1. Introduction
Elizabeth Anscombe argued that if we are to understand intentional action, we must understand the special kind of knowledge agents have of their own actions. For, she claimed, what agents do only is intentional if they have this special kind of knowledge of it. She thereby introduced the topic of “practical knowledge” into contemporary philosophical discussions of action. Very often, it has been Anscombe’s negative characterization of practical knowledge as “non-observational” that has drawn philosophers’ attention.Footnote 1 According to Anscombe, insofar as agents know what they are doing by way of observation, testimony, inference, or any other mode of what she calls “speculative” knowledge—knowledge in which “the facts, reality, are prior, and dictate what is to be said, if it is knowledge” (Reference Anscombe2000, 57)—then whatever it is they are doing is not an intentional action. However, philosophers have increasingly sought to appreciate the importance of Anscombe’s positive characterization, according to which, unlike speculative knowledge which “is derived from the objects known,” practical knowledge is “the cause of what it understands” (Reference Anscombe2000, 87).Footnote 2 The positive characterization contains the idea—neither contained nor strictly implied by the negative one—that an agent’s practical knowledge, unlike their speculative knowledge, is the cause of its object, i.e., their action.
Anscombe’s claim that we have non-observational knowledge of what we are doing aligns well with our ordinary practical self-conception: we do not appear to have to observe ourselves to find out what we are doing. Furthermore, the idea that such knowledge bears a special causal relation to its object holds out the hope that the immediate non-observational character of practical knowledge—something familiar from cases of self-knowledge of mental attitudes like beliefs, desires, etc.—is compatible with such knowledge itself comprising happenings in material reality. For, as Anscombe puts it: “I do what happens”—or in other words: “when the description of what happens is the very thing which I should say I was doing, then there is no distinction between my doing and the thing’s happening” (Reference Anscombe2000, 52–53). The challenge for an Anscombean account of agential knowledge, therefore, is to explain how the kind of immediate self-knowledge agents have of their own actions extends past mere intention all the way to material reality itself.
However, recent accounts that have tried to take seriously Anscombe’s positive claim that practical knowledge bears a special kind of causal relation to intentional action have done so at the expense of Anscombe’s negative characterization of practical knowledge as “non-observational.” I shall examine Richard Moran and Eric Marcus’s accounts as representative of two trends in the literature: “Formalism” and “Normativism” (sections 2–3). I will argue that each account fails to explain how practical knowledge can comprise what the agent is actually doing, as opposed to what they the intend to be doing. Consequently, in each case, it appears observational knowledge must ultimately be appealed to in order to establish the actuality of an agent’s action. This amounts to a rejection of the very idea of practical knowledge. Picking up on a clue from Anscombe herself, I shall argue that an account of practical knowledge as both the productive cause of the action and, thereby, non-observational knowledge of action, requires that we appreciate the significance of the role of know-how or skill in intentional action.Footnote 3 In particular, I shall argue, first, that an explanation of the productive character of an agent’s practical reasoning must appeal to the agent’s exercise of their know-how or skill, and second, that reflection on the exercise of know-how reveals how agents can have genuinely non-observational knowledge of their action (section 4).Footnote 4
2. Formalism
Anscombe’s positive characterization of practical knowledge as the “cause of what it understands,” a phrase borrowed from Aquinas, comes at the end of an investigation into practical reasoning. The goal of practical reasoning, according to Anscombe, is to calculate the means to one’s ends. If successful, the agent arrives at a conception of an order of actions that, if performed, will realize the agent’s ends. The action’s very structure thus has its source in the agent’s practical reasoning, in their working out how to achieve their end. We can therefore say that an agent’s practical reasoning determines their action insofar as it is practical reasoning that determines what the action performed is. Anscombe’s famous example of a man pumping poison into a house serves as an illustration of the kind of reasoning in question. On the basis of his practical reasoning, the man, interrogated by someone observing him in action, is able to answer a series of questions:
Why are you moving your arm up and down (A)?
B: I’m operating the pump.
Why are you (B)?
C: I’m replenishing the house water supply.
Why are you (C)?
D: I’m poisoning the inhabitants. Footnote 5
The questioner extracts from the pumper a description of an order of means and ends in which each subsequent answer—B, C and finally D—gives an end to which the previous answer has stated the means. The “A–D order” thus arrived at is the order of the pumper’s intentional action itself—its instrumental structure. Crucially, the kinds of answers that practical reasoners are in the position to give—the kinds of answers given by the pumper when he is interrogated about what he is doing—do not merely include statements about what they intend to do, are going to do, or hope to be doing if all goes well; they can also include knowledgeable statements about what the agent is doing.
The crucial question is this: even if agents can arrive at a conception of how to realize their intention through calculating an A–D order, how does this reasoning yield the knowledge that they are A-ing, B-ing, etc., as opposed to merely intending or trying to do so? The answer to this question, it seems, is supposed to have something to do with the fact that the knowledge in question—the knowledge gained through practical reasoning—plays a productive role with respect to action. That, at any rate, is what Anscombe’s positive characterization of practical knowledge would appear to suggest. But what kind of “production” is this, and how might it yield knowledge of action?
In saying that practical knowledge is the “cause of what it understands,” Anscombe does not mean that an agent’s practical knowledge causes their action in the sense of “cause” that characterizes the familiar contemporary notion of efficient causation.Footnote 6 A more promising way of understanding the idea that practical knowledge is “the cause of what it understands” is to conceive of it as the formal—as opposed to (merely) efficient—cause of action. Richard Moran provides a clear formulation of this idea:Footnote 7
the sense in which my practical knowledge is “the cause of what it understands” is not primarily in the sense that my intention to do a particular thing is a necessary causal antecedent for my making the movements I do […] practical knowledge, whose object is specified within an intensional context, determines which descriptions of “what happens” may count as descriptions of what the person is intentionally doing. So, the sense of the phrase from Aquinas is not about the efficient causal role of intention in producing movements, but rather concerns the formal or constitutive role of the description embedded in one’s practical knowledge making it the case that this description counts as a description of the person’s intentional action. If the agent didn’t know this happening under this description, then as so specified it would not be “what he is intentionally doing.” It is in this sense that “practical knowledge is the cause of that which it understands.”
(Reference Moran2004, 47)The role of practical knowledge, for Moran, consists in making it the case that there is any intentional action that is describable (by anyone) as being of a such-and-such a kind. It is because I know that I am, for example, ringing the church bells, that there is any intentional action that is describable in this way by anyone. Calling practical knowledge a “formal cause” on this reading is meant to capture the way my practical knowledge makes it the case that there is such a happening, one describable as my ringing the church bells, in the first place.
In order to see what is at stake here, we can distinguish between a stronger and weaker notion of formal causation. To anticipate: it is clear that what Moran means to describe, and what is indeed needed, is the stronger notion. I will argue, however, that the notion of formal causation as articulated in his official account is not sufficiently distinguished from the weaker notion. According to the “weak” notion of formal causation, to cite X as the formal cause of Y is just to say that X contains the form of Y—or we can also say, contains the account of what Y is. If “X” and “Y” stand for an agent’s practical knowledge and intentional action respectively, then according to the weak notion of formal causation, Anscombe’s positive characterization of practical knowledge amounts to the idea that the form of a given intentional action can be specified by the form that is contained in an agent’s practical knowledge. Now, it is clear this weak notion of formal causation does not capture Anscombe’s idea of practical knowledge, for according to the weak notion, an observer’s knowledge of your action could, if they are sufficiently well informed, be the “formal cause” of your action. After all, an observer’s knowledge of what you are doing can still be said to “contain” the form or account of what it is you are doing. A notion of formal causation must therefore also explain (or be supplemented with an account that does explain) how the agent’s practical knowledge is responsible for their action having the form it possesses in a way that an observer’s knowledge does not.
The point of considering this anemic notion of formal causation is simply to bring out what is in fact required: an account of precisely how the agent’s practical knowledge “makes it the case” that what happens when they act falls “under the description” under which the action is known. Moran, however, does not attempt to secure this dimension of practical knowledge. Moreover, he in fact seems to foreclose the possibility of securing it when he remarks that formal causation has little to do with the genesis of action:
The point is not that the knowledge embedded in my intention helps to produce the movements that lead to the picking up of some milk, but rather that those movements would not count as my picking up some milk unless my practical understanding conceived of them in those terms.
(Reference Moran2004, 47)Moran would presumably grant that we must avoid having a merely weak sense of formal causation, and that we must therefore possess an explanation of how the agent’s knowledge is responsible for making it the case that a given action counts as the intentional action that it does. But it is hard to see what such “responsibility” could amount to in the absence of a properly productive relation holding between the agent’s practical knowledge and their action. If the agent’s practical knowledge really plays no role in the production of the agent’s movements—or, what amounts to the same thing, if the formal or constitutive role of practical knowledge is somehow independent of some additional further role in producing intentional action—it is hard to see how it could be said to be responsible for the action being the action that it is.
This worry about the productivity of practical knowledge directly bears on the question of practical knowledge’s epistemic credentials and in particular its allegedly non-observational character. What practical reasoning affords us, on Moran’s view, is a conception of how to achieve one’s end that is informed by one’s reasons for acting. The crucial question, of course, is how such a conception can constitute knowledge that one is intentionally acting. For Moran, the key to answering this question lies precisely in the “formal or constitutive role” of practical knowledge we have just been discussing. Practical knowledge’s non-observational epistemic relation to its object is supposed to be elucidated through appeal to its responsibility for there being an action that is intentional under certain descriptions. However, precisely because the productive character of practical knowledge with respect to intentional action is left obscure, Moran’s account of its epistemic relation to intentional action is also unconvincing.
In order to see why this is, we first need to raise the specter of a sort of account Moran explicitly wants to avoid—but one that Moran’s Formalism, I want to suggest, is unable ultimately to improve upon. This is the account according to which, as Anscombe herself describes it: “what one knows as intentional action is only the intention, or possibly also the bodily movement [and] the rest is known by observation to be the result” (Reference Anscombe2000, 51). She calls this a “mad account.”Footnote 8 Less hyperbolically, we can call it a “two-factor” account since it depicts knowledge of what one is doing as a combination of knowledge of one’s mental states (one’s intentions) and perceptual knowledge—each making its own separate contribution to knowledge of action.Footnote 9 Such accounts face the challenge of securing the proper connection between the two factors they invoke. For example, suppose I know that I intend to paint the wall yellow; lo and behold, my arm moves and, furthermore, I observe the wall becoming increasingly yellow. Are these two pieces of knowledge, in conjunction, sufficient for me to know that I am painting the wall yellow? For all I know, what I perceptually observe is consistent with some other explanation: a muscle spasm perhaps, or a rare neurological pathology. It seems that an additional theoretical inference must be called upon to justify my self-attribution of the action in question. In which case, we are left with a form of speculative knowledge, in Anscombe’s sense, when it comes to the material happening itself, for the knowledge is, seemingly, “derived from the things known.” What goes missing in two-factor accounts, therefore, is the possibility of immediate knowledge of oneself as efficacious or productive of the movements and alterations one makes. Indeed, that the two-factor conception of practical knowledge appears to alienate the agent from their action underlines the original attraction of Anscombe’s notion of practical knowledge. Practical knowledge introduces the possibility of practical self-knowledge, i.e., immediate self-knowledge of oneself as acting, that does not reduce knowledge of one’s action, implausibly, to consciousness of, on the one hand, one’s mental interior, and on the other hand, a merely theoretical knowledge of material reality. It does so, moreover, by raising the prospect of a form of knowledge that is itself productive.
Moran himself recognizes that a two-factor account should be avoided, and states that it is Anscombe’s view of practical knowledge as a formal cause that indeed enables her to avoid it. Furthermore, he claims that, in the same stroke, this account shows how practical knowledge “depends on, but does not reduce to, speculative knowledge of can happen and is happening and thus evades the charge … that the admitted dependence of successful action on ordinary observation must mean that the agent’s knowledge of what he is doing must be observational after all” (Reference Moran2004, 56). But can Moran’s account meet the challenge he sets for it here? Let us grant that the practical reasoning an agent engages in—reasoning through which they arrive at an A–D order—results at least in their forming the intention to realize that order and thereby results also in knowledge of their intention to act. How does that reasoning also provide knowledge that there is a happening that is genuinely describable as the execution of that very intention? It does not, for the account leaves out practical knowledge’s efficacy. Where there is no account of the agent’s efficacy, there is, unsurprisingly, no account of the agent’s knowledge of that efficacy.
It would be inaccurate to describe Moran’s view as being a two-factor view, for he evidently recognizes the need to avoid ascribing to Anscombe or otherwise endorsing a view of agent’s knowledge of action which is somehow built up out of knowledge of intention and perceptual knowledge. On his account, practical knowledge commits itself to the actuality of the agent’s action without needing to advert to an amalgam of these two factors. But although Moran’s account is different in its ambition, it nevertheless misses the same key element that two-factor views do: a form of immediate knowledge of oneself as acting. The two-factor account rules out the presence of such a form of knowledge, whereas Moran’s account posits it, but fails to account for it.
Because Moran’s account fails in this regard, the very danger that Moran himself recognizes—namely, that ordinary observation, which is after all drawn upon in action, will appear to be the true source of knowledge of one’s action—is very much a live one. Moran, of course, wants to resist this suggestion: although practical knowledge is dependent on observational knowledge both as a source of background knowledge about the world and as an aid within particular intentional actions, he denies that practical knowledge is “derived” or “inferred” in any way from perception. In making this point, he draws an analogy between practical knowledge and mathematical knowledge:
[The agent’s] dependence does not by itself mean that the knowledge in question must really be observational after all, any more than the dependence of one’s mathematical knowledge on the good working order of a calculator or a teacher or one’s own brain means that such knowledge is really empirical and not a priori.
(Reference Moran2004, 61)Precisely how we are to read the analogy is not obvious, but in a footnote at the end of this passage Moran approvingly cites Tyler Burge, who defends the claim that knowledge does not lose its a priori status in virtue of being dependent on perception.Footnote 10 Crucial to Burge’s argument, however, is the idea that the role of perception in such cases is merely to transmit or preserve knowledge claims whose ultimate justification nevertheless remains nonempirical. In Moran’s account, by contrast, there is no original explanation of how practical knowledge could ever be a source of knowledge of what is happening in the material world.Footnote 11 Such an explanation cannot be given within Moran’s framework precisely because the conception of practical knowledge Moran articulates is restricted to the formal level, and the productive aspect of practical knowledge is left unilluminated. Until that is provided for, the danger remains that the agent’s knowledge of themselves as efficacious—as acting—will always be left to perception, and in such a way that it is vulnerable to the objections as the two-factor view.Footnote 12
3. Normativism
If an agent’s practical knowledge is not merely the formal cause of intentional action, but is actually productive of it, then perhaps this can be explained by reference to an Aristotelian doctrine that Anscombe subscribes to: that intentional actions are the conclusions of practical syllogisms. Eric Marcus has recently defended this idea, arguing that we should identify action with a normative judgment that constitutes the conclusion of a practical syllogism. Following him, I will refer to this kind of account as “Normativism.”Footnote 13
The Normativist holds that the conclusion of practical reasoning is both a normative judgment to the effect that one should perform a certain action and, simultaneously, the very action that judgment concerns. Through the identification of action with a normative judgment, the account promises to explain both the efficacy of an agent’s thought—its resulting in intentional action—as well as the agent’s knowledge of the action. The special kind of efficacy associated with practical thought is understood in terms of the rational capacity to make practical inferences. This framework, Marcus suggests, can illuminate our understanding of the non-observational character of practical knowledge, for: “If actions are judgments, then whatever explains an agent’s knowledge of the relevant judgment can explain her knowledge of the action” (Reference Marcus2018, 320–21). The identity of judgment and action suggests that whatever explains knowledge of the judgment also secures knowledge of action.
According to Marcus, practical inferences are essentially self-conscious: agents can avow these attitudes authoritatively without evidence or observation. Marcus starts from the observation that, from the first-person perspective, the question whether one believes that p is transparent to the question as to whether p is true.Footnote 14 In other words, the question whether I believe p is, for me, settled in settling the question whether p. Marcus explains this observation as follows: believing p is simply representing p as “to-be-believed” (in the case of inference, on the basis that another proposition, q, is to be believed). In representing a proposition in this way, one is thereby conscious of one’s holding the belief in question. The identification of intentional action with judgment makes available the correlative treatment of practical knowledge. If I judge that “A-ing is to be done,” then my knowledge of my so judging will be transparent to the question of whether A-ing is to be done. If that judgment is my A-ing, then it follows that in making the judgment “A-ing is to be done” I know I am A-ing. Practical knowledge—non-observational knowledge of my own action—is thus understood as a species of a broader genus of non-observational knowledge: the distinctive first-personal knowledge of one’s own rational attitudes.
However, even if I know that I ought to A—and just know, thereby, that I judge as much—it does not obviously follow that the self-consciousness associated with that judgment also comprises my action actually taking place. Concluding that one ought to take a means to an end is consistent with failing miserably, or with not even having begun to act. Knowledge that I am actually doing what I should be doing still requires knowledge of my action as taking place in material reality. The agent’s recognition of the normative status of an action type does not speak to the question of whether it is taking place in material reality. The self-consciousness Marcus describes appears to extend, therefore, only to the normative character of a proposed action, rather than its actual existence. As such, it does not constitute practical knowledge.
There are different reasons one might be skeptical of Normativism’s identification of judgment and action; crucially, I will not here express a general skepticism concerning the claim that a normative judgment could be identified with an intentional action. More specifically, I do not mean to motivate a skeptical attitude towards Normativism by appealing to cases of what Marcus calls “disengaged deliberation”—cases in which agents think they ought to do something, but due to some failure of will nevertheless do not.Footnote 15 For current purposes, I am happy to grant such cases can be explained satisfactorily. My concern is this: even in Marcus’s paradigm cases—the cases in which agents decide they ought to do A and thereby A, their conclusion being identical to their action—it is left mysterious how the content of the normative judgment could be such that, simply in virtue of the agent’s consciousness of their normative judgment, they are conscious of the material happening that is their action.Footnote 16
Now, in fact, Marcus appears to concede that, without observation, the agent can only know what they are “up to” and not what they are “actually doing”:
To know that I’m actually doing it [i.e., actually performing an action], I do rely on perceptual knowledge. But “actually” signals that what’s at issue is a judgment that my plan is working or perhaps that my chances of success exceed a certain threshold. It is the sort of judgment that an observer of my action might make; it is a theoretical judgment. When the observer makes such a judgment, she does so against the background understanding that that is what I am up to.
(Reference Marcus2018, 327)The idea that knowing what I am actually doing is the result of a theoretical judgment, expressive of perceptual knowledge, marks a retreat from the idea of practical knowledge. Practical knowledge understood as knowledge of “what I’m up to” is ex hypothesi consistent with ignorance that my intentional action is happening in material reality. If that is the case, then it is no more than knowledge of what I am doing if my intention is being realized. The agent’s reliance on perceptual knowledge to know more than that raises the concern that Marcus’s Normativism (like Moran’s Formalism) is at risk of failing to significantly improve on the two-factor view. Although Marcus sets out to respect the kind of self-consciousness characteristic of intentional action, it is precisely the crucial element—the agent’s own efficacy—that goes missing from the agent’s self-consciousness.Footnote 17 Knowledge of that requires more than the self-consciousness characteristic of a normative judgment that takes an act type as its object.
Of course, Normativism, like Formalism, is not a two-factor account. Marcus thinks that his concession is compatible with the claim that practical knowledge, as he conceives of it, does by itself comprise intentional action. For he thinks it is possible to know “what I am up to” (where this is more than knowledge of “mere” intention, but is knowledge of action) without nevertheless having knowledge “that I’m actually doing it.” Why think that, though? The idea might partly stem from a view of intentions as ontologically continuous with intentional actions. On Marcus’s conception of intention, “mere intentions are such representings [i.e., the normative judgments identified with actions] in their earliest (according to flexible standard) phases, or highly defective representings” (Reference Marcus2012, 90). Consequently, even in cases where, for example, I intend to bake a cake but haven’t yet started to do anything—or cases in which I am failing in my attempt to bake a cake—I am still “up to” something, and can know “what I am up to” where what I thus know is an action in either an incipient, or else highly privative, form. I do not wish to object to the inclusion of such phenomena under the general category of “intentional action.” However, an account of the proper object of practical knowledge should at least not treat these cases as paradigmatic, for then practical knowledge would, in general, be thought only to comprise whatever the agent does know when, in fact, they have yet to lift a finger, or are flailing about hopelessly. To treat such cases as paradigmatic risks unwittingly restricting practical knowledge to knowledge of intention. For knowledge of “what I am up to,” where that excludes what I am actually doing, seems only to comprise that much. I shall return to this point below (section 4.b).
What I have said thus far has not been aimed at casting doubt on Moran’s claim that practical knowledge is the formal cause of what it understands, nor Marcus’s claim that actions are identical to thoughts concerning the normative status of those very actions. What I have meant to cast doubt on is the capability of these doctrines, as articulated by their defenders, to illuminate the non-observational character of practical knowledge. I shall now turn to consider how we might stand to do better.
4. Practical knowledge and know-how
4.a A clue from Anscombe
In order to make headway, I propose we consider an oft-overlooked passage from Intention:
“Practical knowledge” is of course a common term of ordinary language … A man has practical knowledge who knows how to do things; but that is an insufficient description, for he might be said to know how to do things if he could give a lecture on it, though he was helpless when confronted with the task of doing them. When we ordinarily speak of practical knowledge we have in mind a certain sort of general capacity in a particular field; but if we hear of a capacity, it is reasonable to ask what constitutes an exercise of it … In the case of practical knowledge the exercise of the capacity is nothing but the doing or supervising of the operations of which a man has practical knowledge … Although the term “practical knowledge” is most often used in connexion with specialised skills, there is no reason to think that this notion has application only in such contexts. “Intentional action” always presupposes what might be called “knowing one’s way about” the matters described in the description under which an action can be called intentional, and this knowledge is exercised in the action and is practical knowledge.
(Reference Anscombe2000, 88–89)Anscombe notes here that “practical knowledge,” the term she has been using all along to denote an agent’s knowledge that they are performing a particular action, already has a meaning in our ordinary language. In ordinary usage, to have practical knowledge is to know how to do something, and not only to the extent that one can speak about performing actions but insofar as one is able to do the things in question oneself: ride a bike, construct a wall, fix a vending machine and so on. Although two people, one capable of enacting a particular series of means-end actions, one only capable of lecturing on how such a thing might be done, could both be described as in some sense knowing how to do something, it is only the one who can carry it out that has genuinely practical knowledge in this ordinary sense. What Anscombe goes on to say suggests that there is, however, more than a verbal similarity between her notion of practical knowledge and the ordinary one. The ordinary sense of practical knowledge, “‘knowing one’s way about’ the matters,” is presupposed in every case of intentional action, and thus in every case of practical knowledge in her technical sense.
To see why this might be the case, consider the difference between Anscombe’s poisoner and a historian intimate with, what has later become recognized as, a turning point in world history. The latter might know in minute detail the A–D order enacted, perhaps from the account given by the perpetrator himself. Nevertheless, when charged with performing the same act for a historical recreation on her university’s “Humanities Day,” she may be incapable of carrying off the deed correctly. Perhaps she is not used to operating this exact kind of pump and is therefore failing to apply the right kind of simultaneous pressure on its different parts; or perhaps she cannot fix the leaks that inevitably occur in this kind of setup quickly enough to make headway. Past a certain point, further means-end information will be useless. The historian, we would like to say, simply lacks the know-how or skill necessarily to pull it off.
It is in exercising their know-how or skill—their capacities to act—that agents are genuinely efficacious. For this reason, the possession of know-how distinguishes genuinely practical knowledge of A–D orders from those of someone who, like our historian, can merely lecture on them. I shall argue that it is precisely in its “guise” as the know-how exercised in action that practical knowledge is the source of the agent’s non-observational knowledge that they are actually acting, as opposed to merely intending to act. Consequently, recognizing the role of practical knowledge in the “ordinary” sense is key to understanding practical knowledge in Anscombe’s sense.Footnote 18
4.b The “actuality” of action
To appreciate the role of know-how in practical knowledge, we need to consider more closely what distinguishes genuine knowledge of action—that is, knowledge that one’s action is actually happening—from mere knowledge of intention. This requires a brief foray into the topic of the temporality of action. In recent years, philosophers of action have been paying renewed attention to the temporal structure of intentional action—in particular, to that temporal feature of the structure of action that is reflected in English grammar by the contrast between perfective and imperfective aspect. Footnote 19 Aspect, unlike tense, marks distinctions in the internal temporality of processes or events. The importance of aspect can be brought out from a simple observation: the fact that someone was doing something, does not entail that they did it. For example, if I say that I was crossing the road—a statement in the past tense, but which displays imperfective aspect—what I say does not entail that I ever crossed the road, a claim exhibiting perfective aspect. Indeed, the truth of sentences in the present progressive tense that display imperfective aspect, i.e., sentences of the form “S is A-ing,” rule out that the action is completed at the time of the assertion—that is, they rule out the corresponding perfective claim “S has A-ed.”
These grammatical distinctions articulate the temporality of intentional actions, which are unfolding processes that admit of interruption (the kind of thing Aristotle would class as a kinêsis). The fact that the conclusion or “perfection” of such processes rules out that they are still ongoing is clear enough. To be actually crossing the street one must not have already crossed it (unless one is doing so a second time). However, there is a related question to which the answer is less obvious: To be actually crossing the street, need one have made any progress to the other side? While the fact that an agent is presently doing something means they have not yet done it, it does often entail that they have at least done something. Footnote 20 I need not have actually gotten to the other side of the street in order to be able to truly declare myself to be crossing the street, but unless I am making some progress, my claim that I am crossing the street should, at best, be taken with a pinch of salt. Accordingly, Matthias Haase argues that when it comes to intentional movement, knowing that I am A-ing essentially involves knowing that I have A-ed. This is, as Haase points out, simply a special instance of a general truth about movement:
Consider watching a tree fall over. The experience of this motion in progress always involves the apprehension of the bits of space the tree has already traversed in its fall. When you haven’t experienced any bit of falling through a segment of space, then you are not watching it fall. Perhaps you saw that it got uprooted. Consequently, you may be said to know that it is going to fall eventually. But if it hasn’t moved yet, then it is not falling. So all you know is its present tendency to fall. A tendency that might be prevented from manifesting itself in actual motion. Perhaps the next moment something gets jammed into the tree so that it doesn’t end up falling.
(Reference Haase2018, 229)The general point about movement is this: knowledge of the movement of some substance, S, always requires knowledge that S has completed some phase of the movement that is in question. Consequently, if we “subtract” from the knowledge that “S is falling” the knowledge that “S has fallen this* much”—where “this*” refers to whatever phase of falling the perceiver is comprised of—we are left, it seems, merely with knowledge of the tendency of something to fall, not knowledge of the falling motion itself.
When we turn to intentional movement, the point shows up in the inferences that can be made from statements with prospective aspect used to express future intention (“I am going to A”) and statements in the present progressive (“I am A-ing”) used to characterize actual movements. From the statement “I am going to A,” the inference to the fact that there is some phase of A-ing which I have already completed is in no way obligatory. Nothing in the fact that I am going to cross the road entails I have made any progress in doing so. However, from the claim that “I am A-ing”—used to speak of an actual crossing in progress—the inference that there is some phase of A-ing which I have already completed is obligatory.
Haase draws a crucial consequence of this observation: were the content of one’s practical knowledge strictly limited to the present progressive—to “I am A-ing”—and did not include “I have A-ed” (or “I have A*-ed,” where A* is some phase of A-ing) one’s practical knowledge would be confined merely to knowledge of the tendency to act, rather than the actuality of acting—and, as Haase notes: “one already knows that when one knows what one intends or wills to do” (Reference Haase2018, 229). My practical knowledge that I am A-ing, therefore, if it is to be more than knowledge of intention and amount to knowledge of the actuality of my acting, must contain within it at least some knowledge of my actual progress, of what I have actually done (thus far) in A-ing.
This point about the content of practical knowledge is intimately related to its productive character. The knowledge of someone who merely intends to act comprises only an order of act types: general ways of doing things that could be instantiated, on different occasions, indefinitely many times. However, once we recognize that practical knowledge comprises what the agent has done thus far, we can see that agents who have practical knowledge of what they are doing must have knowledge of an A–D order of actions as actually, although perhaps only partially, instantiated in material reality.
Now, it is through the exercise of know-how that the form that figures, at the outset of action, merely in the agent’s intention, progressively becomes instantiated in material reality. It is therefore an appreciation of know-how’s contribution to practical knowledge that was missing from Moran’s Formalist account. Recall that Moran offered no explanation of how the form of action contained in an agent’s intention—i.e., the A–D order to be carried out—is the same as that found in the agent’s action. In the absence of such an explanation, the scope of practical knowledge could only extend to the intention to do B by doing A, etc. It is the productivity of practical knowledge—insofar as it involves the exercise of know-how—that marks the shift in its content; that is, the shift from comprising something merely formal, to comprising a form that is instantiated in action. Absent the exercise of know-how, an agent’s practical knowledge can only move from knowledge of a general order of act types, to knowledge of those act types as instantiated, through the offices of perception.Footnote 21
The question that remains is this: Can the exercise of know-how not only secure the “actuality” of action—thus making available the proper object of practical knowledge—but also yield non-observational knowledge of that object through its very exercise? If an agent’s exercise of know-how in carrying out an A–D order does not only bring about genuine alterations in material reality, but is also the source of knowledge of those alterations, it must be the case that it is through the making of those very alterations that they are known.Footnote 22 But how is it that in bringing about some movement or change one can come to have knowledge of that movement if not by observation?
4.c Know-how and non-observational knowledge
The province of material reality that an agent has the most immediate knowledge of in the exercise of know-how is their own body, for the agent’s body is the substance whose intentional movements and alterations are known through their very movement of it. I do not need to observe the movements of my limbs as I swim, walk, or run in order to know what is happening with them. I know of their progressive movement through space simply in virtue of moving them.Footnote 23 Why is this? In bodily movement, there is an identity between the bearer of the capacity to move, and the substance in which the movement occurs. Consequently, there is no distinction between agent and patient, and thus no need to posit an act of perceptual or receptive cognition by means of which an epistemic gap between agent and the patient is closed.Footnote 24
Of course, the claim that there is no room for a perceptual relation to the object of practical knowledge in the case of one’s own body is only the first step in a full account of the nature of bodily awareness within action.Footnote 25 That account is beyond the scope of this paper. Nevertheless, at its core it would have to be the following idea: the special kind of awareness that attends the exercise of bodily know-how is grounded in the fact that the agent’s ability to move their own limbs is largely constitutive of what their limbs are. That is to say, an account of what our limbs are “in themselves,” one that takes into view the role they play in the life of a self-conscious living being, is inseparable from an account of our capacities to move them. This distinguishes our limbs from everything else we can move or alter, which need not be understood as something the essence of which lies in its being so moved. If that is correct, the distinction between one’s capacity to move, and the object of that capacity, collapses in the case of the body. Our arms, legs, hands and feet, should all be understood as capacities for movement “made flesh,” as it were. What is important for our purposes is that it is precisely because in the case of bodily movement an agent-patient gap, and thus the need for a mediating relation, is absent, that the knowledge one has of the movement of one’s body is not receptive, and therefore not perceptual. Insofar as the knowledge one has of the movement of one’s limbs is essentially grounded in one’s ability to move them, such knowledge is distinctly practical. This is why insofar as exercising know-how involves moving one’s own body, knowledge of the actuality of that bodily movement is non-observational.
However, knowledge of bodily movement, even when fully accounted for, is not sufficient to explain the knowledge of actions involving external substances. Recall Anscombe’s slogan: “I do what happens.” “What happens” when we act extends past the physical boundaries of our bodies. The range of things I am ordinarily said to know how to move or alter is not exhausted by my own limbs, but comprises the movement and alteration wrought in all those material realities I utilize and transact with that exist external to my body.Footnote 26
This is the point where it can seem as though we must give up on practical knowledge. If knowledge of one’s action comprises knowledge of the actual obtaining of changes in things external to one’s body, how could it be anything but observational? It seems agents must stand in some kind of receptive relation to the changes wrought by them in external materials if they are to have knowledge of them. Consider an example: while sautéing some onions, I see that they are browning. This knowledge is apparently had through observation. Specifically, given the premise that the knowledge that I am sautéing onions necessarily includes the knowledge of my having sautéed them (at least to some degree)—and given that that must include knowledge of their altering appropriately in response to the heat, namely, knowledge of their browning—it seems to follow that my knowledge of my action is, after all, observational.
As a first response to this problem, we could try saying the following: it is misleading to say that in simply observing the change in the color of the onions I have observed my progress in acting. What I observe is not my acting but the particular objects involved in my action. Indeed, what I observe is in fact consistent with someone else having gotten the onions into that state. If what I know, in perceiving the state of the onions, is my progress in sautéing them, it is only because my perception of the onions presupposes knowledge of my having gotten them into that state in the first place.Footnote 27 Thus although perception is involved, in order for it to provide me with a basis for judgment about my actual progress, it must be interpreted in the light of that genuinely practical knowledge. Therefore, this reply goes, my knowledge is never simply observational, for observation can ground my knowledge only in light of presupposed practical knowledge.
This response is a step in the right direction, but notice that it stands in danger of simply reinstating a version of the “two-factor” view. The difficulty is this: although on this account only an agent with practical knowledge could, by means of her observations, have knowledge of the actuality of her action, the observations she makes are of a kind that could be had by an observer of the action. This suggests that ordinary perceptual observation provides a distinct contribution to knowledge of the actuality of action. As a result, we end up chasing the agent’s genuinely practical knowledge back to the movements of her body, leaving the rest to observation.Footnote 28 Although the perception involved in the exercise of know-how contributes to knowledge of the progress of action when interpreted by the agent in the light of what they already know themselves to be doing, the latter extends no further than the contours of the agent’s body.
Perhaps the kind of perception agents have of the tools and objects they handle could be enjoyed by mere observers. However, cases in which observation makes a unique and separable contribution are, I suggest, at best a secondary phenomenon. The primary way perception figures in the exercise of know-how marks the contribution of perception as inseparable from the exercise of skill itself, and, I shall argue, in a way that can be called “non-observational” in Anscombe’s precise sense, i.e., not derived from the thing known.
Consider another example: I am tying my shoe laces when I realize, halfway through, that they are not tight enough, and so start over again.Footnote 29 Coming to that conclusion requires being perceptually related to my laces. However, the way in which I am perceptually related to my laces is not one in which a mere observer could be related to them. This comes out in the fact that in my perception of the shoe laces as I tighten them, there is little separation between what I perceive and the act of tightening them. If I were pushed as to how I know they are tight, I might describe the ease with which the laces did or did not yield to my various movements. But, to make sense, those descriptions have to make reference to the movements appropriate to lace tying: movements that are the exercise of the skill of tying shoe laces. The content of my perception that the laces are (or are not) tight cannot be separated, in this case, from the very act of tightening them.
The example suggests the existence of perceptual capacities where neither the exercise of the capacity, nor the perceptual contents those exercises afford the agent, can be understood as independent of the exercise of know-how. The way in which this kind of perception is partly constitutive of the exercise of know-how distinguishes it, I want to suggest, from the kind of observation that could, in principle, be had by an observer.Footnote 30
Recall Anscombe’s characterization of observational knowledge, which is “speculative,” as “derived from the things known,” as opposed to practical knowledge, which is “the cause of what it understands.” Now, there is clearly a sense in which my knowledge of the shoe laces is derived from the things known in the example above. After all, the laces themselves are not the product of my action. However, the same cannot be said for the change that I make in the laces by tying them. My knowledge of that, of the alteration I make, only counts as “derived from the things known” in a highly qualified sense, for the content of my perception is in that case intrinsically determined by the productive exercise of skill: what I perceive in tying them is inseparable from my act of tying them. It is only insofar as I am the cause of the reality known that I have knowledge of it, and although that knowledge is partly perceptual, it is not “derived from the things known” in a sense that excludes its being knowledge that is “cause of what it understands.” The idea that perceptual knowledge of what one is doing could never be practical knowledge, turns on the assumption that an exercise of perceptual knowledge could never be productive of its object. The possibility of the kind of perception I have drawn attention to undermines this idea.Footnote 31
The fundamental mark of practical perceptual knowledge, therefore, is its content’s resistance to being factored into two independent elements: (a) what the agent perceives, such that anyone suitably positioned could perceive it, and (b) a separate interpretation of the significance of what is perceived from the perspective of someone with know-how. It stands to reason that the sense of touch provides the clearest prima facie example of such unfactorable perception, for insofar as touch registers the presence of external surfaces on our body, it is the sense most obviously bound up with our capacities for movement. But there is no reason to think that the kind of perception in question is limited to touch. Consider, for example, an archer aiming at a moving target—they know, if properly experienced, when to loose their arrow in order to have the greatest chance of success. Part of knowing that they are actually doing that involves knowing when they have the target properly in view. But what they perceive, in that case—that they have gotten themselves in a position to shoot—is not something they arrive at merely through an interpretation of what just anyone, suitably positioned, could perceive.
We can now state the sense in which the exercise of ordinary observational capacities is secondary to the core case of the involvement of perception in action. Sometimes agents do evaluate objects in material reality in order to be sure they are making progress and in a way that will involve perceptual knowledge that could be had by anyone, not merely someone who knows how to do the thing in question. After changing the light bulb, we flick the switch a couple of times to make sure all is in order; after sanding wood, we run our hands over the surface to make sure it is smooth. Such knowledge is secondary insofar as perceptual knowledge had by the agent in those cases is only knowledge of my action against the presupposed background of the genuinely non-observational knowledge of my action—knowledge that involves practical perceptual knowledge—the source of which is the exercise of know-how.
Much more clearly needs to be said in order to account for the phenomenon of practical perceptual knowledge as I have just outlined it. All I have done here is try to motivate the necessity of such an account, as well as make some room for it against a backdrop of assumptions which seem to foreclose its very possibility. A full account would have to do no less than vindicate the idea of a form of receptivity the affordances of which are not capable of proper characterization independently of an attendant sense of know-how or skill.
5. Practical knowledge, practical reasoning and know-how
The preceding section proposed a framework for understanding how the exercise of know-how can be understood as both the cause of action and the source of non-observational knowledge of action. In the process, I hope to have shown something of the unity of Anscombe’s negative and positive characterizations of practical knowledge. It is insofar as practical knowledge—in the form of the exercise of know-how—constitutes the efficacy of the agent, that it is a source of non-observational knowledge of what the agent is doing. I want to end by saying something about how we should think of the place of know-how among the other capacities that are involved in intentional action.
I have claimed that the exercise of “ordinary” practical knowledge (i.e., know-how or skill) is the source of practical knowledge in Anscombe’s sense. However, know-how cannot be the only source of practical knowledge. For the practical reasoning described in Intention that arrives at an A–D order is not identical to the exercise of know-how, but is nevertheless a prerequisite for any knowledge of one’s action that stems from the exercise of skill. How precisely should we think of the relation between practical knowledge in Anscombe’s technical sense, and “ordinary” practical knowledge or know-how?
The capacity to arrive at an A–D order of actions—i.e., the capacity for practical reasoning that Anscombe illuminates through her discussion of the practical syllogism in Intention— presupposes the ability to perform such actions, and thereby presupposes skill or know-how. For the practical reasoning that arrives at an A–D order of actions would be merely idle, if not fantastical, if it did not always have in view the agent’s know-how—what they can and cannot do. Viewed from the other direction, however, the exercise of know-how always presupposes some A–D order of actions the agent is performing. Together these two capacities are exercised in the performance of a particular A–D order of action.
Nevertheless, the capacity to perform a particular A–D order is not simply an actualization of one or several aggregated skills the agent possesses. For one’s general know-how always interacts with one’s conception—had through one’s instrumental reasoning—of the particular A–D order the agent is to perform. One exercises a skill—in the good case—in a way that is appropriate to the particular means-end structure one enacts. For example, one pumps ever so quietly on a stealthy mission, or perhaps slowly if the next step in the plan is arduous enough that one shouldn’t tire oneself too quickly. You might tie your laces loosely if you’re only popping out for a moment and need to pull your shoes off easily when you get back, but tightly if you are about to scramble down a ravine. When an agent’s practical reasoning and skills are working in lockstep, we can ascribe to them the capacity to perform a particular A–D order.
We can therefore agree with Anscombe that both things she describes in the passage quoted above as “practical knowledge” are indeed practical knowledge, for together they make up, with practical reasoning, the general capacity for intentional action.Footnote 32 Practical knowledge, so construed, is a capacity with two core actualizations: intentional actions and knowledge of one’s intentional actions. Crucially, a relation of priority holds among these actualizations: the agent’s knowledge of their action has its source in their acting, for, as we have seen, it is insofar as one exercises know-how in carrying out an A–D order one has arrived at, that one comes to have knowledge that one is actually acting and not merely intending to act.Footnote 33
Acknowledgments
An audience at the University of Chicago made helpful comments on an early draft of this version of paper; special thanks are due to Christian Kietzmann for further helpful discussion just after that occasion. Two anonymous reviewers for this journal also gave me helpful comments and pushed me on several points in a way that improved the paper. Finally, thanks to Matthias Haase for discussion of this paper, as well as many other illuminating conversations on Intention more generally.
Rory O’Connell is a teaching fellow at the University of Chicago. He works primarily in philosophy of action and ethics.
