flow experience” (Csikszentmihalyi 1991) characteristic of skilled activities like athletic or musical performance (see footnote 17). Another is Tellegen and Atkinson’s (1974) classic account of the personality trait “absorption”: “a disposition for having episodes of “total” attention that fully engage one’s representational (i.e., perceptual, enactive, imaginative, and ideational) resources. This kind of attentional functioning…result[s] in…imperviousness to distracting events” (ibid.). One might argue that Tellegen and Atkinson’s absorption is (at least in part) a disposition to become absorbed in Ribot’s sense. For present purposes, it is relevant that flow and absorption (in Tellegen and Atkinson’s sense) can arguably proceed without meta-awareness. Consider that the performance deficit called “the yips” seems to occur when athletes become meta-aware of their performance, which disrupts the flow state (although this is contentious: see Papineau 2013). Tellegen and Atkinson (1974, p. 274) similarly argue that meta-awareness is rare during absorption. If flow and Tellegen and Atkinson’s absorption often proceed without meta-awareness, these are empirically grounded cases where an antithetical phenomenon to mind-wandering fits Metzinger’s definition.

  11. 11.

    Metzinger could stipulate such counter-examples away. Indeed, he considers apparent cases of meta-aware mind-wandering and proposes a “terminological solution”: “intentional episodes of daydreaming, to the extent that they do involve [mental autonomy]…, thereby do not count as episodes of mind wandering, which refer only to unintentional episodes of stimulus-independent thought” (Metzinger 2013a, footnote 4). Metzinger argues (convincingly) that his stipulative definition does explanatory work: it clarifies the distinction between tuning and zoning out (ibid., footnote 4). However, it’s not clear why we should identify mind-wandering with zoning out; instead, we could say that Metzinger has clarified an important type of mind-wandering. Furthermore, we’ve seen that Metzinger’s definition does not answer a more fundamental explanandum: why is mind-wandering non-agentive? For cases like Watching the Mind Wander are recognizably non-agentive, even though they proceed with veto control. A unified explanation of why mind-wandering is non-agentive (if available) would surely be fruitful. I thank an anonymous reviewer from Philosophical Studies for pressing me to consider this response.

  12. 12.

    An anonymous reviewer from Philosophical Studies has wondered if we should characterize mind-wandering in terms of the contents (rather than the dynamics) of our wandering thoughts. I think this approach would be unfruitful for two reasons. First, a wide range of mental states can count as mind-wandering (imagining, remembering, visual imagery, inner speech, etc.). These states will presumably have widely varying contents. Second, the same sequence of mental states can unfold in a wandering way and in a goal-directed way. For example, I can compose lines for lecture, muse that it’s sunny in California, etc. while my mind wanders. Or I can call these thoughts to mind in a goal-directed fashion (e.g. in order to relive the vignette from my paper). On my view, whether this sequence of thoughts counts as mind-wandering depends on whether one is guided to token the thoughts in question.

  13. 13.

    There are deep issues about whether mind-wandering is a type of attention, which I cannot fully address in this paper. I classify mind-wandering as attention for the following reasons. When your mind wanders, the lights don’t seem to go off. Rather, you seem to have a vivid stream of thoughts, in which your consciousness is selectively focused on various bits of information (e.g. memories, internal speech, imaginings). And many theorists define attention in terms of the selective focusing of consciousness, where focus is variously characterized by enhanced vivacity (e.g. Titchener 1910), rational accessibility (Smithies 2011), a primitive structuring relation (Watzl 2011a), or something else. Given such considerations, it is natural to classify mind-wandering as attention (see Watzl 2011b for a similar argument). For example, William James claims that “the natural tendency of attention when left to itself is to wander to ever new things” (James 1890/1981, p. 422). Yet theorists who define attention in terms of attentive task performance (Mole 2011; Koralus 2014) have reason to classify mind-wandering as the reverse of attention, as distraction. Mind-wandering is not guided in the right way to count as performing a task (Sects. 2.2, 4.1). A fortiori, mind-wandering cannot count as performing a task attentively (see Mole 2011, p. 57 for a similar argument). Thus, mind-wandering may be a choice point between contemporary accounts of attention.

  14. 14.

    Other philosophers draw similar distinctions. For example, Searle (1983) contrasts Davidsonian intentions with intentions in action. Mele (1992) and Pacherie (2008) contrast Davidsonian intentions with proximal intentions. Of these philosophers, Pacherie’s account is closest to my own, since her account of proximal intentions is grounded in a notion of guidance similar to my habitual guidance. I thank Elizabeth Pacherie for clarifying my thoughts on this point.

  15. 15.

    My account of habitual guidance is adapted from Railton (2009, p. 8). Similar accounts are developed by Brownstein and Madva (2012), Rietveld (2008), and Kelly (2010), among others. In particular, my account of attentional guidance is indebted to conversations with Aaron Henry, who is writing a dissertation on perceptual activity.

  16. 16.

    Sometimes, rumination also seems goal-directed: e.g. when I obsessively consider how to improve my teaching, in light of a sub-par evaluation. I can subsume such cases under my account of goal-directed attention.

  17. 17.

    Absorption bears some similarities to the construct of “flow”. Flow occurs when “attention is completely absorbed” (Csikszentmihalyi et al. 2005, p. 601; cf. Bruya 2010) in skilled activities like athletic or musical performance. During both absorption and flow, attention is structured around one’s goals: “flow tends to occur when the activity one engages in contains a clear set of goals. These goals serve to add direction and purpose to behavior. Their value lies in their capacity to structure experience by channeling attention” (Csikszentmihalyi et al. 2005, p. 601). And during flow, attention and behaviour are subject to the kind of fine-grained adjustments characteristic of habitual guidance: “flow is dependent on the presence of clear and immediate feedback…[that] informs the individual how well he or she is progressing in the activity, and dictates whether to adjust or maintain the present course of action” (ibid., p. 601). One dis-analogy is that my examples of absorption are purely intellectual, whereas the paradigm examples of flow are sensorimotor activities. However, Csikszentmihalyi (1991, Chapter 6) argues that intellectual cases of flow can occur while reading or solving mental puzzles, for example. I thank an anonymous reviewer from Philosophical Studies for suggesting that I relate absorption and flow.

  18. 18.

    Using these resources, I can subsume Dorsch’s distinction between mind-wandering and focused daydreaming (Dorsch 2014; cf. footnotes 1 and 4). Dorsch argues that, unlike mind-wandering, focused daydreaming is a form of mental agency and possesses narrative structure (ibid., Sect. 1.5). On my view, what Dorsch calls ‘focused daydreaming’ is one among many types of goal-directed attention. Conceptions of a goal can have various types of structure, including narrative structure in Dorsch’s sense. Focused daydreaming inherits this structure: e.g. the daydreamer is guided to attend to each segment of her narrative in order. One advantage of my account is that I provide explanatory unification, showing how focused daydreaming is an instance of a more general category. Furthermore, I elucidate the sense in which focused daydreaming involves agency (it is guided and goal-directed). I thank an anonymous reviewer from Philosophical Studies for suggesting that I engage with Dorsch.

  19. 19.

    This example is adapted from Williamson (2000, p. 123). Williamson’s stability is comparable to robustness under perturbations to initial conditions. Robustness plays a central role in philosophical discussions of biological purposefulness (e.g. Thompson 2007; Walsh 2013). I thank Jonathan Weisberg for suggesting that I use Williamson’s notion of stability.

  20. 20.

    I thank Kalina Christoff and an anonymous reviewer from Phil Studies for urging me to clarify these points.

  21. 21.

    I have presented earlier versions of this paper at The Society of Philosophy and Psychology in June 2013, the University of Toronto’s Graduate Form in March 2014, Berkeley’s Wollheim Philosophy Society in May 2014, Evan Thompson’s graduate seminar at the University of British Columbia in September 2014, and the University of Antwerp’s Conference on Intentions in December 2014. I am grateful to everyone who participated in discussions on these occasions. For their comments on drafts of this paper, I thank Diana Raffman, Evan Thompson, Jennifer Nagel, Andrew Sepielli, Ronnie de Sousa, Jonathan Weisberg, Alex Madva, Kalina Christoff, Aaron Henry, Chris Mole, Dominic Alford-Duguid, Sheisha Kulkarni, Sina Fazelpour, Jelena Markovic, Kousaku Yui, and two anonymous referees for Philosophical Studies. This research was supported by funding from the Social Sciences and Humanities Research Council of Canada.


  1. Baird, B., Smallwood, J., & Schooler, J. W. (2011). Back to the future: Autobiographical planning and the functionality of mind-wandering. Consciousness and Cognition, 20(4), 1604–1611.CrossRefGoogle Scholar
  2. Block, N. (1996). On a confusion about a function of consciousness. Behavioral and Brain Sciences, 18, 227–247.CrossRefGoogle Scholar
  3. Branch, T. (1738). Thoughts on dreaming. Tully’s Head in Pall-Mall: R. Dodsley.Google Scholar
  4. Brownstein, M., & Madva, A. (2012). The normativity of automaticity. Mind & Language, 27(4), 410–434.CrossRefGoogle Scholar
  5. Bruya, B. (2010). Effortless attention: A new perspective in the cognitive science of attention and action. Cambridge: MIT Press.CrossRefGoogle Scholar
  6. Callard, F., Smallwood, J., & Margulies, D. S. (2011). Default positions: How neuroscience’s historical legacy has hampered investigation of the resting mind. Frontiers in Psychology, 3: 321.Google Scholar
  7. Callard, F., et al. (2013). The era of the wandering mind? Twenty-first century research on selfgenerated mental activity. Frontiers in Psychology, 4: 891.Google Scholar
  8. Christoff, K. (2012). Undirected thought: Neural determinants and correlates. Brain Research, 1428, 51–59.CrossRefGoogle Scholar
  9. Christoff, K., Gordon, A., & Smith, R. (2011). The role of spontaneous thought in human cognition. In O. Vartanian & D. R. Mandel (Eds.), Neuroscience of decision making (pp. 259–284). New York: Psychology Press.Google Scholar
  10. Csikszentmihalyi, M., Abuhamdeh, S., & Nakamura, J. (2005). Flow. In A. Elliot & C. S. Dweck (Eds.), Handbook of competence and motivation (pp. 598–608). New York: The Guilford Press.Google Scholar
  11. Csikszentmihalyi, M. (1991). Flow: The psychology of optimal experience. New York: Harper Perennial.Google Scholar
  12. Davidson, D. (1963). Actions, reasons, and causes. The Journal of Philosophy, 60(23), 685–700.CrossRefGoogle Scholar
  13. Davidson, D. (1980). Essays on actions and events (2nd ed.). Oxford: Clarendon Press.Google Scholar
  14. Dictionary, O. E. (2008). Oxford English dictionary online. Oxford: Oxford University Press.Google Scholar
  15. Dorsch, F. (2014). Focused daydreaming and mind-wandering. Review of Philosophy and Psychology (Advance online publication). doi:10.1007/s13164-014-0221-4
  16. Filevich, E., Kühn, S., & Haggard, P. (2012). Intentional inhibition in human action: The power of no. Neuroscience & Biobehavioral Reviews, 36(4), 1107–1118.CrossRefGoogle Scholar
  17. Fox, K. C.R., et al. (2014). Mind-wandering on the good, the bad, and the useful: Neural correlates of goal-related and positive spontaneous thoughts. Mind and Life Institute.Google Scholar
  18. Frankfurt, H. (1971). Freedom of the will and the concept of a person. Journal of Philosophy, 68(1), 5–20.CrossRefGoogle Scholar
  19. Frankfurt, H. (1978). The problem of action. American Philosophical Quarterly, 15(2), 157–162.Google Scholar
  20. Hobbes, T. (1651/1928). Leviathan, or the matter, forme and power of a commonwealth ecclesiasticall and civil. New Haven: Yale University Press.Google Scholar
  21. James, W. (1890/1981). The principles of psychology. In F. Burkhardt & I. K. Skrupskelis (Eds). Cambridge, MA: Harvard University Press.Google Scholar
  22. Kane, M. J., et al. (2007). For whom the mind wanders, and when: An experience-sampling study of working memory and executive control in daily life. Psychological Science, 18(7), 614–621.CrossRefGoogle Scholar
  23. Kelly, S. D. (2010). The normative nature of perceptual experience. In Perceiving the world (pp. 146–159). Oxford: Oxford University Press.Google Scholar
  24. Killingsworth, M. A., & Gilbert, D. T. (2010). A wandering mind is an unhappy mind. Science, 330(6006), 932–932.CrossRefGoogle Scholar
  25. Klinger, E. (1971). Structure and functions of fantasy. New York: Wiley-Interscience.Google Scholar
  26. Koralus, P. (2014). The erotetic theory of attention: Questions, focus and distraction. Mind & Language, 29(1), 26–50.CrossRefGoogle Scholar
  27. McVay, J. C., & Kane, M. J. (2009). Conducting the train of thought: Working memory capacity, goal neglect, and mind wandering in an executive-control task. Journal of Experimental Psychology: Learning, Memory, and Cognition, 35(1), 196–204.Google Scholar
  28. Mele, A. R. (1992). Springs of action: Understanding intentional behavior. Oxford: Oxford University Press.Google Scholar
  29. Metzinger, T. (2013a). The myth of cognitive agency. Frontiers in Psychology, 4: 931.Google Scholar
  30. Metzinger, T. (2013b). Why are dreams interesting for philosophers? The example of minimal phenomenal selfhood, plus an agenda for future research. Frontiers in Psychology, 4: 746.Google Scholar
  31. Mole, C. (2011). Attention is cognitive unison: An essay in philosophical psychology. Oxford: Oxford University Press.Google Scholar
  32. Morsella, E., et al. (2010). The spontaneous thoughts of the night: How future tasks breed intrusive cognitions. Social Cognition, 28(5), 641–650.CrossRefGoogle Scholar
  33. Nolen-Hoeksema, S., Wisco, B. E., & Lyubomirsky, S. (2008). Rethinking rumination. Perspectives on Psychological Science, 3, 400–424.CrossRefGoogle Scholar
  34. Pacherie, E. (2008). The phenomenology of action: A conceptual framework. Cognition, 107(1), 179–217.CrossRefGoogle Scholar
  35. Papineau, D. (2013). In the zone. Royal Institute of Philosophy Supplement, 73, 175–196.CrossRefGoogle Scholar
  36. Railton, P. (2009). Practical competence and fluent agency. In D. Sobel & S. Wall (Eds.), Reasons for action (pp. 81–115). New York: Oxford University Press.CrossRefGoogle Scholar
  37. Regis, M. (2013). Daydreams and the function of fantasy. New York: Palgrave Macmillan.CrossRefGoogle Scholar
  38. Ribot, T. (1890). The psychology of attention. Chicago: Trans. by Open Court Publishing Company, Open Court.Google Scholar
  39. Rietveld, E. (2008). Situated normativity: The normative aspect of embodied cognition in unreflective action. Mind, 117(468), 973–1001.CrossRefGoogle Scholar
  40. Schooler, J. W., Reichle, E. D., & Halpern, D. V. (2004). Zoning out while reading: Evidence for dissociations between experience and metaconsciousness. In D. T. Levin (Ed.), Thinking and seeing: Visual metacognition in adults and children (pp. 203–226). Cambridge, MA: MIT Press.Google Scholar
  41. Schooler, J. W., et al. (2011). Meta-awareness, perceptual decoupling and the wandering mind. Trends in Cognitive Science, 15(7), 319–326.Google Scholar
  42. Searle, J. R. (1983). Intentionality: An essay in the philosophy of mind. Cambridge: Cambridge University Press.CrossRefGoogle Scholar
  43. Smallwood, J. (2010). Why the global availability of mind wandering necessitates resource competition: Reply to McVay and Kane (2010). Psychological Bulletin, 136(2), 202–207.CrossRefGoogle Scholar
  44. Smallwood, J. (2013). Distinguishing how from why the mind wanders: A process-occurrence framework for self-generated mental activity. Psychological Bulletin, 139(3), 519.CrossRefGoogle Scholar
  45. Smallwood, J., McSpadden, M., & Schooler, J. W. (2007). The lights are on but no one’s home: Meta-awareness and the decoupling of attention when the mind wanders. Psychonomic Bulletin & Review, 14(3), 527–533.CrossRefGoogle Scholar
  46. Smallwood, J., & Schooler, J. W. (2006). The restless mind. Psychological Bulletin, 132(6), 946–958.CrossRefGoogle Scholar
  47. Smallwood, J., et al. (2008). Going AWOL in the brain: Mind wandering reduces cortical analysis of external events. Journal of Cognitive Neuroscience, 20(3), 458–469.CrossRefGoogle Scholar
  48. Smithies, D. (2011). Attention is rational-access consciousness. In Attention: Philosophical and psychological essays (pp. 247–273).Google Scholar
  49. Sutton, J. (2010). Carelessness and inattention: Mind-wandering and the physiology of fantasy from Locke to Hume. In C. T. Wolfe & O. Gal (Eds.), The body as object and instrument of knowledge (pp. 243–263). New York: Springer.CrossRefGoogle Scholar
  50. Tellegen, A., & Atkinson, G. (1974). Openness to absorbing and self-altering experiences (“absorption”), a trait related to hypnotic susceptibility. Journal of Abnormal Psychology, 83(3), 268–277.CrossRefGoogle Scholar
  51. Tenenbaum, S. (2010). The vice of procrastination. In C. Andreou & M. White (Eds.), The thief of time. Oxford: Oxford University Press.Google Scholar
  52. Thompson, E. (2007). Mind in life: Biology, phenomenology, and the sciences of mind. Cambridge, MA: Harvard University Press.Google Scholar
  53. Thompson, E. (2015). Waking, dreaming, being: New light on the self and consciousness from neuroscience, meditation, and philosophy. Oxford: Oxford University Press.Google Scholar
  54. Thompson, M. (2008). Life and action. Cambridge, MA: Harvard University Press.CrossRefGoogle Scholar
  55. Titchener, E. B. (1910). Attention as sensory clearness. The Journal of Philosophy, Psychology and Scientific Methods, 7(7), 180–182.CrossRefGoogle Scholar
  56. Velleman, J. D. (2008). The way of the wanton. In K. Atkins & C. MacKenzie (Eds.), Practical identity and narrative agency. New York: Routledge.Google Scholar
  57. Walsh, D. M. (2013). Mechanism, emergence, and miscibility: The autonomy of Evo–Devo. In P. Huneman (Ed.), Functions: Selection and mechanisms (pp. 43–65). Dordrecht: Springer.CrossRefGoogle Scholar
  58. Watzl, S. (2011a). Attention as structuring of the stream of consciousness. In C. Mole, D. Smithies, & W. Wu (Eds.), Attention: Philosophical and psychological essays (Vol. 145). Oxford: Oxford University Press.Google Scholar
  59. Watzl, S. (2011b). Review of Christopher Mole’s “Attention is cognitive unison”. Notre Dame Philosophical Reviews.Google Scholar
  60. Williamson, T. (2000). Knowledge and its limits. Oxford: Oxford University Press.Google Scholar

Copyright information

© Springer Science+Business Media Dordrecht 2015

Authors and Affiliations

  1. 1.Philosophy DepartmentUniversity of TorontoTorontoCanada

Personalised recommendations