kinds of action explanation

I've noticed over the years that scientifically minded psychologists tend to collapse all explanation and understanding into a causal form. If I ask 'how?' or 'why?' questions about human behaviour, thought, or feeling, what I'm inexorably after, they seem to suggest, are causes. The idea of there being different kinds of causes - e.g. the three kinds met with in the bio-psycho-social model - is of course utterly widespread. But the idea that many psychological explanations may be non-causal in character seems very often to be missing from view. And all too often explanations which are not at all causal in form get confusingly dressed up in causal garb.

My own contention is that most of what is valuable in psychological explanation is non-causal in form. But let me begin with a word about this term 'cause'. In English today we tend to describe causes as what makes something happen. "Why did the chimney pot fall?" "It's because the heavy rain dislodged the mortar". Or: "because the brick layer didn't put the right mortar in." If we followed Aristotle we might say that whenever, in response to a "why?" question, we come up with a "because..", what follows after the "..." will be a cause. And Aristotle distinguishes four kinds of cause: efficient, formal, material, and final. But of these, only efficient causes really correspond to what we typically call "causes" today. (This is just a semantic difference, not a substantive disagreement of any sort. ... An exception would be when we talk about, say, protesting for 'a just cause', where causal language is used to articulate matters teleological.) When the the other kinds of Aristotelian cause are invoked, we now typically talk of "non-causal explanation". This should all become clearer in a moment.

In what follows, then, I set out various different ways that we answer "why?" questions when we're dealing with human behaviour.

Causes

As noted above, causal explanations are offered in response to questions like "what made that happen?" "how did that come about?" "what precipitated that?" "Why?" questions, let's note, are not only concerned with matters of meaning; sometimes they truly are questions about causes: "why did the tree explode?" "it was hit by lightning". Causes are often events ("Why did the monkey keel over?" "It ate a poisonous insect"), but can sometimes be objects, and sometimes be the absence of such events or objects ("Why did the car seize up?" "The oil hadn't been replaced"). 

Talk of the causes of human behaviour is typically talk of what interferes in our agency. "Why did Geoff leap like that just now?" "Well the train blew its whistle as it came out of the tunnel, causing him to jump". "The hypnotist caused him to make a fool of himself." But we may also take an interest in causes when we're interested in agency's exercise. However when we do that, we step outside of psychology's ownmost domain. We might for example be interested in what inside our bodies (e.g. what goings on in our CNSs) causes the movements that partly constitute our actions. ... Or, sometimes people talk of agents causing their own actions ('agent causation'). There's perhaps nothing wrong with that, but it should be noted that it rather boils down to our making clear that our action in such cases is free, i.e. isn't being caused by anyone or anything else.

One used to often hear the idea that causal relations are law-like. To say that B is caused by A, it was said, is to imply that, at least under some or other description, Bs are reliably caused by As all else being equal. It is true that it's hard to see how, if everything else really is equal at t2, and if A caused B at t1, A wouldn't do so again at t2. But it's not clear to me that this helps much with the project of developing a reflective understanding of causal explanation. (The idea of a distinctly causal form of "necessity" such that A somehow necessitates B is equally obscure to me.)

It is also common to hear the idea that causal explanations essentially invoke distinct relata. What there is to this idea is perhaps the idea that formal causes are not efficient causes. (My hailing the cab did not cause me to raise my arm; in this context it was me raising my arm.) But when we think of the diverse actions and events which we typically collocate with the notion of "cause" (cutting, scraping, burning, pushing, breaking...) it's not obvious that causes and effects really are that distinct. "Why is this piece of paper in two pieces?" "I cut it with the scissors". Here the activity designated with the causal verb "cutting" is not a separate event than the cleaving of the paper.

These days one sometimes hears about counterfactual analyses of causation. The idea is that part or all of what is meant by the idea of A causing B is that were A to not have happened, then B wouldn't have happened. The idea perhaps is that if B were going to happen anyway regardless of whether A occurred, then it's peculiar to think that A really contributed to B. A problem I find with this is that it risks covering cases which intuitively we wouldn't count as causal. For example, we don't tend to find Russian sentences in French language books unless they are quotations. But does this mean that quoting the Russian sentence causes it to appear? No: the quoting is constitutively, rather than causally, related to the sentence. (In the Aristotelian terminology, the sentence is the material, not the efficient, cause of the quotation is the sentence.) And if - as I suggest below - we are thinking of the relation of intentions to actions, it's true that were someone to not intend to X it's unlikely that they would do X. But to my mind this speaks to the rarity of unintentional action in human life rather than to a causal relation between intention and action.

Dispositions

It's sometimes said, especially by those who are really only interested in causal explanations, that reference to dispositions is non-explanatory. One can see why. To answer "Why did Tim fall asleep?" with "He's disposed to fall asleep" may often be rather empty. But note, even toy examples such as that of Tim's sleepiness needn't always be simply empty, and this is because they in effect serve to rule out certain other kinds of explanation. Everyone else stayed awake just fine. So why did Tim fall asleep? Well, it's not that he was, say, up all last night. It's just that he often falls asleep like that. It's in his nature. He's like that. References to dispositions are in fact rather like this. "Why did the bubble burst?" "Bubbles just are apt to do that": Hardly explanatory, but still illuminating for the bubble novice. Such a disposition-citing answer may be little more than a way of saying "silly question!" But "Why did the rock break?!" "It turned out to be a particularly fragile rock": Less of a silly question, and illuminating because reference to ('efficient') causes ("it was struck extra-hard") are effectively ruled out.

References to character typically provide dispositional explanations of behaviour. Through these references we become acquainted with what someone is like. With how they are disposed - how they tend - to act and react, that is.

References to someone's desires are at least somewhat dispositional. To be thirsty is both to feel a certain way (parched, perhaps) and to be disposed to drink. We may have sensations of thirst or hunger but, in certain instances, not be disposed to drink or eat. But this isn't to say that "thirst" and "hunger" are primarily the names of sensations (nor that sensations aren't anyway partly picked out in terms of the behaviours to which they dispose us). We can see the dispositional connection between desire and action in the following way. If someone is thirsty, is it so much as intelligible that they wouldn't drink if a) they had the opportunity, b) there wasn't something they desired more than drinking, c) there wasn't something wrong with them? If there were no dispositional relation between desire and action here then it ought to be possible to say "Yes she was thirsty, and no it's not that she wanted anything else (e.g. staying in bed) more than drinking, and no there wasn't anything wrong with her; it's just that she didn't drink despite being thirsty." In truth, however, this is unintelligible.

To use the Aristotelian language again: referencing a disposition is referencing something like a formal, rather than an efficient, cause of something. We understand why something is done by understanding that it's in the nature of the thing. Why did the rock break apart? It's a fragile rock. Not an empty claim, since other possibilities were available. Perhaps a different, far more robust, rock broke not because it was fragile but because it was hit by Thor's own hammer. 

Reasons

Most often when we ask why someone did something, what we're after are their reasons. In being after those, we're sometimes after what they're after. That's to say: we're sometimes interested in what they were aiming at, what they were trying to achieve. "Why did you become a psychologist?" "I wanted to make myself useful to people who were in psychological trouble". "Why did the chicken cross the road?" "To get to the other side". In such cases we're particularly interested in what people themselves say about their own actions. And to respond sincerely to a request for reasons just is to do what's called "giving one's reasons". We won't always accept such answers even when they're sincerely offered, since at times our behaviour is motivated (see below for motivations) in ways which defeat reason self-ascriptions. On the whole, though, we shall, and it's important to note that those times when reasons are defeated by reference to (e.g.) unconscious motivations are necessarily exceptional.

At other times, when we're after their reasons, we want to know why what they did is the thing to do. When I ask you, who are teaching me maths, "Why did you carry the ten to the next row?", I'm not interested in what your aims are ("To do the sum correctly"). Instead I want to know why one should do what you did here. It's worth noting though that reasons of this sort are typically rather less personal than those of the first sort. We are interested in someone's reasoning, but here we're more interested in the reason than in their reason for doing what they did.

The character of reason-giving explanation is sometimes teleological. Sometimes, that is, an action is made rationally intelligible to us when we see it as aiming at a worthwhile goal. The other side of the road is by itself a rather thin reason for our chicken's road-crossing, but when we bear in mind the nice pile of corn on the far verge the whole enterprise makes a good deal more sense. Here we're interested in the chicken's aims rather than its action's precipitants. Or, in Aristotle's more capacious sense of "cause", here we're interested in the final, rather than the efficient, causes of action. Other reasons are what Anscombe calls "backward looking": "Why did you punch him this afternoon?" "Because he punched me this morning." Once again the idea isn't that the prior punching caused the later punching, but instead that the earlier punching was the rational ground of the later punching.

Reason explanations of action often belong, one might say, more to ordinary life rather than to psychology. Or, they do so long as one has something like the following understanding of "reason" and "psychology" in mind: "reason" meaning what a person sincerely and definitively says about their action's end, "psychology" meaning the understanding we develop of human beings to account for why we don't act in rational ways. 

Context-Placing Explanations

A typical way that we come to understand an action is through what Julia Tanney has called 'context-placement'. Someone walking past a sixth form classroom is puzzled as to why the teacher has just written "anal" on the board. "Eww that doesn't look right!" thinks our observer. Their puzzlement shifts somewhat when, adjusting their angle of vision, they see the word "philosophical" written before "anal". And then they see the teacher retrieve the pen she had dropped, and come back to the board to complete the phrase "philosophical analysis". The act of writing "anal" is made intelligible by placing it in an already intelligible context. The understanding dawns not because they see the precipitating cause of the writing, but by placing it in an intelligible context. (If the board in question had instead belonged to a gastroenterologist there mightn't have been a puzzle to be resolved in the first place.) Clearly, writing the word "analysis" on the board didn't cause the writing of the word "anal" in the sense of "cause" which I've been stalking in this post. It didn't precipitate it; it didn't come before it (but, in fact, was only finished after "anal" was already written); in this context it constitutes it. Understanding arises not from seeing what caused "anal" to arise, as it were, but from the prior understanding that analysis is the kind of thing that philosophy teachers talk about. We typically mark this by talking of meaning. "What's the meaning of all these vigorous arm movements young fellow?" "Well sir, I'm busy hailing a cab!" It's their human intelligibility that's at stake. Once again the textual analogy is helpful. When we understand a text we aren't understanding it in terms of its causes, but in terms of what it's saying. 

It's sometimes suggested that such reading of meaning amounts to interpretation. Whilst we are free to use the word "interpret" that way if we really must, there are, Wittgenstein taught us in Philosophical Investigations §201, reasons to avoid doing so. Think of when it is that we're in need of interpretations, and what they then consist in. We need them when something is obscure to us. And they consist in the substitution of an opaque piece of discourse for another that's clear in meaning. If we think of interpretation this way, it should be apparent that no interpretation is involved when we understand the clearer piece of text. Instead we simply "get" what it's saying without employing any method or means. Much of our ordinary understanding of people is like this. We apprehend the actions of others in the joint social and environmental contexts of our lives. We take a certain facial configuration over time as a smile, automatically seeing it as expressive of happiness, all in the context of meeting up with a long-departed loved one. This happiness doesn't need to be fathomed; the fact of it doesn't require interpretative labour. Our perception of others is itself run through with such meaning, so we don't have to work it over afterwards by thought.

It's not only when thinking of reasons, but also of the related concepts of intentions and motivations, that we are interested in matters teleological. We understand the arm movement as pumping: this is what it aims at. We understand the pumping as an act of poisoning when we see the intentions that specify the acts goal. When we ask for reasons we are interested in the reasoning which someone would give for why they act as they do. So too when we're interested in intentions. And it's worth noting that the concepts of "intention" and "trying" gain their special significance in our lives because of a possible contrast between what we did and what we intended or tried to do. "Why did you just hit your thumb with the hammer?" "Well I didn't intend that; what I was trying to do was to hit the nail!" In this way the concepts serve to preserve the intelligibility of action which may otherwise look deranged to us. 

The concept of "motive", though, covers a somewhat different territory, and is especially important for psychology. It's also often more relevant when what we're aiming to do is to understand not so much action as behaviour. (I take it that the concept of "action" belongs more properly to ethics and law, and that of "behaviour" to psychology.) Thus we can be motivated in ways that we don't ourselves recognise. Explanations of behaviour that reference motive are still teleological. 

Intentions

Such explanations of action as reference intention take a couple of forms. First, consider the kind of intention which someone avows. If one is seduced by a causal model one might think that the avowal announces the presence of an inner item of some sort - which inner item will, all going well, later cause the behaviour which is thereby intended. With that model there are famous difficulties - including the difficulty in identifying any such item when we look within, and the related difficulty of separately individuating the intention and the action as causal explanation requires. But rather than thinking of intention as a thing of some sort, we can instead think of avowals of intention as promissory notes: "I intend to change the bedlinen later" means something like "you can count on me to do it". Another use of "intend" signals such behaviour as is non-accidental. "Did you mean to shove him into the river?" "No! I tripped on a root and stumbled into him!" Our action did not have its effect as its telos, and it becomes intelligible to us when we see that it wasn't intended. 

Explanatory references to actual intention can be both teleological and context-placing. Recall Elizabeth Anscombe's well-known example from Intention of the water-pumping person. Tim is moving his arm. Why? To pump water? Why? He is pumping poison into the house. Why? To kill the Nazis inside. We make sense of the doing of one action - the arm moving - not by looking to just any of its effects, but instead by looking to what the person was doing (poisoning Nazis) in doing what they did (moving their arm). The actions become 'larger' as they progressively incorporate the intended aim of the 'smaller' acts. The explanation proceeds by redescription of the action, and the redescription relates the action to the sincere avowal of the agent. We are particularly interested in what people say about the ambitions they had when pursuing their own past, present, and future actions, and our interest takes the form of considering their sincere avowal as largely definitive of the actual character of the action. (A remark on the concept of "intentional action".)

Motives

Regarding this last point there's an interesting contrast between the concept of "motive" and of "intention". For sure there's some slack in how we use the words, but we're rather more likely to talk of unconscious motivation than of unconscious intention. Relatedly, whilst "motive" and "reason" are sometimes used interchangeable, we speak properly of motives even when reasoning doesn't really come into it. (Reasons bear a closer connection to reasoning than do motives.) "Why did George kill James?" "Well he knew perfectly well that it wasn't the right thing to do. But he was motivated by jealousy and a lust for revenge." Our new year's resolutions may be full of good intentions, but our being motivated by our vices makes it hard to keep to these intentions. Intentions can remain in play even whilst motivation rises and falls. The distinction is also legally important. George is charged with murder rather than manslaughter because he intended James to die, but considerations of motive - whether the act was motivated by jealousy or anger or instead by greed - are often irrelevant when it comes to liability's discernment.

Regardless of such differences between motivation and intention, explanations which reference motivation have a similar character to those which reference intention. Somebody performed an action. We want to know why. There are various different possible motives that anyone, or perhaps even the someone in question, could have had for performing the action. Were there not then there would be no explanation required and no place for a concept of motivation. The explanation narrows the field. It places the action in relation to one particular way of being human - i.e. in relation to one set of intelligible drives and ends. In this way the intelligibility of the explanation depends on the intelligibility that motivated action already has for us. What I mean is: when we see someone act in a straightforwardly intelligible way, the motivation of it quite obvious to us - Gloria is hungrily devouring her dinner - then we already understand the action without any need for explanation. So too when Gloria is acting more strangely: here we will need an explanation if we're to retrieve our sense of her rationally or at least (in the case of motivated irrational acts) humanly intelligible acts. The understanding arises by placing the act in an already intelligible context. However the intelligibility of this more broadly described act doesn't itself depend on the explanation. Nor does it depend on our seeing the efficient causality alive in it. Instead we now grasp it as the sort of thing which we accept at least some humans just are disposed to do, or as something with a rationally intelligible end.   

Postscript on an argument by Davidson

In the 1960s the American philosopher Donald Davidson, a chief pundit of causalism in the theory of action, came up with an argument against non-causal accounts which to my mind has been surprisingly influential. It goes like this: Suppose that Geoff goes into the George & Davis ice-cream shop. There are many reasons for which he could have done this. His girlfriend works there and he likes to go see and surprise her. He's hungry. Oh, and he wants to escape this peculiar man who is following him down the street. All of these desires of his are actual - they all, as Davidson might put it, in combination with the relevant knowledge or belief (he knows his girlfriend works there, he believes he has the money to buy an icecream, he thinks the man won't follow him into such a shop), 'rationalise' his action. And yet on this occasion he acts for only one of these reasons. The prima facie reasons give him a motive - but only one constitutes an intention. Thus if you ask him why he went into the shop he would say 'To escape the man', albeit that he will acknowledge that he genuinely also had the other prima facie reasons. Now, according to Davidson, we can't understand the difference between the reason for which someone acts, and a reason which would, as Davidson might say, merely 'rationalise' the action, without the aid of 'causal glue' to stick the first, 'operative', reason and the action together. It is causality, he says, which makes a mere relation of intelligibility into something actually determinative of the action.

Perhaps somewhat peculiar about this scheme is Davidson's treating the desires and beliefs in question as reasons ('primary reasons' he calls them). The end adverted to with 'To escape the man' is after all hardly the (or any) kind of thing or event which could intelligibly function as a cause. But let that pass, since we can all readily agree both that actual reasons can only be intelligibly ascribed to someone who at least has the requisite desires, and that 'I did it because I wanted to escape the man' is anyway a statement about a reason. And since I've already suggested that such reasons don't stand to actions as causes do to effects, what's surely necessary for me to show at this point is instead that Davidson's 'causal glue' isn't after all required. But this seems rather easy to do. First of all, it is, let's remember, perfectly intelligible that Geoff acted on all three desires. And in fact, absent reasons to think that certain of his desires are in any given case (in the causalist lingo) inoperative, this should I suggest simply be our default presumption. And what reasons for 'inoperativity' might those be? They will I think be that despite, say, Geoff genuinely liking seeing his girlfriend he is today in a real rush, or despite his being hungry he is trying to better manage his diabetes. In other words, what makes such desires inoperative, what makes them mere wishes rather than realised intentions, is not their failing to stand in causal as well as sense-making relations to the actions in question, but simply their less significant place in a larger set of sense-making relations. To argue against this it would be necessary to think of a situation in which there are, given his general desires, various reasons why it would make sense for Geoff to do X, but, despite none of them being trumped by other desires (to not be late, to not put on weight), they nevertheless cannot to be counted as reasons for the action. I can't myself think of such a situation. Can you?