Here's how the American philosopher Donald Davidson described getting up one day:

Davidson was interested in how we distinguish between the things we do, such as washing, shaving and dressing, and the things that happen to us, such as stumbling on the edge of a rug. In his own words, the question is this: ‘What events in the life of a person reveal agency; what are his deeds and his doings in contrast to mere happenings in his history; what is the mark that distinguishes his actions?’

Davidson wasn't the first philosopher to raise this question. Descartes wrote about it in his treatise Les Passions de L'Âme. His approach was more physiological than Davidson's, and to my mind it was better for that reason. (It will become clear why I think this in due course.) Descartes regarded the problem as explaining the difference between the changes in your body that you yourself cause personally, such as the movements of your legs or lips when you walk or speak, and the changes that happens in your body that you do not cause personally, such as the growth of your toenails, the contraction of your heart, or the way your foot bobs up and down slightly when your legs are crossed. But it is essentially the same problem, and Descartes and Davidson solve it in essentially the same way. They both claim that every act a person does is intentional or stems from an intentional movement of some kind. So intention – or volition as Descartes calls it, he uses the French word volonté – is the mark that distinguishes a person's actions.

I shall argue that this was a plausible solution from Descartes's point of view – that is, taking some of the main elements of his philosophy for granted – but it is not plausible otherwise. So this is a case – and not by any means the only one – in which a philosophical idea has long outlived the intellectual context in which it made good sense.

Descartes believed that each person – each one of us – is a mind or soul that interacts with a body. For example, your body causes your sensations and perceptions, when you feel hungry or thirsty, when you see the bread on your plate or smell the coffee in your cup, and when you feel and taste it in your mouth; and you cause the movements in your body that occur when you put the bread in your mouth and chew it and when you sip from your cup. But you yourself are a purely spiritual entity, and whereas your body has limbs and organs you are completely simple, without parts. Unlike bodies, minds or souls are not made of matter, and do not behave in the ways that things that are made of matter behave. For example, they do not grow, move, or get hot or cold. According to Descartes, all that minds or souls or persons do is think.

If we assume this is right, intentions or volitions provide a plausible way of distinguishing between activity and passivity in our lives. First, any event caused by a person – in other words, any event that ‘reveals his agency’ – must either be, or be caused by, one of his own thoughts. (For comparison: if the only thing Parliament could do was legislate, then anything that reveals its agency would have to be, or be caused by, legislation; and if the only thing Tiger Woods could do was play golf, then anything that reveals his agency would have to be, or be caused by, his playing golf.) Second, presumably the thoughts that cause our active movements are active thoughts. Third, our active thoughts, Descartes says, are our volitions:

In fact the steps in this argument are not equally convincing. But the first is incontrovertible, and I shall leave it as an exercise for the reader to decide what may be wrong with the other two.

Descartes's way of singling out the events that reveal a person's agency was adopted by John Locke, and in the eighteenth century a number of philosophers expressed the same idea. For example, Thomas Reid said that if an act was done ‘without [a person's] will and intention, it is as certain that […] it ought not to be imputed to him as the agent’, and Jeremy Bentham insisted that ‘if the act be not intentional in the first stage, it is no act of yours’.Footnote ³ The idea is still popoular with philosophers today, mainly because Davidson gave it a new lease of life.

Here is Davidson's argument:

The argument is interesting and quite persuasive when one reads it for the first time, but not everyone has been persuaded by it. For example, the American philosopher Harry Frankfurt rejects the idea that intention is present whenever people act:

Frankfurt insists that the contrast between these kinds of movements is the same whether we are concerned with a spider or a human being, and hence that it cannot ‘be explicated in terms of any of the distinctive higher faculties which characteristically come into play when a person acts’, such as intention. But unfortunately he stops there: ‘the general conditions of [human] agency’, he comments, ‘are unclear’. I think he is thinking along the right lines. But we can progress further, and define human agency in a more satisfactory way than Descartes or Davidson does, if we approach it via the agency of complex things in general.

Complex agents can be animate (a spider) or inanimate (an engine), and complex animate agents can be divided into three kinds: organisms; parts of organisms, such as cells and bodily organs; and superorganisms, that is, groups of organisms, such as colonies of ants or bees, that act as functionally integrated wholes. An institution – such as a business, a university or a government – is an unusual kind of complex agent, being an inanimate whole with inanimate parts analogous to organs, and animate members somewhat analogous to cells.

Every complex agent has various active and passive powers, in other words, abilities to cause and liabilities to undergo various kinds of change. Some of them are purely aggregative. For example, the weight of an engine is simply the combined weight of its parts. But others are not. For example, an engine's ability to drive a flywheel depends on the interaction between functionally differentiated parts. We want to explain what distinguishes the kinds of motion that are attributable to the non-aggregative agency of a human being as a whole, such as the motion of your legs when you walk. But human agency is not the best place to start. In fact, the most straightforward cases are institutions, because their powers and the powers of their parts are conventionally defined. For example, a university has the power to award degrees, which none of its parts or members has alone. But the members and parts of the university – professors, examining boards, administrative offices and so on – have to follow complex procedures when it does so, as laid down in the university's statutes and regulations. The power to award degrees belongs to the university as a whole, but the exercise of this power depends on the integrated exercise of the distinctive powers accorded to its members and its parts.

Superorganisms also provide examples where distinguishing between the agency of wholes and functionally differentiated parts is relatively straightforward. For example, when bees swarm, they gather in a convenient place, for example in the branches of a tree, while scouts set out to find a place for a new hive. When a scout encounters a candidate place it returns to the swarm and communicates what it has discovered by performing a ‘waggle dance’, running through a figure-of-eight pattern, vibrating its body laterally as it moves through the central axis. Different features of the dance communicate the distance and direction of the place and how attractive or suitable it appeared. Other scouts respond to dances by investigating the places they advertise. More attractive candidate places elicit dances that are longer and more vigorous, which in turn are more likely to elicit a response from other scouts. So, gradually, and without any individual comparing one candidate place with another, the support for the more attractive places grows. When a sufficient number of scouts are indicating the same place with their dances, the swarm goes there and builds a hive.

In cases of this kind, it is not difficult to distinguish between the agency of the colony or swarm and the agency of its parts, even though there is no constitution or a set of regulations to refer to, where their various powers are defined. As in the case of a university, the swarm as a whole has a power that none of the individuals it is composed of has alone – the ability to select the place for a new hive – but the exercise of this power depends on the interaction between functionally differentiated parts.

The same model applies to the agency of a spider, which – like any multicellular organism with specialized tissues – is in effect a highly integrated colony of functionally differentiated but genetically similar cells. The similarity between an animal and a colony is especially clear if the individuals that make up the colony are compared with organs as opposed to cells. The colonial medusa Nanomia cara illustrates this point. As the biologists John Maynard Smith and Eörs Szathmáry explain,

It differs from a colony of bees in that it develops from a single fertilized egg, so the parts of its body are as genetically similar as those of an single insect or vertebrate. But unlike the organs of higher animals, its body parts evolved from individual organisms, similar to the present-day Hydra.

The lesson of these examples is that the agency of complex things with functionally differentiated parts depends on the integrated operation of these parts, rather than on the operation of a specific part. A spider spins a web, eats its prey, selects and copulates with a mate. Some of these activities involve parts specifically adapted for them, such as poison and silk glands. But since they all involve complex interactions with the spider's environment – and mostly with moving targets – they all involve the integrated operation of its sensory and motor systems, as well as the metabolic systems on which its life and activity continuously depend. One option is therefore to regard the movements that result from the integrated operation of its motor and sensory systems as attributable to the agency of the spider as a whole. This would mean that the spider is not moving its legs itself if a boy makes its legs move by pulling on threads he has tied to them, or by using an electrode to stimulate the nerve tissue that controls them. But it is moving its legs itself if the boy tries to catch it, so that it scurries away. These movements of the spider's legs would be attributable to the agency of the spider as a whole because they result from the integrated operation of its motor and sensory systems.

If this is plausible in the case of spiders, is it also plausible in the case of human beings? Not quite. Activities which depend on the integrated operation of motor and sensory systems are attributed to a human being as a whole, as opposed to particular organs, tissues or cells – feeding, sex, locomotion, communication, and so on – but these activities involve the physiological systems that are responsible for our intellectual and emotional lives as well. So it is better to think of the integration of motor and cognitive systems as the mark of human agency, using ‘cognitive’ with the same broad meaning as it has in ‘cognitive science’. Adopting this proposal involves some tidying up of our usual ways of thinking and talking about human agency, but it matches it more closely than the idea that intention is definitive of human agency and it also has the merit of being consistent with a general conception of agency by complex substances with functionally differentiated parts.

The cognitive-motor theory matches our usual ways of thinking and talking about human agency better than the intention theory, because many kinds of human action clearly do not stem from an intentional movement and (in Davidson's terminology) cannot be described under an aspect that makes them intentional, but do involve the integrated operation of sensory and motor systems: for example, reflex acts, such as ducking or drawing back one's head involuntarily to avoid a blow, or making an involuntary adjustment to one's posture to maintain balance; habitual action, for example, verbal tics such as echolalia, the automatic repetition of words and phrases spoken by the person one is conversing with, or interspersing speech with words or phrases like ‘like’ and ‘you know’; some uncontrolled action done in abnormal or pathological states of mind, such as panic or psychosis; and, most importantly perhaps, the spontaneous expression of emotion in gestures, vocalizations and facial expressions, such as smiling, scowling, pouting, shrugging and laughing or crying out with pleasure or pain.

It is true, of course, that a great deal of behaviour expressing emotion is intentional, some of it spontaneous and not fully controlled, for example kissing a lover or throwing crockery at a spouse. It is also true that many changes in the body that express emotion are not imputed to the person as agent, for example, blushing and shedding tears. And others are borderline cases or perhaps treated inconsistently. For example, we do not regard goosebumps as the result of personal agency, but we describe a dog as raising its hackles, and seem to think of this as an act, on a par with growling or baring its teeth. But that leaves a large range of cases that are undeniably unintentional expressive acts, such as the ones Norfolk mentions in a speech about Cardinal Wolsey in Henry VIII, which Darwin quotes in The Expression of the Emotions in Man and Animals:

The truth is that most human acts, or most that matter enough to be reported or recorded, stem from an intentional movement, with the important exception of spontaneous expressions of emotion. The reason is simple. In most cases, movements are considered unintentional either because the agent wasn't aware of making them, or because he wasn't able to control them. But it is unusual for us to move parts of our bodies without being aware of doing so, or to make movements we are unable to control, at least when these movements are liable to have significant consequences. It is even more unusual for us to move our bodies unintentionally in ways in which we cannot also move them intentionally, although even this may happen on occasion. For example, Darwin suggests that few people are able to control the muscles involved in in the facial expression of grief.

So most significant human acts stem from an intentional movement, but not all. And even if every human act without exception stemmed from an intentional movement, it would still be a mistake to define human agency in terms of intention. For comparison, if every human act was selfish, or stemmed from a movement of the agent's body that had a selfish purpose, we wouldn't be bound to say that this is why it qualifies as an act. If every human act involved intention, it would still be a mistake to think that intention is the key to understanding human agency because human beings are complex agents with functionally differentiated parts and the non-aggregative agency of complex agents with functionally differentiated parts always depends on the integrated operation of these parts, and not on the operation of a specific part. This is true of human beings in just the same way as it is true of other animals, colonies of animals, plants, institutions and machines. So if a basic kind of human activity that involves the integrated operation of cognitive and motor systems, such as feeding or copulating, were only conscious and controllable to the extent that breathing is, a smaller proportion of human acts would involve intention than is actually the case, but human agency would be defined in exactly the same way.

This does not prevent us from asking which part of the body initiates motion in a limb. We can still ask Locke's interesting question, ‘my right hand writes, whilst my left hand is still: what causes rest in one, and motion in the other?’,Footnote ⁸ and choose between Aristotle's answer, which is the heart, and the Stoic philosopher Chrysippus's, which is the brain. We can even distinguish the part of the brain that causes motion in a hand from the part that makes the heart contract. But this will not tell us why we assign one kind of motion and not the other to the agency of a human being as a whole. The key to answering this question is integration, and not the activity of a specific organ or mental faculty or a specific kind of thought. It was impossible for Descartes to understand this, because he conceived of a person as a simple entity, without parts. But it should be obvious once we abandon this idea, because the activity of any part of a complex agent is exactly that: the activity of a part as opposed to the activity of the whole.

Next time you're woken up by the sound of music on the radio, doze a bit, then get up, brush your teeth and get dressed, pour yourself some coffee, stumble on a loose tile in the kitchen, and spill your coffee while you're fiddling with your phone, you can decide whether you agree.