One and a half answers - Personhood

Reader beware! While most of the material on this site deals with leading or at least, recognised theories, this is where I get to tell you straight out what I personally think.


the pale cast of thought

Existential angst - or that old groin strain?

The idea of "personhood" is essential to our normal understanding of the world. It bears on many immediately practical issues, helping to determine how we deal with property, punishment, and vital medical procedures, for example. Doubts about our own personhood seem to be self-refuting: only people, surely, can doubt anything. But somehow a clear, robust, and intuitively appealing account of personhood remains elusive, and its significance and even its foundation in reality have been challenged from more than one direction.

To a na´ve eye the key point about people, the difference between people and mere objects, is that objects don't move by themselves. The movement of a football across the pitch is visibly no more than the result of the preceding kick: a kind of continuation of the kick, in fact. The swing of the player's leg, by contrast, seems to arise spontaneously out of a wish to score and plans for doing so. The voluntary action of a person seems to divert the natural flow of cause and effect.

This perception that willed action is somehow an exception to the normal run of causality underlies the moral concepts of freedom and responsibility. We act freely because our actions are somehow determined by our wishes, not purely by preceding events; equally, we are responsible for those actions because in some important sense, the explanatory buck stops with us, instead of being traceable back to ever-earlier causes. We are, or seem to be, the real origin of our own intended actions. If this na´ve view is correct, it is agency, the ability to perform this trick of intentional action, which distinguishes us from mere objects, and makes a person a person.

There is certainly an air of mystery about this apparently unique human power of initiation; nevertheless, I think the 'na´ve' view is basically correct. 

Two formidable questions


Anyone who wishes to defend the "na´ve" view set out above must address two formidable questions, easily recognisable as versions of classic philosophical problems. The first is: how can intentional action interfere with the flow of cause and effect? Surely the laws of physics ultimately dictate the footballer's decision just as much as they determine the movement of the ball (albeit through a much more complex mechanism)? If people are the ultimate source of their own actions, we should be unable to predict those actions without taking into account their desires, intentions, and so on. But science seems to offer us a complete account of the world which gives us the power, in principle, to predict any event through the operation of physical cause and effect without any need to take account of "mentalistic" entities.

The second formidable question is; what exactly does "intentional" mean anyway? Intentions, in the ordinary sense of the word, are part of the wider phenomenon of intentionality in the special philosophical sense of "aboutness". Intentions, like beliefs and desires, are not just self-contained mental or emotional states: they are about something; intentions, in particular, are about the intended act or goal. The question of exactly what it means to be "about" something, the way "aboutness" works, and the mysterious powers which it seems to have, remain controversial and in many respects profoundly unclear. To define people in terms of intentions is therefore, on the face of it, merely to replace one mystery with another.


Surprisingly enough, the answer to the first formidable question lies in the second. The problem of personhood, and with it the issues of responsibility and free will, turn out to be members of the family of problems raised by intentionality. How so? One of the more mysterious aspects of intentionality is the ability of intentions (along with beliefs, desires, and so on) to address the future. The footballer's kick, for example, occurs because of a possible goal which hasn't happened yet (and indeed, may never happen). If intentions really allow future possibilities to influence present events in this way, they clearly do break the normal causal series: it isn't that people are somehow uncaused causes, but rather that the causes of intentional behaviour lie in the future rather than the past. It's the intentionality of intentions, and specifically their future-directedness, which runs counter to normal causality.

The dilemma created by this answer, however, seems as bad as the original problem. On the one hand, it seems clear that we do routinely think about things that haven't happened yet, and base our actions on those thoughts. It surely can't be true, however, that the future causes our thoughts in any direct physical sense. For one thing, if that were so, our beliefs about the future would be as definite and particular as our memories of the past: for another, we can think about things that don't exist and never could just as easily as we can think about the genuine future. In fact, it seems clear that there is no fundamental difference between these two cases; when we think about the future it is an imagined future we address.

A natural response to the impasse is therefore to say that our intentional acts are caused, not by the future itself, but by mental images or representations of it. But this merely pushes the problem back a step. Aren't images caused by the thing they represent? It is the shape and colour of the object which determine the shape and colour of its picture. If our mental representations are caused by possible future events, we are back with the original problem of how something in the future can affect events now. If, on the other hand, the images are not caused by the events they represent, how do they come to have anything to do with them, and in what sense can they be representations?



There is one relatively unproblematic way in which our behaviour can address future events. When I see an apple, I reach out to grasp it: the reaching is a response to the present sight of the apple but addresses, and is motivated by, the future situation in which my hand will be near enough to grasp it. Without the prospect of my grasping the apple a moment or so later, reaching towards it would be pointless. How does that future possibility influence my present behaviour? My action doesn't have to be inspired by a mysterious prevision of the apple-grasping future: it is simply that reaching and grasping present themselves to me as a single action, a natural response to the sight of an apple, and one which happens to extend over a short period of time.

What is happening here is that I have, implicitly at least, recognised as a single entity the behaviour made up of the separate acts of reaching and grasping. This ability to recognise a larger entity, only part of which is currently accessible or perceptible, applies to objects as well as actions. I may, for example, pull down a branch from the mass of foliage above me in the expectation of being able to pick a hidden apple from it. I recognise the visible part of a branch, recognise it as a component part of a larger, apple-bearing branch, and recognise that an apple is another part. By recognising the over-arching entity "apple-bearing branch", I can produce behaviour inspired by the apple even though the apple itself is at present invisible and out of reach. It may turn out, of course, that in fact there is no apple on this particular branch. In that case, my behaviour addresses something that doesn't exist. If I carelessly pick on a pear tree, it may even address an absurd entity: the apple which grows on pear trees.

This seems, on a small scale at least, to be the kind of process we need in order to explain how intentional behaviour comes to address the future; and it requires no magic. The process of recognition is highly complex in practice, but in principle it comes down to the chiming of patterns in the mind with those in the world, and with each other. All that needs to happen in the example above is that part of one of the patterns evoked in my mind by seeing the branch is sufficiently similar to part or all of the pattern which would have been evoked by direct sight of the apple to cause me to behave as if the apple were already visible at the end of the branch; that is, to pull down the branch so that I can reach it.


Once we have noticed the possibility of this kind of "extended recognition"- recognising wholes from parts and parts from wholes - it becomes clear it has considerable potential. For one thing, it provides us with an open-ended way of linking perceptions and actions into complex structures. By recognising "over-arching entities" like the apple-tree branch we can jump, via overlapping patterns of activity in the brain, from something we can see to something we can't; from there to other unseen objects, and so on. In addition, recognising the relevant over-arching entity helps us understand the nature and likely behaviour of the perceived object. In fact, I believe the recognition of over-arching entities is the mechanism which underlies such apparently fundamental mental processes as logic and the association of ideas; it is the "just seeing" which provides the bedrock of all forms of understanding. *

These grand claims may seem out of proportion with the rather modest examples considered so far, but the idea of extended recognition opens the way to further progress.

Point and purpose home

The "over-arching entities" involved in different cases of extended recognition may be obscure, or they may virtually force themselves on our attention. "Teleological" qualities generally exemplify the latter category. The providential ways in which many animals behave (building nests or storing food for the winter), and the cogent and effective structure of all living things strongly evoke the processes of which they are part. They seem undeniably to have a purpose in almost the same way as tools and machines consciously designed for a particular task. In such cases, we can hardly help but recognise the over-arching entity whose parts include, on the one hand, the pattern of behaviour, or the structure of the organism, and on the other the occurrence of a particular result.

In discussing these teleological properties, we normally fall back on the language of intentions even when we know it isn't strictly appropriate. If asked why birds have wings, we may say that they were meant for flying with, just as we would about the wings of a plane, though only in the latter case were the wings the result of a conscious design. It is helpful to have a vocabulary which allows us to describe the kind of apparent purpose possessed by living things without implying actual intention, and here I will use the word "point" in two related ways: the point of a spade is that it helps dig holes: equally, a spade points to the digging of holes. The point of the eagle's wings is that they allow it to fly, though they were not intended or meant to lift it into the air, whereas flight is both the point and the intention in the case of the wings of a jumbo jet.

Seeing the point of things, in this sense, is clearly an example of the kind of extended recognition discussed above. We can recognise wings as part of the process of flying, and a spade not only as a spade, but also as part of the larger entity of hole-digging. Pointing, in the ordinary sense of pointing with a finger, is a particular case of the wider phenomenon, namely the case in which the larger entity is a simple line through space. We recognise the finger, recognise the extended line of which it is part, and recognise the object which occupies a more distant part of the same line.


 People's ability to recognise the point of the behaviour patterns of animals, and more particularly, the behaviour patterns of other people, opens the way to an ability to think about things. H.P. Grice perceptively pointed out that an intention to communicate is successful if the target recognises the intention. If A correctly recognises that B is trying, with discreet hand-signals, to tell him that his flies are undone, then ipso facto B has succeeded, because B now knows about his flies. In recognising the content of the intention, A necessarily recognises the content of the intended message.

Short of telepathy, however, intentions cannot be directly recognised; they can only be inferred from observable behaviour. It isn't, in fact, the intention that needs to be recognised so much as the point of the communicative behaviour. The motion of the hand points to zipping up a trouser fly; the process of zipping up a fly points, among other things, to flies. If we recognise that a particular piece of behaviour is well adapted to making us think of trousers, then ipso facto we think of trousers, whether or not that was the other party's aim.

From point to meaning home
early adopter

'Early adopter' problems

It must surely be in this kind of process that the beginnings of real communication and intentionality are to be found. We can imagine, with no claim whatever to historical or anthropological accuracy, an ascent into meaningfulness by stages.

Suppose I am a languageless anthropoid living with others of my kind in an ancient village. Fellow villagers who see me excavating a waterhole may see that the digging points to accessible water, a desirable object which prompts digging behaviour in them too, so that they come over and help. I can now recognise another point to digging - as well as getting water, it makes other people come and help dig. In future I may start digging in places where I couldn't finish the job myself, because I foresee the arrival of help.

Recognising the need for help even before setting out, I may be prompted to perform some digging behaviour (lunging towards the ground with a pick, giving the tennis-player's grunt normally emitted when the pick strikes home) in my home village where others can see it. The other villagers can recognise the mime as part of the action of digging, which in turn is part of the action of getting water, which in its later stages features the desirable element of accessible water. In short, the mime points to water: so some of them respond by coming along to get water, too. After many such occasions, familiarity means that the grunt alone, in the right circumstances, comes to be a sufficiently recognisable cue to recruit help.


I'm not alone in seeking help this way, and after a while, a number of recruiting signs or sounds are in use. When a villager steps outside in the morning, there may be a digger, a fruit-picker, and a hunter all performing the characteristic gestures and noises that enlist assistance for their own chosen pursuit. The villager is faced with a set of simultaneous stimuli, and responds to the one with the strongest current appeal. Using the ability to see each potential expedition as a whole entity and recognise later stages within it, the villager may spot flaws in some proposals: recognition of the possibility of fishing, say, extends to recognition of the annoyingly ripped net still lurking in the back of the hut.


When I wake in the morning, I recognise the beginning of the time of day when various activities are canvassed. From that first part, I can recognise the overall process of canvassing. By recognising in sequence each of the canvassing efforts encountered in my normal walk through the crowd, I can 'review' the options usually on offer without actually going outside.


One day, while resting in my hut, an errant smell from outside causes me to recognise the possibility of apples. I may recognise these possible apples as coming from a group of thriving trees, which might belong to a particular landscape, the bare earth of which resembles a real plot I know; the growing of apple trees on which would be the end of a recognisable process involving seeds which got there through a placement process, part of which is my right hand and the rest of the body which goes along with it... Recognising my way up and down the long chain of entities, I can work back from the imagined future orchard to what I can do now; and forward again to search for problems, until the plan (and surely we can call it that) is distinctly formulated.


We can now see how intentional, deliberate actions arise. When I make plans through processes like the ones described above, the cause of my actions will involve my having recognised probable consequences of the options entertained. Actions which result from the recognition of their probable future consequences surely deserve to be considered intentional. In such cases, it can fairly be said that I "knew what I was doing".

There is scope for actions to be premeditated in some stronger senses, too. Since starting a review of my options is itself among the actions which can be inspired by the ongoing mental planning process, my behaviour may result from a second- or higher-order review: one in which I chose when and whether to choose and what to choose between.

Besides "knowing what I was doing", moreover, I may also have asked or told myself what I was planning, because as well as talking to my fellow anthropoids about my water-seeking projects, I can surely talk to myself, too, and even do so silently. Reviewing the different sequences of grunts which form part of the recruitment process in my village, I may work step by step through rival sequences (thinking, we might surely claim, about what to say); and the grunt sequences (or words), may be reviewed not as candidates for utterance, but for their direct effect on me; a process which amounts to explicit thinking in words - talking to myself in my head - rather than wordless consideration of alternatives.

It is impossible to resist the conclusion that internal processes of this kind, though essentially only complex series of connected acts of recognition, amount to thinking about things as well as forming intentions. It follows naturally that such processes also constitute consciousness or, for those who subscribe to the idea that phenomenal and 'access' consciousness could be different, at least the latter variety - the kind of consciousness which is causally relevant to our actions.

Coming back at last to our starting point, since these processes are the ones from which intentional acts arise, and since we began by defining a person as a source of intentions, it follows, plausibly I hope, that wherever such processes operate (typically, and perhaps invariably, in a human being) there is indeed a person. This person, the common origin of a particular set of intentions, is itself an over-arching entity, recognition of which, in the appropriate case, amounts to self-awareness.



Although there is nothing in the processes described above which contradicts normal physics or the standard materialist understanding of the world, this account also provides an indication that intentional actions nevertheless remain beyond the reach of calculated prediction.

At a "mentalistic" level, the nature of the problem is clear. In order to work out the outcome of any process, mental or otherwise, we need to know which causal factors to take into account. In predicting the path of a ball across a football pitch, we have to exclude, or take into account, the influence of the wind, for example. Now our intentional behaviour is caused by what we think about, and very small differences in our thoughts can lead to very large differences in our actions. But we can think about anything . The list of potential causal factors for mental processes is, accordingly, indefinitely long, and includes imaginary and absurd entities. At any moment, we may think thoughts no one has ever thought before, and any one of our implicit beliefs and desires may become relevant. As a result the outcome of processes which depend on our beliefs are not calculable.

This strange fact belongs to a family of oddities about meaning and mental processes. A notable member of this family would (in my view at least) be the indeterminacy of translation pointed out by Quine, for example. Another is the mysterious ability we all have to hold an infinite number of beliefs at once, at least in an implicit form, and yet to pull out the relevant ones effortlessly and instantly. The appearance of mystery in all these cases arises from the same fundamental misapprehension; as we have seen, our minds don't encode our beliefs in lists of symbols; they simply contain patterns which point.


There is a curious double ambiguity inherent in pointing. Every pointer points to an infinite number of things. When I point a finger I may mean to indicate the clock on the mantelpiece, but I also point to the wall, the people next door, the small volume of air immediately in front of my finger, and so on. Equally, objects used for indication may contain an infinite number of pointers. How many cross-country routes does an ordinary map show? Unless we impose some limitation on such factors as the number of start and end points to be considered, and the length of acceptable routes, there are as many as you care to pick out; in this sense, the finite map provides an infinite amount of information. In a similar way there is no end to the list of distinct patterns which can occur in a brain, and no end to the list of implicit contents of any particular mind.


Our thinking is currently so coloured by the seductive analogy between brains and computers that this is hard to accept, but it must be remembered that the human brain is not a discrete-state machine. Computers are constituted that way precisely in order to guarantee that their behaviour is predictable with certainty (or in practice, with a very high degree of probability), and can be specified by the designer or programmer. Human beings, on the other hand, are not artefacts; they were never designed, or specified; they are the products of intentionless evolution, and no one ever declared that they should have only a finite number of valid states. In fact, just as the map shows an infinite number of routes, our brains have an infinite number of valid states.

It's clear, back on the mentalistic level, that if we had been the products of design rather than evolution, with our behaviour carefully specified in advance by a designer or programmer, we really could not be free, responsible, or indeed, people. The causal relations which operate within a brain are determined by the detailed physical constitution of that particular brain, but the causal relationships which operate within a computer are independent of its physical details, barring accidents.

Curiously enough, then,  it turns out that those who see people as programs or data have something in common with those who would see them as immaterial spirits. The computationalists, at the end of the day, are not materialistic enough: they want people to be Platonic abstractions rather than the one-off physical objects they really are...


In the case of footballs and other simple physical phenomena, the potential interactions and the forces involved in any event can be listed and approximated accurately enough to allow predictions with a negligible margin of error. But the reliance of conscious processes on the pointing qualities of mental patterns, means that unforeseen patterns cannot be excluded, nor their potential impact on the course of events kept within any limits. So the ultimate explanation of intentional behaviour really does lie in the incalculable patterns which embody our desires, beliefs, and intentions.


* I certainly find this perspective intuitively appealing. When I think about a problem it feels as if I am contemplating it in slightly different ways and waiting to recognise a useful aspect. It feels much more like that, at least, than the application of an unconscious algorithm. But in addition, this view seems to cast a helpful light on other problems about mental activity. It provides, for example, an illuminating way of looking at induction, causality and counterfactuals. Humean problems in these areas arise because we look for formal reasoning to justify a belief, or prediction, whereas often it is simply the recognition of consistent over-arching entities (the momentum which persists through the collision of billiard balls, for example) which is at work. It is hard to shake off the assumption that our ability to recognise things needs to be underpinned by logical procedures, but in fact the reverse is true: logic itself depends on our ability to recognise consistent patterns of truth values.Back to main text