Human Agency, Reasons, and Inter-subjective Understanding

William Hasselberger. Philosophy. Volume 89, Issue 1. January 2014.

The Standard Story of Action and Reasons-Explanation

Mainstream action theory proceeds, roughly, in the following three steps: (1) What distinguishes intentional actions (e.g., dancing) from non-intentional bodily motions (e.g., a leg spasm) is the fact that the former, but not the latter, are performed for reasons, i.e., intentional actions spring from the agent’s first-person perspective on what-to-do (on what’s sensible, worth doing, from the standpoint of agency); (2) the perspective/reasons for which an agent acts, her ‘motivating reasons’, are internal mental states or mental representations with a certain ‘content’ that the agent seeks to bring about or produce; and (3) the role of the agent herself in acting can be reduced to that of certain mental states—often called ‘desires’, ‘belief-desire-pairs’, or ‘intentions’—causally triggering the bodily motions that are her actions. Call this:

The Standard Story of Action: intentional actions are bodily motions causally triggered in the right way by certain internal mental-representational states—i.e., the bodily movements count as actions in virtue of this psychological-causal history.

Actions are, minimally, motivated self-movements; and motivation consists in the presence within an agent’s ‘psychic economy’ or ‘psychic stew’ of causally-efficacious internal mental states that possess semantic or representational content and that dispose the agent to alter the world to ‘fit’ their representational content. These internal states (i) are the crucial elements in (causally) moving or pushing the agent from a state of inactivity and (ii) they also rationalize the agent’s behavior by supposedly revealing what the agent saw in so acting, why she saw her action as fit to choose.

What internal mental states are the ‘right’ action-triggers? Broadly, propositional attitudes: a Standard Story theorist may hold that ‘belief-desire-pairs’ trigger action directly, or by way of causing the formation of a distinct psychological state of ‘intention’ that then causes the bodily motions—or that the chain includes some further type of formally-specifiable internal state that grants the behavior the full status of being ‘intentional’, ‘autonomous’ action, or action ‘par excellence’. The action-theoretic ‘game’, for the Standard Story, is then arguing over which internal psychological states, with what particular functional and causal roles, are necessary and sufficient for the task: this is the general approach to human agency associated with Donald Davidson and more recent exponents such as Michael Bratman, Alfred Mele, Michael Smith, and David Velleman, among others.

The Standard Story is primarily offered as an account of the underlying nature and structure of intentional action, but it implies—and is often motivated by—a particular account of our everyday practices of describing, explaining, and understanding one another’s actions: an account of so-called ‘intentional’, or ‘reasons-based’, or ‘commonsense’ psychological explanation and understanding.

A bit of bodily behavior—say, my reaching for a cylindrically shaped object containing liquid, bringing it to my lips and ingesting its contents—can be ‘explained’ or ‘understood’ in a variety of ways involving different explanatory requests and citing different ‘reasons’ or explanatory factors. One kind of explanation of my reaching/grasping behavior would cite facts from neurophysiology (electrical discharges in my brain, the release of sodium and calcium in my muscle fibers), another kind, facts from evolutionary biology, or ethology.

A different and seemingly distinctive, yet perfectly familiar form of action explanation and understanding cites the agent’s reasons: those considerations for which, or in light of which, the agent acted (e.g., ‘Because I’m thirsty and here’s some water’). Citing the agent’s reasons seems to explain her bodily behavior in a distinctive way in that it, unlike (say) neurophysiological explanation, purports to reveal the purpose or point or rationale of the behavior, from the agent’s perspective. As many put it, everyday explanations of behavior in terms of the agent’s reasons make the behavior rationally intelligible as an action: this explanatory-interpretive perspective enables us, in Davidson’s phrase, to see the human animal ‘in his role as Rational Agent’ and to thereby understand the movement of his or her body as an exercise of agency and practical reasoning, rather than a meaningless motion on par with a spasm or hiccup. Our commonsense practices of asking questions like ‘Why did you do (are you doing) that?’,’What’s the point of doing that?’, ‘What were you after?’ (and other cognate forms), all seek to make an action intelligible by revealing the agent’s reasons for so acting. This apparently distinctive way of understanding human activity calls for adequate philosophical construal.

The Standard Story construes our everyday practices of explaining and understanding others’ actions as a practice of ascribing to others ‘motivating reasons’, qua internal, representational (propositional-content-bearing) mental states like belief-desire-pairs. The hope is that by attributing these internal states to an agent we can, looking back, interpretively understand the ‘point’ of her outward behavior, and, looking forward, predict her future behavior.

The Standard Story of Action-understanding: understanding another agent’s action, in the ‘intentional’ or ‘reasons-based’ sense, consists in ascribing certain internal mental states to the agent: those that can be plausibly be seen to rationalize and cause her outwardly observable bodily motions.

The Standard Story’s claims about motivation, rationalization, and inter-subjective understanding fit neatly together: motivation travels from the internal psychic economy of mental states (proxies for the agent and her perspective, whose ultimate natures are to-be-discovered by an ideal neurophysiology) to the ‘outer’, publically observable domain of bodily motions; and explaining an action in terms of the reasons for which it was performed re-traces this course: from the outwardly observable bodily motions to the ascribed internal states that (purport to) rationalize and cause them.

The Standard Story is often treated as a basic tenet of a ‘naturalistically respectable’ account of human mentality and the metaphysics of human action and agency, one according to which actions are ontologically reducible into composites of more fundamental, non-agential components: mental states, bodily motions, and causal relations/profiles. The hope is that the Standard Story can thereby show how human agency can be located within an ontologically austere picture of the world that need not recognize metaphysically-irreducible ‘human agents’, irreducible teleology, or sui generis ‘powers’ or ‘capacities’ of human agency.

Though these elements of the Standard Story—claims about motivation, reasons-explanation, and the naturalistic metaphysics of agency—may be, in a strict sense, separable, they collectively compose a highly influential picture of the nature of human action. The coherence of the overall picture, and its promise of parsimony and naturalistic respectability, lends it particular force.

In what follows I’ll argue that the Standard Story involves a deeply distorted conception of how we perceive and understand the actions of others in everyday life. I’ll argue, drawing on themes from Elizabeth Anscombe and Gilbert Ryle, that the problem lies in certain core assumptions of the Standard Story; so I will neglect various finer-grained distinctions among theories grouped under this rubric.

The Direct Perception of Psychological Meaning in Action

Notice: by decomposing actions into bodily motions and ‘motivating reasons’ that trigger motion, and then treating motivating reasons as internal mental states, the Standard Story implies that we never really directly perceive another agent’s reasons, motivations, intentions (much less their efficient-causal powers, or, absent an ideal neuroscience and fMRI machine, their neurophysiological ‘realizers’). What is there to be directly ‘seen’, on the Standard Story, is a series of not-intrinsically-intentional bodily motions—the event or occurrence of such motions in public space—which may or may not be actions (depending on their causes) and must be rendered intelligible via interpretative ascription of internal psychological states to the agent (governed by a ‘folk’ understanding of how such states relate to one another and tend to produce behavior). On this account, then, parties to the inter-subjective encounter must ascribe internal states to one another to achieve any workable psychological understanding.

Both Anscombe and Ryle (like their shared point of influence, Wittgenstein) claim that:

(i) We cannot understand human actions as bodily occurrences that are causally triggered by certain distinct, internal psychological states.

(ii) Psychological terms like ‘intention’, in their everyday, pre-theoretical uses, don’t refer to internal mental states of which only the subject can be directly aware.

(iii) Our characteristic and conceptually primary inter-subjective understanding of human action is not constituted by a procedure of ascribing to other agents unobserved internal mental causes of their outwardly observable (not-intrinsically-intentional) bodily motions.

Ryle disparagingly calls this account ‘psychological divining’ of ‘para-mechanical’ causes, and says that this whole idea rests on ‘the mythical bifurcation of unwitnessable mental causes and their witnessable physical effects’. And Anscombe too emphasizes that giving a reason in answer to a ‘Why?’ question, in the distinctive reasons-seeking sense, is not giving predictive ‘evidence’ that the behavior will occur nor is it citing any distinct internal ‘mental cause’. Agency and inter-subjective action-understanding must be construed in a fundamentally different way.

Consider the following phenomenological questions: What is the character of an intelligible experience of human agency in normal, everyday contexts? What’s it like to encounter and understand the actions of others in mundane situations?

Imagine walking into a busy local restaurant to pick up a carry-out order. Upon entering, you’re presented with the familiar bustling sights, sounds, smells: people are seated at tables eating and conversing; waiters taking orders; busboys shuttling about with trays; a line of people waiting to pay at the register. After a brief, no-nonsense interaction with the cashier, who smiles politely as you pay and pick up your food, off you go.

The phenomenology of encountering and understanding the actions of others in everyday practical affairs such as these—something we tend to overlook since it occurs so naturally and spontaneously—has the following character. Our characteristic experience of other agents’ behavior is not of bodily motion but of immediately intelligible and structured human activity falling under directly perceptible patterns. Just as we typically and characteristically perceive speech in our native language as immediately meaningful, without first perceiving meaningless noise that we then interpret as meaningful, so too we typically perceive human behavior as patterned and unified, and immediately meaningful. In the run of everyday affairs we don’t primarily perceive humanoid bodies undergoing changes in position. Characteristically, human behavior in ordinary contexts is primarily encountered as structured, intentional, and psychologically significant from the start: as a unity of intelligible activity that falls under a perceptible and meaningful pattern or type of action: ‘eating’, ‘conversing’, ‘ordering food’, ‘paying at the register’, ‘busing food to a table’ (all concepts firmly at home in what Sellars called the pre-scientific ‘manifest image’). Importantly, we may not even be able to give a description of what is directly perceptually present to us in non-psychological, purely ‘behavioral’ or physiological/geometric terms: imagine trying to give an account in non-intentional, purely physiological and geometric terms of what is seen when you see two people eating dinner and conversing (a mess of jaw, lip, eye, hand motions is obviously insufficient).

This direct and non-inferential phenomenology of inter-subjectivity is perhaps clearest in the perception of others’ emotions: as Wittgenstein points out, we typically perceive the emotion in another agent’s face or behavior immediately, i.e. without having to infer the presence of the emotion from an initially emotion-neutral and unstructured presentation, like the physical-geometrical shape of another’s mouth, eyes, and hand motions, or something of the sort.12Likewise, in ordinary contexts, our grasp of the psychological meaning of another’s ‘outward’ behavior seems phenomenologically immediate: it is not the product of first interpreting a presentation of the behavior that is initially non-intentional and unstructured—e.g. a sequence of bodily motions and gyrations that, in itself, is meaningless (it could be a mere happening)—which we then interpret as triggered by, and meaningful in virtue of, postulated internal states of the agent. Rather, intentional actions, like emotionally-expressive faces, are typically phenomenologically present as directly perceptible psychologically-meaningful patterns of activity such as:

(a) intelligible types of action (eating, conversing, dancing, cooking),

(b) social roles being enacted by agents (teacher, traffic cop, cashier, bus driver).

And, characteristically, we don’t encounter agents in ‘blank’ and practically ‘neutral’ locations of detached observation, but rather within:

(c) pragmatic situations or contexts that themselves have meaning in terms of a range of intelligible action possibilities (grocery stores, classrooms, airports, gyms).

Throughout much of daily life, we can understand what an agent is doing in terms of these informationally-rich agential patterns: patterns that are often perceptually apparent to a suitably situated and constituted interpreter. Call this ‘mundane action intelligibility’: I just see people shopping in the grocery store, working at a computer in the office, eating in the Chinese restaurant, driving a taxi, mowing their lawn, etc. And, crucially, nothing is immediately puzzling about the activities unfolding, they don’t require the unmasking of anything hidden: they’re intelligible elements of a familiar world of patterns of practice and interaction. In mundane and familiar contexts, it’s often not necessary to gain a particularly detailed grasp of the agent’s practical perspective in order to make sense of what she’s doing. Often it’s enough for practical engagement and interaction to grasp the perceptible action-type, or role-enactment, or purposively-structured situation, and thereby to grasp some range of intelligible interactive possibilities those perceptible patterns make apt.

I’ll take these phenomenological data—the seemingly direct and non-inferential perception of bodily actions as structured, meaningful instances of intelligible action-patterns—as given, as everyday data for theorizing. Though the Standard Story is not primarily an account of the phenomenology of action and action-understanding, but rather of its essential conceptual structure, it still must make sense of the pre-theoretical experience, the important features of the manifest image—and an inability to do so would be a serious flaw.

Psychological Understanding as Knowing-How

If the phenomenology of ordinary, inter-subjective interactions typically presents, not bodily motions requiring explicit Standard Story interpretation, but actions with ‘outwardly’ perceptible psychological meanings, what explains this? Consider:

Anti-Psychologism: the perception of psychological meaning is a distinctive practical skill: the knowledge-how-to recognize or discern inherently meaningful patterns and structures of human activity. This practical-perceptual knowledge-how, like our ability to directly perceive emotions, is a basic element of human sociability, and cannot be grounded on or reduced to the Standard Story.

Anti-psychologism may seem unfamiliar in the contemporary context. Both Anscombe and Ryle criticize modern philosophy for having an overly ‘contemplative’ and ‘intellectualistic’ understanding of knowledge that misses distinctively practical forms of knowing. When we learn how to do something—e.g., how to read Swedish, how to play guitar, how to surf, how to cook on an open flame—we don’t just know that certain stateable facts are true (after all, we can know that playing guitar involves such-and-such without knowing how to play at all). Rather, we have a certain practical mastery, skill, competence. On Ryle’s influential account, knowing-how is a sui generis form of distinctively practical knowledge or competence that (i) cannot be reduced without loss to the knowledge of certain propositions (what Ryle calls ‘knowledge-that’); and (ii) does not require (and typically doesn’t involve) applying ‘regulative propositions’ to particular cases. Indeed, Ryle argues that intelligently or properly applying propositions, rules, or principles to cases itself requires, on pain of regress, a form of knowledge that’s not itself knowledge of a further proposition or rule—it requires knowledge-how, practical knowledge.

Ryle applies the idea of non-propositional knowing-how to inter-subjective psychological understanding itself:

If [psychological] understanding does not consist in inferring, or guessing the alleged inner-life precursors of overt actions, what is it? … [A] spectator who cannot play chess also cannot follow the play of others; a person who cannot read and speak Swedish cannot understand what is spoken or written in Swedish … Understanding is a part of knowing how. The knowledge that is required for understanding intelligent performances of a specific kind is some degree of competence in performances of that kind.

The claim is striking: psychological understanding—including, centrally, action-understanding—is a form of non-propositional knowledge-how; and this know-how involves the interpreter sharing a degree of ‘competence’ in the observed performances and practices with the agent being interpreted, akin to the mutual grasp of a language or a game. Action-understanding is made possible by the actor and interpreter sharing a body of knowledge-how—a kind of general practical sensibility—that enables mutual discernment of meaning:

In making sense of what you say, in appreciating your jokes, in unmasking your chess-stratagems, in following your arguments and in hearing you pick holes in my arguments, I am not inferring to the workings of your mind, I am following them. Of course, I am not merely hearing the noises you make, or merely seeing the movements you perform. I am understanding what I hear and see. But this understanding is not inferring to occult causes. It is appreciating how the operations are conducted.

As I read him, Ryle is describing a kind of practical-perceptual attunement among interpreter and agent: given the right knowledge-how, I can directly ‘understand’ what the agent is doing without, first, ‘merely seeing the movements [the agent] perform[s]’ and, second, ‘inferring to the workings of [the agent’s] mind’. The relevant knowledge-how is a practical-perceptual capacity to discern patterns of psychological intelligibility or significance: a capacity to see what the agent sees in her action.

This has a consequence that stands in radical opposition to the Standard Story: an agent’s intention—the purposive character of her behavior—can be a directly perceptible feature of the everyday world, for the appropriately situated and constituted interpreter. The claim is not that our inferences about mental causes are (simply) very quick: it’s a rejection of the Standard View’s account of what can be present or given in our experience of other people’s actions. ‘I am sitting in a chair writing, and anyone grown to the age of reason in the same world would know this as soon as he saw me, and in general it would be his first account of what I was doing’, writes Anscombe. The description under which the action is intentional, from the perspective of the agent, and the intrinsically meaningful pattern via which the action is directly presented in the experience of the interpreter, can converge. In the veridical case, the (knowledgeable) perception of the embodied action of another agent (qua intelligible action-type or social-role enactment) is, at one the same time, perception of intention.

It’s a perfectly general point that, with the right practical-perceptual knowledge-how, an agent can directly perceive meaningful features of the world that are there-to-be-perceived, but are only apparent given the relevant practical-perceptual capacities and conceptual resources. An experienced botanist can directly see oriental bittersweet among the grapevines, where I just see a mess of undifferentiated ‘plants’; and a properly trained musician can hear the minor seventh chord and the return to the root, where others hear ‘piano.’ In a crude and thin sense of ‘perceive’, these subjects perceive the ‘same’ things: the same shapes and colors, the same sound-pressure changes. But the former, through training and the acquisition of practical-perceptual skill, are alive to the meaning and intrinsic structure of what they perceive, the latter not.

In the same way, Anti-psychologism claims that mundane forms of psychological meaning and intelligibility—emotions, unfolding types of actions, social roles, familiar situations of practice—are there to be directly perceived, given the right perceptual skills and knowledge-how. In successful exercises of this knowledge-how, the agent’s intention is directly perceptually evident to the skilled observer (without her needing to make an internal state ascriptions/inferences from an initial presentation of mere bodily motions). And directly perceiving mundane psychological meanings in what others are doing is often sufficient for the kind of everyday action-understanding required for ordinary practice and interaction among agents. (These claims do not imply the absurd view that the actions of other agents are ordinary totally transparent to us, or that we can’t be mistaken or corrected, or that others cannot hide their intentions: I’ll return to problematic cases below).

Knowledge-how, Psychologism, and Infinite Regresses

Phenomenology is not decisive, however. The Standard Story can accept that it often seems like (phenomenologically speaking) we directly, non-inferentially perceive psychological meaning in other’s embodied actions; likewise, the Standard Story can accept that, in veridical cases, it seems like we directly, non-inferentially perceive behavior as structured from the outset by the very same description under which it’s intentional. Nonetheless, for the Standard Story, this seemingly ‘direct perception’ is not basic or fundamental.

The Standard Story Redux: the seemingly ‘direct’ or ‘immediate’ perception’ of meaningful action patterns results from the perception of some bodily-motion-facts B, and the rapid and possibly non-conscious, post-perceptual ascription of mental-state facts M. (The fact that this interpretive process does not show up in our conscious experience is largely irrelevant to the truth of the theory.)

Herein await serious problems for the Standard Story. First, the response suggests a radical constriction of what we can, intuitively speaking, perceive or experience—and, consequently, a radical inflation of what we must infer and ascribe. If I cannot directly see someone intentionally playing poker, say (but only infer this rapidly/non-consciously from the perception of body motion facts), then it seems I can’t see people playing games, conversing, cooking, resting, or smiling happily or scowling angrily (rather than making shapes with their faces). But why accept such counter-intuitive claims: what’s the motivation, other than to save a problematic theory?

More problematically, the above ‘rapid/non-conscious inferences’ response gives rise to a dilemma: either it (a) presupposes the kind of practical-perceptual knowledge-how central to Anti-psychologism, or (b) it falls into an infinite regress. Consider: if the seemingly ‘direct’ perception of psychological meaning (e.g., of an agent performing a certain type of action, or enacting a certain role) is always itself the upshot of prior internal state ascription, we may ask, on the basis of what are such internal psychological states initially ascribed? The Standard Story interpreter is going to have to, on pain of infinite regress, immediately discern some features of the perceived situation and perceived behavior as relevant to the interpretive process. On pain of regress, certain features of what’s encountered have to be seen, prior to the ascription (theorizing, simulating), as salient and making such-and-such range of psychological states with such-and-such content even possibly applicable and relevant. Otherwise the interpretive process could never so much as get started. How could we possibly know when to start, what to look at, what is relevant to the interpretation (or whether to interpret, here and now, at all)?

Recall the example above of a trained musician hearing a minor-seventh chord and a return to the root, while someone else hears, simply, ‘piano playing’. It’s not that the two individuals have the ‘same’ experience (this is only true in a highly attenuated sense) and then the musician judges or infers that such-and-such is a minor-seventh chord or a return to the root. Not only is this a totally distorted description of the phenomenology (phenomenologically, certain chords are simply perceived as such, from the start); more importantly, any ‘judgment’ of this kind would already presuppose a particular, perceived item, picked out from the perceptual flux, as it were, as fit or apt for the judgment. This is already a practical-perceptual discrimination—i.e. selective focusing on and bringing-into-view some particular thing under some intelligible aspect. It can’t be brute unconceptualized ‘given’ and then conceptual inference, if the inference is to have any rational grip. In the same way, any Standard Story interpretation will itself presuppose some prior and immediate perceptual-recognitional knowledge-how, some set of capacities and skills to see what is immediately relevant and significant—out of the infinite array of features and properties instantiated by any given situation and series of bodily motions—and thereby to get the mental-state ascription process started. Even for the Standard Story, an observer has tofirst exercise practical-perceptual knowledge-how to bring an ordered situation into view and knowledgably recognize the perceptual cues of relevance in a situation; only then can she ascribe to others some internal mental states (belief-desire-pairs, etc.) that those cues suggest; and only then, finally, can she arrive at a provisional sense-making interpretation that sheds rationalizing light on what is seen.

Therefore: even the Standard Story can’t rely on psychologistic inferences all the way down, it has to presuppose and rely on knowledgeable perceptual sensitivity to a whole array of situational relevancies/saliencies even to get started. But once we recognize this essential role of practical-perceptual capacities (even for the Standard Story), then the supposed necessity of an additional step of internal state ascription is undermined—as is the Standard Story’s alleged virtue of parsimony.

This opens the door to the (Rylean) knowledge-how-based Anti-psychologism: the view that suitably constituted observers can directly perceive psychological meanings in outward behavior, such as the type of action intentionally performed, the social role enacted, or the pragmatic situation at play. For Anti-psychologism, in everyday cases, this is a one-step process of perceiving mundane action-meanings (not a multi-step, outer-to-inner psychologistic process). The upshot is that perceptual knowledge-how enables the interpreter to, at once, bring the action and the intention into view, from out of a limitless number of other true descriptions of the situation—rather than interpretively ‘constructing’ the intentionality of others’ behavior out of pre-action elements of bodily motions and posited internal states.

The Perception of Unity in Unfolding Action

Anti-psychologism also explains a further, related feature of our experience of psychological meaning in human behavior. In a key section towards the end of Intention, Anscombe writes:

Consider a question ‘What is the stove doing?’, with the answer ‘Burning well’ and a question ‘What is Smith doing?’, with the answer ‘Resting’. Would not a parallel answer about Smith really be ‘breathing steadily’ or perhaps ‘lying extended on a bed’? Someone who was struck by this might think it remarkable that ‘What is—doing?’ should be understood in such different ways…And ‘resting’ is pretty close to lying on a bed; such a description as ‘paying his gas bill’, when all he is doing is handing two bits of paper to a girl, might make an enquirer say: ‘Description of a human action is something enormously complicated, if one were to say what is really involved in it—and yet a child can give such a report!’

Notice there are an infinite number of true statements that one could make about what Smith is ‘doing’ lying there (he’s pumping blood, exerting gravitational force on surrounding objects, displacing less than a million liters of air). Yet we characteristically converge on the description—or, in more phenomenological terms, the ‘presentation’—under which it’s an intentional action: he’s resting. Just below this passage, Anscombe says that her reasons-seeking ‘Why?’ question ‘can be looked at as a device which reveals the order in this chaos’, and later says that ‘the term “intentional” has reference to a form of description of events’ (her emphasis). She gives the examples ‘building a house’ and ‘writing a sentence on a blackboard’ as ‘concepts of human action’ with this form. Any description of this kind requires some immediate, non-theoretically-based (or inferentially-based) grasp of what is relevant to get off the ground.

Notice that when we perceive and understand behavior as the unfolding of some particular action-type—’cooking’, ‘taking an order’, ‘paying at the register’—we perceive a diverse range of physically and temporally discrete bodily motions as exhibiting a kind of unity over time. Almost all actions are processes extended in time and are internally structured and can be divided into sub-actions: parts or steps of the whole action. Sub-actions have distinctive rational-teleological relation (not merely an efficient-causal relation) to the whole action, qua parts or steps that compose it (doing A, B, and C, as part of doing D). Consider the motions of someone’s hands while eating. If you ‘un-focus’ your practical-conceptual gaze (so to speak), what shows up is utter disarray of plunging, pulling, rising, etc., motions of a humanoid body at time1, t2, t3… What holds these together as a single, unified and intelligible action?

On the Standard Story, it’s the (‘right’, non-deviant) causal connection to a guiding internal rationalizing cause (a belief-desire pair or intention). But, again, interpreters don’t see internal mental states like belief-desire-pairs or intentions (sans an ideal neuroscience and fMRI). So in virtue of what these motions initially show up as a unity—as a single, intelligible action-in-progress with successive steps—rather than a thousand discrete motions requiring (and making apt) some second-step, Standard Story interpretation to unify them? The same dilemma for the Standard Story looms again.

For Anti-psychologism, our knowledge-how, including our fluency with action-type concepts like ‘eating’, enables us to directly, perceptually grasp this order/structure of reasons present in unfolding actions themselves (not their internal causes) and, thereby, enables us to perceive unities that are not apparent at the level of bodily motions. In normal contexts, like walking in to restaurant, we immediately perceive, not disordered bodily motions to-be-explained, but someone eating: a perfectly intelligible action-type, phenomenologically present as a unified action-in-progress with a series of rationally-connected, constituent parts (grasping the noodles, raising the chop sticks, etc). Given the appropriate knowledge-how and a grasp of the relevant action-concepts, a suitable observer will perceive the multifarious motions in terms of the action-type that unifies and subsumes them within a teleological step-whole structure, prior to ascribing or ‘postulating’ any internal mental items that trigger them. We are attuned to, or home in on, this encompassing action-description—which immediately presents a number of different motions as a single, unified unfolding process of doing such-and-such—out of an infinite number of other true descriptions of an agent’s behavior that segment it in infinitely various ways (and we have a non-inferential, intuitive sense of what counts as the ‘same’ unified action-type, even in novel circumstances).

Attunement to psychologically meaningful features of behavior is, according to Anti-psychologism, a basic starting point, rooted in native human dispositions and socio-cultural formation—not a product of some prior, Standard Story mental-state ascription. For Anscombe (as for Wittgenstein on meaning and rule-following) the fact that human beings do converge, given normal social formation, on the same inter-subjectively intelligible descriptions/patterns, that we are attuned to certain ‘order’ of intelligibility rather than other true descriptions, is a bedrock feature of human sociability—something that grounds further, more elaborate explanations, not something that could be itself grounded by the Standard Story.

Anti-Psychologism: Intentions, Explanations, and Problematic Cases

In the limited space that remains, I’ll briefly elaborate on Anti-psychologism’s unorthodox claim that intentions, reasons, or motives are not internal mental states in terms of which actions can be reductively-causally explained. My aim is merely to provide a basic sketch of what a more fully elaborated Anti-psychologism would involve.

Ryle and Anscombe rightly stress that the basic conceptual shape of psychological predicates like ‘intention’, ‘motive’, ‘reason’ must be gleaned from careful attention to our actual everyday interactions and cognitive practices (e.g., using ‘ Why?’ questions in giving and asking for reasons, following someone’s play in a game); these give us the basic data for philosophical-psychological theorizing. The Standard Story provides a particular philosophical construal of psychological predicates like ‘intentions’ and ‘reasons’—namely, as internal mental states—and then gives an account of how those concepts function in action-explanation and understanding—namely, as rationalizing causes. But this is neither compulsory nor commonsense: rather, it’s a questionable, metaphysically-motivated reconstruction of ordinary practice.

First, intentions/motives/reasons need not be seen as (internal) states that are conceptually prior to and distinct from bodily actions themselves; rather, we can explain intention in terms of action (not vice versa, as on the Standard Story). Following Anscombe and recent work inspired by her, we can conceptualize ‘intentions’ as, not states at all, but doings or forms of activity: namely, being-embarked-on an intentional action (an unfolding process that may only be in its initial phases). In this sense, the agent’s intention does not exist apart from (i.e. as an isolable, prior causal trigger of) the bodily activity she does: instead, it is her doing-of-the-action-intentionally, which holds together the various temporally-extended bodily motions into a single, unified action-in-progress. Hence, we use phrases like ‘I’m doing A because I intend to do B’ interchangeably with ‘I’m doing A because I am doing B.’ In this sense of ‘intention’—as the activity of being-embarked-on-a-type-of-action—intentions are neither necessarily ‘internal’ nor ‘ascribed’; intentions can be outwardly perceptible in embodied action.

While the intentions and psychological significance of other’s behavior is often perceptually evident to us in familiar situations of practice, there can be problematic cases where we fail to understand another person and seek explication via ‘Why?’ questions. And Anti-psychologism can grant that we typically only have a rough-and-ready perceptual grasp of others’ intentions (though this is usually, but not always, sufficient for everyday pragmatic interactions). And it can grant the possibility of error: that we can take ourselves to see certain action-types unfolding that are not, in fact, unfolding at all. Moreover, the meaning of another person’s behavior can be intentionally veiled or deceptive.

Consider problematic cases where we seek further explanation via the practice of explicitly asking for and giving reasons (posing ‘Why?’ questions). Davidson famously argued that the ‘because’ of an explicit reasons-claim like ‘I sent her the letter because I wanted to cheer her up’ must be causal—by which he meant an efficient-causal relation holding between relata that were both particular, datable events: a bodily occurrence and a distinct mental event (that is itself re-describable as an internal physical event). This, he claims, is the only plausible account of how a reason (or motive, intention) really explains an action. But many garden-variety reasons-explanations of actions make no reference at all to internal psychological states of the agent—and, a fortiori, no reference to those states as rationalizing internal causes of distinct bodily motions. It’s perfectly ordinary for an agent, in response to a ‘Why?’ question, to give her reason for acting as she does, thereby explaining her action, by citing things ‘outside the head’ or ‘in the world’, such as:

(a) A re-description of the queried behavior in terms of an action-type :

‘Why are you handing those papers to that girl?’

–‘I’m paying my gas bill.’

(b) A social role-description that functions as an action re-description :

‘What are you doing in my garden?’

–‘I’m the new gardener.’ (read: ‘…and I’m gardening’)

(c) Another action of which the queried action is intelligible as a part or step :

‘Why are you climbing up on the ladder?’

–‘I’m replacing all the old light bulbs in the house.’

Explanations like (a)-(c) are unexceptional moves in our everyday cognitive practice of explaining actions with reasons and they can be genuinely explanatory. Moreover, the world-citing reasons given in (a)-(c), if taken at face value, are not even eligible to be ‘reasons’ and ‘reasons-explanations’ on the Standard Story: the Standard Story has to see explanations like (a)-(c) as non-literal shorthand for, or otherwise reducible to, the ‘real’ explanations that cite internal mental states as rationalizing causes. Anti-psychologism has no such reductive commitment: it can reject the idea that action explanation must always trade in something more basic than action (i.e., internal mental representations and body motions) and that everyday action-understanding needs to be vindicated by some account of causal relations among the lower-level phenomena.

Notice that in everyday discourse, ‘Why are you doing that?’ and ‘What are you doing?’ are basically the same reasons-seeking question. If we take world-citing reasons-explanations (a)-(c) at face value, what holds them together is that they articulate or re-characterize a given bit of behavior so that it is no longer immediately puzzling and we’ve got a workable sense of what the ‘point’ is. A reasons-explanation may fulfill this explanatory function by drawing on explanans that are themselves actions or inherently action-involving, as in (a)-(c) above. Explaining an action in one of these ways is not, primarily, to describe the agent’s internal ‘psychic economy’, but to show how the action fits into a familiar and intelligible world of human practice.

As Anscombe saw, internal mental state explanations of behavior are simply insufficient on their own to secure the minimal inter-subjective intelligibility that is essential to understanding one another’s behavior. I’ll reconstruct Anscombe’s thought as follows:

(1) Reasons-explanations of actions must make the perspective from which the agent acted at least minimally intelligible—i.e. such explanations must show how the action makes sense (or is desirable, rational, fit to choose) given the agent’s point of view, and, hence, must provide a suitable grasp on the agent’s point of view.

(2) Reasons-explanations that meet the condition in (1) involve more than just citing a ‘want’, ‘desire’, or ‘intention’, in the (Standard Story’s) formal sense of mental state with some propositional content and a desiderative direction of ‘fit’ (disposing the agent to change the world to ‘fit’ the content of the attitude).


(3) Citing a (Standard Story) ‘want’, ‘desire’, or ‘intention’, by itself, is not sufficient for a reasons-explanation of an action.

Anscombe’s idea is that, if have an incomprehensible bit of behavior B, citing a formal want/intention (in the Standard Story sense) to perform B, with no further account of the sense in which B appears desirable or even intelligible to the agent, will not provide a suitable reasons-explanation of the agent’s B-ing. Her famous cases involve things like someone scooping mud into saucers or putting green colored books on her roof. Citing a formal, Standard-Story-type intention to do these things will not make the behavior intelligible: it does not ‘rationalize’ the behavior and enable us to understand it as intentional action, as opposed to a compulsion or pathology. Indeed, it doesn’t advance understanding whatsoever.

What is primarily sought by an everyday request for action-explanation is an account of what is being done that provides some sense (at least vague and minimal) of how it could possibly show up as desirable, attractive, meaningful, fit to choose, in the agent’s eyes. This grasp of the agent’s perspective not be particularly rich or deep at all—it can be a dim sense of the what the agent was up to and drawn by—but it cannot be utterly missing, otherwise citing some internal desire/intention to engage in the (incomprehensible) behavior does no explanatory work. The everyday ‘Why?’ characteristically seeks some fuller, broader descriptions of the type of action being performed, or the role/identity being enacted, or the end sought that will show what the agent saw in so acting (this doesn’t require interlocutors to accept the agent’s perspective, of course: only to make minimal sense of it). The Standard Story’s ‘intentions’ and ‘belief-desire-pairs’ are simply insufficient, by themselves, to provide this understanding: hence, reason-giving practices can’t be reductively explained in terms of them.

This shows the crucial importance of what I’ve called ‘mundane’ action intelligibility: our bedrock, pre-theoretical, knowledge-how-based fluency with psychological meanings and intelligible patterns of activity in ordinary circumstances. This is a primary, sense-making ‘frame’ or ‘lens’ that we bring to bear on everyday situations of interaction and within which our everyday reasons-explanations function.

While many patterns of action are clearly pan-cultural, with purposes tied to human nature in a broadly biological sense (characteristically human needs and purposes: ‘eating’, ‘drinking’ and ‘resting’ are obvious examples), social formation has a crucial role as well: many action-types, roles, and situations, are organized by purposes that themselves can only be grasped from within particular socio-cultural practices. Recall Ryle’s examples of psychological understanding as practical knowledge-how: speaking a natural language and following someone’s play in chess. The examples are telling: we can’t bring into view the relevant patterns and purposes except by acquiring, at least to a limited extent, the concepts and perspective—the lens of practical-perceptual knowledge-how—that comes with initiation into that particular practice. The meaning, desirability, and very possibility, of actions-types like making a move with one’s knight in a game of chess depend on a socio-cultural context: to understand and explain what a player is doing in grasping a block of wood and thrusting her hand, we need to know how to ‘read’ the meaning of that motion against the background context of the game. It is familiarity and skillful competence with the wider pattern of the game, the socio-cultural practice, which plays the crucial sense-making role.

For Anti-psychologism, inter-subjective understanding results from the exercise of practical-perceptual knowledge-how, and this knowledge-how is gained through processes of socialization and inculcation into social practices—processes themselves build on and presuppose innate, pre-rational dispositions to react to and perceive other human beings in certain ways (e.g., an infant’s responsiveness to her mother’s gaze, facial expressions, tone of voice). An agent’s repertoire of knowledge-how and mastery of action-concepts constitutes a largely spontaneous, second nature of practical-perceptual skills that enables her to ‘know her way about’ the social world in which she is embedded—a standing capacity to perceive, prior to theoretical inference, meaningful categories of action, enactments of roles, and pragmatic scenarios, and thereby to pre-reflectively understand others in terms of their participation in a familiar socio-cultural world of practice (or, to use an admittedly thorny Wittgensteinianism, a recognizably human form of life).

Conclusion: Reductive Metaphysics and Everyday Cognitive Practice

The positive account of Anti-psychologism above is only a rough sketch of a rival account of agency and inter-subjective understanding to the Standard Story. However, I suspect that what ultimately underlies the attractiveness, for many philosophers, of the Standard Story is not its inherent plausibility (such as it is). It’s that the Standard Story promises to ‘place’ human agency and action-understanding within a metaphysical world-picture to which adherents are antecedently committed: roughly put, one in which the world has a single, fundamental (efficient) causal structure that is ultimately the provenance of the natural sciences to describe and explain, and whose only genuine ontological commitments are to those items that appear (or will appear) in our best natural scientific explanations. Agency and action-understanding, if they are to be vindicated, must be shown to fit into the same kind of (efficient) causal explanatory structure we find in physics and chemistry; any other purported ‘explanation’, ‘knowledge’, or ‘understanding’ would not carve nature at the joints—it wouldn’t be genuinely explanatory or revelatory of reality (in an ontologically austere sense).

Anti-psychologism, in contrast, holds that the practical knowledge and skills at play in our everyday pragmatic interactions with one another can be genuinely cognitive, revelatory of aspects of everyday reality, whether or not this practical knowledge can be fully codified, rendered explicit, and then integrated into reductive and minimalist metaphysics. We need not accept that our practical skills require ‘vindication through reduction’ in the sense that the Standard Story seeks to provide. Such purported ‘vindication’, as I’ve argued, distorts our most basic experience of perceiving, understanding, and making sense of actions in everyday contexts, and threatens to render the manifest image of everyday life incoherent.

We should also resist, as Ryle and Anscombe teach us, the metaphysical temptation to characterize ‘the mental’ as a range of items somehow hidden ‘behind’ the perceptible world of embodied, human practical activity (and to draw sharp, internal-mental-life/outer-bodily-action metaphysical contrasts in the first place). In our most familiar, mundane experience, ‘intentions’, ‘intelligence’, ‘mindedness’, pervade embodied human activity, and, as such, can be perceptually manifest to the knowledgeable observer. This is not ‘behaviorism’ (a charge leveled against both Anscombe and Ryle, mistakenly) but a distinctly Anti-psychologistic account of how human doings are, most basically and immediately, understood and experienced in everyday life: as inherently intelligible and permeated with psychological significance, from the start.