Philosophy of action inspired by the work of Elizabeth Anscombe has been connected to ethics and metaphysics; and sometimes to psychology. Fewer attempts have been made to connect it to political philosophy, the philosophy of history, or economic theory. This may be due to how we usually talk about historical, economic or political occurrences. Take revolutions. No one would dispute that revolutions have economic, political and historical implications. However, people often talk about revolutions as they talk about the weather or about seismic occurrences. Ten days that shook the world, John Reed remarked apropos the Russian Revolution. This way of speaking raises questions. If revolutions were natural events they could in principle take place in the absence of any agent and would require no agents; and thus would be no business for philosophy of action, however conceived.
Most of us would nevertheless agree that the Russian Revolution was the outcome of many actions, caused by many agents acting in all sorts of ways. Still, some difficulties arise. Were we to ask those agents about the reasons for their actions, only a small fraction would admit to having intended to bring about what we now call “the Russian Revolution.” Also most likely the description the Russian Revolution would be understood differently by different agents.
Perhaps revolutions are instead like Berkeley’s trees in the garden, that is, perhaps they only come about insofar as “there is somebody by to perceive them.” It seems indeed to matter that we only name such things retrospectively, that what we call the Russian Revolution corresponds to aggregate descriptions of many past actions. This is easily reconcilable with assuming the lack of awareness of all, most, many, or some of the putative agents. You can well act relative to an end that someone else (possibly even yourself) will later describe as something that hadn’t occurred to you while you were acting.
The consensus among historians is that the Russian Revolution was the outcome of many different intentional actions; and of many unintended consequences of many of these intentional actions; and that the many agents had different reasons for acting in the many ways they did; and that some of these agents did not know what they were doing; or were wrong about what they thought they were doing. It is possible that no majority of agents ever intended to bring about anything like what we now Berkeley-like call the Russian Revolution; and even if they had it would not have made any difference. In matters of action even Bolsheviks appear to be Mensheviks. Majority-intentions and wishes add nothing special to causation, though the intentions of particular agents often might.
This admittedly is not saying much. Ben Laurence has made the stronger claim that “people are acting together intentionally if and only if their actions can all be straightforwardly instrumentally rationalized by the same action.” (282) His example is that of a gang robbing a bank. Each gang member, he argues, can instrumentally rationalize their bank-robbing tasks. However, not all cases of acting together appear to be like robbing a bank. For instance, what Laurence calls the instrumental rationalization of the action is in the case of the Russian Revolution mainly ex post facto, whereas in the case of robbing a bank it is frequently ante factum. This difference does not seem to be one of scale. It may not always be the case that very large gangs can only rationalize their actions after the fact; and it is conceivable that the joint cooperation of very few people (and indeed actions by single agents requiring no cooperation at all) may eschew instrumental rationalization altogether.
Later in his paper Laurence puts forth an additional distinction that implies an even stronger claim. He states that “an individual is acting as part of some group if and only if his action is subject to the special collective sense of the question ‘Why?’” (289) This rests on the notion collective sense of a question. The notion feels prima facie mysterious. Surely any “collective sense” would at least imply more than one agent. You can mostly tell whether in the question Why are you doing X? “you” is singular or plural. For robbing a bank it can be either. For making a revolution it will likely be plural (singular-you revolutions we would not call revolutions). This additional distinction may be of some help in separating cases like robbing a bank (all kinds of “you”) from both cases like flying a kite (mostly singular “you”) and cases like revolutions (mostly plural “you”). It does not however clarify any “special collective sense of the question ‘Why?’,” since it does not explain why or when any two people would have to rationalize their tasks identically.
It could be that the collective sense of an action is then more an effect of retrospection than a property of a certain class of actions. The argument sounds congenial. However it entails a Pyrrhic price tag. If no “special collective sense” of the question “Why?” applies, then, as Laurence says, either the action (collective action, one would assume) “is straightforwardly unintentional” (287), or the agent “does not know that he is acting as part of a group [or again, does not know what it is that the group is after].” (idem) We would be facing an unenviable choice: either the Russian Revolution was a mere coincidence, or it was not a case of collective action. The first alternative offends the widespread belief that not everything was random in the Russian Revolution. The second one offends the widespread belief that the Russian Revolution was the outcome of many individual actions. If that was not a case of collective action surely nothing could be.
A difficulty with explaining the “collective sense” of the question “Why?” is that there is no such thing as a common mind for people synchronously to entertain wishes and intentions. Laurence acknowledges that there is “no one mind shared by a collective agent” (271). What we sometimes call a common mind is really more like a shared task or like a series of drills, as in synchronized swimming. Most explanations of “collective actions” seem to require something like division of labour, at the very least one between planners and operatives, though also often one between kinds of operatives. However, explanations of how such divisions come about tend to be labile. Adam Smith, who coined the phrase “division of labour” and who occasionally wrote about minds, believed that only creatures such as dogs, i.e. creatures that for him do not have minds of their own, could be of one mind. Thus, with allowances to what he says elsewhere about sympathy, no large number of agents ever can. He also believed that what looks like division of labour is often not the outcome of any agreement as to who should do what. Individuals, as Laurence says, may well never know “what it is that the group is after.” “Two greyhounds,” Smith remarks, “in running down the same hare, have sometimes the appearance of acting in some sort of concert. . . . This, however, is not the effect of any contract, but of the accidental concurrence of their passions in the same object at that particular time” (Wealth I.2).
Smith’s point seems to be that dogs never know “what it is that the group is after.” They are not good at answering “Why?” questions. Thus they do not attempt to instrumentally rationalize their goings-about, or anything else for that matter. What the Smith dogs share instead is a Humean “accidental concurrence of passions,” caused by something (“A hare! A hare!”) to be sure, but not amenable to any intentional description. Smith never considers the Hobbesian hypothesis of brute beasts entering into covenants and deciding e.g. that it is in their common interest to run a hare.
Joint arrangements as to who would do what job regarding a particular robbery or the logistics of power takeover are nevertheless common. They rely on covenants and contracts, if not always written ones; they require instrumental rationalization; and they require agents. The wider point of Smith’s hound example may thus have little to do with the metaphysical difference between dogs and agents. It may instead caution against a tendency to infer agent agreement from phenomenological synchronicity. For Smith, the “appearance to act in some sort of concert” is often just an appearance. Agents appearing to be acting in concert may often not be.
Explanations of revolution seem to come in two kinds. You may say that revolution is caused by concurrence in passions, by a spontaneous overflow of powerful feelings. We could call this Caninism. And you may say that revolution is the product of previously shared designs and intentions, that is, the effect of an explicit agreement. Leninism is an extreme version of this second kind of explanation. Leader and masses, Leninists notice, always seem to agree. They “are as little of a problem to each other as colour is a problem to the painter” (21, qt. cxlii). Contractualism often is a milder version of Leninism, though it may matter how the actual fact of agreement is described.
What Leninism and contractualism generally share (with allowances to late Rawls) is the idea that concurrence in intentions by large numbers of agents seems to be required to explain the bringing about of large-scale events. To be sure, for contractualism the scale model for collective action is the constitutional convention, whereas for Leninism it is bank robbery. Both models however tend to assume division of labour to explain the success of certain productions. Historical events are for both Leninism and Contractualism essentially productions. They are quasi-artifacts brought about by agents familiar with certain technical procedures, and among whom division of labour (such as who does, or can do, what) applies.
The explanations put forth by Caninism and Leninism (in any degree thereof) appear to represent two extreme ways of understanding revolution, respectively as a very general sort of event and as a very specific class of action. Leninisms emphasize the causal role of intentions to a tremendous degree; and this entails the assumption that the right sort of intention will obtain into the right sort of consequence. To that extent Leninisms seem to exclude the possibility of unintended consequences of an action. Caninisms, on the other hand, deny the role of intention and practical reason if not to human action altogether, at least to collective action. They are “straightforwardly unintentional.” Extreme as both kinds of explanation might be they end up describing collective actions much as they would describe the weather, though substituting different things for meteorology: the historical process in the case of Leninism stricto sensu, and psychological processes in the case of Caninism sensu Smithiano.
Robert Nozick endorsed Caninism, with an approving nod towards Smith. In the second chapter of Anarchy, State and Utopia, in a section entitled “Invisible-Hand Explanations,” he discusses “a certain lovely quality of the explanations of this sort,” namely that “they show how some overall pattern of design, which one would have thought had to be produced by an individual’s or a group’s successful attempt to realize the pattern, was produced and maintained instead by a process that in no way had the overall pattern or design ‘in mind’” (18). Among invisible-hand explanations he mentions evolutionary explanations of “traits of organisms and populations”; ecological explanations; Mises’ description of the markets; microeconomics; the Austrian theory of trade cycles; non-conspiratorial historical explanations of counterintuitive decisions; Hayekian theories of social cooperation, and several others (20-21).
The notion may raise the occasional eyebrow. “Having the overall pattern in mind” can both mean having an overall pattern in mind among other things, and having an overall pattern in mind with the exclusion of everything else. Nozick’s formula indicates that he might lean towards the second alternative. However, his cautious conditional tense also suggests that these are cases when one would have thought a pattern to have been produced by an individual or a group though it wasn’t. Is Nozick hinting at the possibility of agentless intention? What could such an animal look like? His theory is contained, or perhaps locked, into his elusive word “process,” as in “a process that in no way [has an] overall pattern or design ‘in mind’.” To be sure, process-talk may leave some room for action-talk; but not much, and not always.
Despite its explanatory loveliness Nozick does not commit exclusively to invisible-hand epistemology. He contrasts invisible-hand explanations with “straightforward explanations,” i.e. explanations that include “full-blown pattern-notions” (19) such as desires, beliefs and intentions. This amounts to a contrast between intentions and processes. Only processes appear to require invisible-hand explanations. “An invisible-hand explanation,” he adds, “explains what looks to be the product of someone’s intentional design as not being brought about by anyone’s intentions.” (19) Of course this means that there really could be no invisible hands, since there really are no agents. “Invisible hand” is an attribute or a garish pseudonym of “process.”
The contrast for Nozick amounts thus to a contrast between what is intentional and what seems intentional but isn’t. It is meant as a reductive correction to the ontology of actions, rather than as a distinction between kinds of actions. It relies on a straightforward opposition between actions-as-causes and events-as-causes, and possibly between human actions and natural processes. As we remarked, the ways in which we talk about revolutions often suggest facts of nature. To this extent invisible-hand explanations invite naturalized versions of history, economics and politics. They may even suggest that considerations of collective action might in time become soluble in the right kind of science, in a science of natural processes.
Would Smith have endorsed the Nozickian distinction between kinds of explanations? Was he the inspirer of what Nozick calls invisible-hand explanations? The occurrences of the phrase “invisible hand” in his writings are scant. In The Theory of Moral Sentiments he remarks that the rich, “though they mean only their own conveniency,” end up “divid[ing] with the poor the produce of all their improvements.” “They are,” Smith adds,
led by an invisible hand to make nearly the same distribution of the necessaries of life, which would have been made, had the earth been divided into equal portions among all its inhabitants, and thus without intending it, without knowing it, advance the interest of the society, and afford means to the multiplication of the species. (IV.1)
This is the argument that will recur in a famous passage in The Wealth of Nations (IV.2):
As every individual, therefore, endeavours as much as he can both to employ his capital in the support of domestic industry, and so to direct that industry that its produce may be of the greatest value; every individual necessarily labours to render the annual revenue of the society as great as he can. He generally, indeed, neither intends to promote the public interest, nor knows how much he is promoting it. By preferring the support of domestic to that of foreign industry, he intends only his own security; and by directing that industry in such a manner as its produce may be of the greatest value, he intends only his own gain, and he is in this, as in many other cases, led by an invisible hand to promote an end which was no part of his intention. Nor is it always the worse for the society that it was no part of it. By pursuing his own interest he frequently promotes that of the society more effectually than when he really intends to promote it.
Smith makes here massive use of intentional verbs and phrases (italicized in the quote). As in the Theory of Moral Sentiments, he is giving pride of place to unintended consequences. He describes a number of intentional actions (such as ‘meaning one’s own convenience’, ‘intending one’s own security’ or ‘intending one’s own gain’ or generally ‘pursuing one’s own interest’) in relation to their unintended consequences (such as ‘advancing the interest of society’, ‘promoting public interest’, and so on). Unintended consequences are for him the most relevant aspect of those intentional actions. The structure could be described as follows: good things may come from the fact that agents acting intentionally produce unintended consequences and/or do not always know the full extent of the consequences of their actions (the coupling of intending and knowing in this context actually occurs in both The Theory of Moral Sentiments and The Wealth of Nations.)
There are two aspects to this structure: one, belonging to ethics, namely that the nature of intention is separate from the nature of consequences, vz. bad intentions can yield and often do yield good consequences. This essay has no wish or ideas to contribute to modern moral philosophy. Another aspect, however, is relevant to Smith’s theory of action and may further our discussion. It intimates at least two, and possibly three, claims. The first, very general, is that intentional actions often produce unintended consequences, that is, that any agent may “promote an end which was no part of his intention.” The second claim is that certain intentional actions have necessary contents; for example, if someone works for a company (a “society”) then he “necessarily labours to render the annual revenue of the society as great as he can” (emphasis added). The agent’s intentions are thus necessarily bound to the particular context of his actions (you never quite f; you company-f, so to speak). A third claim, to a certain extent combining the first two, would be that even when an action has necessary contents that action produces unintended consequences; working for a company entails producing unintended consequences of the kind that befalls company-work (as opposed e.g. to flying kites or fighting battles).
Only apparently is this third claim trivial. Smith’s theory seems to exclude the possibility of mere self-interest, or of self-interest in the familiar, normative, sense of the term. A full description of the motives for an action always requires a description of a particular context (and this entails that there is no fixed repertoire of possible motives for action; self-interest is not a type of motive). Of course one may wonder how particular “particular” need be; and this requires a separate discussion. Still for Smith it is the fact of there being institutes such as companies rather than any psychological laws governing intention-forming, or principles establishing action kinds, that suggests that mere self-interest is not a fruitful explanatory option for economic occurrences.
An invisible-hand explanation then, for Smith, would always suggest that intentional actions are required to account for unintended consequences, and so that invisible-hand explanations, contra Nozick, always require “full-blown pattern-notions” such as desires, beliefs and intentions. Thus Smith has no use or need for the distinction between “straightforward explanations” and “invisible-hand explanations,” and so between intentions and processes. The reason is that for him intentionality remains an ineliminable idiom of practical explanation. It is not to be assimilated to any form of social thought, let alone locked into any of its verbal creatures, such as Nozick’s (and Lenin’s) process.
This takes us back to our initial example of the Russian Revolution. Caninism and Leninism, or so we argued, share a similar view of collective action. You may now better see why, and thus why they combine early and often. Relevant differences notwithstanding, Caninism and Leninism both provide attempts to substitute causal non-intentional accounts of action for causal-intentional descriptions. They do so by eliminating consideration of the connection between intention and unintended consequences in collective action, and consideration of action particulars. This however is what Smith suggests can never quite be eliminated. No state of affairs brought about by the actions of a great number of agents, such as the functioning of the simplest market, can be accounted for without considering that their actions are both intentional and particular, and so that what obtains is both connected e.g. to the institutes of trade and caused e.g. by the unintended consequences of trade-related intentional actions. Human interactions can thus neither be fully naturalized nor fully regulated. Wanting to correct markets by limiting the number of intentional acts allowed is like believing that enlarged willpower or a very effective penal system would eradicate certain consequences of human action, not to mention undesirable forms of economic exchange.
Perhaps in our dealing with collective actions we should follow Smith’s suggestion. Perhaps we should attempt to retain in our historical descriptions, even and especially those of agents acting for all sorts of reasons and aiming at all sorts of goals, what Anscombe, talking about animals, calls the “manner perfectly characteristic of [the] use of intention concepts” (§47). We should do it not because there are collective intentions (there aren’t) or because agents are all like Smith dogs (they aren’t). We should do it because, as Anscombe also shows, “the term ‘intentional’ has reference to a form of description of events” (idem). Given that the alternative varieties of description currently on offer, including Caninism and Leninism, raise substantial difficulties, we ought to take her suggestion seriously, and give intention an ampler chance. Philosophers of late have only wanted to change the world; the point now seems to be to describe it.
 A substantially different version of this essay was presented under the same title as a lecture at the University of Chicago, May 2014.
 Different solutions that also bravely attempt to “make corporate agency intelligible and exciting” have been put forth by Philip Pettit, and alas by myself (Chapter 4).
 Passage in brackets only in original ms. version, 22
 See Bratman, esp. Ch. 4.
 See Rawls 49 and 49n3, as well as his remarks on Hegel and Smith, 457.
 In Anarchy, State and Utopia Nozick does not purport to offer any substantial definition of invisible-hand explanations (20, and 336-7 n13).
 The earliest comes up in an essay on the history of astronomy (written before 1758), where Smith uses the phrase to tellingly make fun of those who believe in “the invisible hand of Jupiter.”
 There is an obvious symmetry between Smith’s “realist” claim and Anscombe’s well-known discussion of “under a description.”
 How particular need we be about circumstances, about companies, kites, or battles? Extreme particularity may be morally vile; extreme absence of particularity is always descriptively inept. Extreme particularity is a tool for universal unimpeachability; it has no use for ethics. Absent particularity assumes universal peccability; it really describes no actions. This difficulty is in turn connected to what Anscombe calls “a natural consequence of the uncountability which is characteristic of the concept of action or event” (229).
G. E. M. Anscombe. 1957. Intention. Cambridge: Harvard UP. 2000
___________. 1979. “Under a Description.” Noûs. 13:2. 219-33.
George Berkeley. 1734. A Treatise Concerning the Principles of Human Knowledge. K. Winkler ed. Indianapolis: Hackett. 1995.
Michael Bratman. 2014. Shared Agency: A Planning Theory of Acting Together. Oxford: Oxford UP.
Joseph Goebbels. 1933. Michael. Ein deutsches Schicksal in Tagebuchblättern. München:Eber. 21. Passage quoted and translated by E.M. Wilkinson & L.A. Willoughby. “Introduction.” Friedrich Schiller. On the Aesthetic Education of Man. Oxford: Oxford University Press. 1967. cxlii.
Ben Laurence. 2011. “An Anscombian Approach to Collective Action.” Anton Ford, Jennifer Hornsby & Frederick Stoutland (eds.), Essays on Anscombe's Intention. Cambridge: Harvard University Press. 270-96, and ms. version, 30 pp.
Robert Nozick. 1974. Anarchy, State and Utopia. NY: Basic Books.
Philip Pettit. 2018. “Corporate Agency: the Lesson of the Discursive Dilemma”. M. Jankovic & K. Ludwig (eds.) The Routledge Handbook of Collective Intentionality. NY: Routledge. 2018. 249-59.
John Rawls. 1999. A Theory of Justice. Revised Edition. Cambridge: Harvard UP.
Adam Smith. 1759. The Theory of Moral Sentiments. (D.D. Raphael & A. L. Macfie, eds.) Indianapolis: Liberty Fund. 1982.
___________. 1776. An Inquiry into the Nature and Causes of the Wealth of Nations. (R.H. Campbell & A. S. Skinner, eds.) Vol 1. Indianapolis: Liberty Fund. 1982.
Miguel Tamen. 2001. Friends of Interpretable Objects. Cambridge: Harvard UP.