2. Minds and The Mind

For Anscombe, the question ‘What is intention?’ is not a question about the states and properties of an individual or – worse – the inner workings of an individual mind. As she puts it in the opening section of her book Intention – it is rather a question about ‘the character of the concept of intention’.Footnote ² A question about the ‘character of a concept’ is not a question about the character of mental representation but a question about the sorts of patterns and abilities that make up a human life shaped by that concept, and about the social, institutional and empirical background that give those patterns their meaning and those abilities their point.Footnote ³ I will say more about that in a moment. To answer a question about the character of a concept, one must attend to the context of human life within which we employ that concept, and to the sorts patterns of explanation and expectation that an application of the concept requires, imposes and supports. What is presupposed in the application of the concept of intention? – about an individual's past and present, about the society in which she lives, and about what will happen in the future?

I said that Anscombe's ability to keep her focus on this background was staggering because the temptation to turn one's attention toward an individual's mind – often one's own – when one's topic is human psychology is almost irresistible. But what Anscombe and Wittgenstein both recognised is that a version of Frege's anti-psychologist argument in logic applies here too.Footnote ⁴ As Frege de-psychologised logic, Anscombe and Wittgenstein de-psychologised psychology. Let me pause on that idea and use it to make clearer the object of Anscombe's – and Wittgenstein's – attention.

Frege said that his topic was ‘the mind, not minds’.Footnote ⁵ He studied The Mind by studying and revealing the laws of thought. He insisted that the laws of thought are not like laws of nature; they are not ‘psychological laws in accordance with which [mental process] takes place’.Footnote ⁶ So, for Frege, when one is studying the laws of thought one is not interested in how particular humans in fact think and reason – it is not relevant, for example, that most of us will infer p from p&q nor that attempting to entertain a contradiction causes psychological discomfort in all reasonable people. Nor is he concerned with the psychological processes – be they in the mind or in the brain, conscious or sub-conscious, personal or sub-personal – that attend thinking and reasoning and that one might discover by introspection or by empirical study. It is of no interest to Frege's study of the laws of thought that one does or does not feel a transition of thought in moving from antecedent to consequent to conclusion, nor that brain process XYZ is always found to attend an act of judgment. Rather, to describe the laws of thought is to describe a formal order in which anything, if it is a thought, will participate. To be a thought is to stand in logical relations to other thoughts, and the character of those logical relations can be described. To describe the structure of that logical space is to give the essence of thought. Anything that does not fit within that structure is not a thought, whatever an individual human being feels or experiences or determines by introspection or brain scans. This is what Frege means when he says he is undertaking a study of The Mind not of minds.

The Fregean idea of a law of thought, and the identification of thought's essence with the structure of logical space, take on a distinctive form in Anscombe's work, as they did in the later Wittgenstein's. This is because where Frege views language – the medium of thought – as an abstract symbolism, Anscombe and the Wittgenstein of the Philosophical Investigations picture language as a communicative tool, whose logic is shaped by and shapes the lives and actions of the humans who fashioned that tool for their own employment and to serve their collective ends. This organic and concrete conception of language means that the task of studying The Mind not minds requires one to attend to more than laws of logic; one must take as one's object human life shaped by concepts.Footnote ⁷ For Frege, to describe the structure of logical space is to give the essence of thought; for Anscombe and Wittgenstein, to describe the structure of human life (the human form of life) is to give the essence of thought and action. Frege's limits of thought – a boundary drawn from the inside when the structure of logical space is mapped – now become the limits of the intelligibility of a human life.

Because Anscombe takes as the object of enquiry the whole ambit of human affairs, she sees most other philosophers’ attempts to address the topic of human action as mistaken not merely because they generate false theories but because they miss their subject matter entirely. They are studying minds but the proper topic for philosophy of action is Mind. She says of Davidson that his project of action explanation fails not because he cannot find a way to rule out deviant causal chains – though that is a failure too – but because he takes his object of enquiry, human action, as ‘already given’.Footnote ⁸ He takes it, that is, that he can identify and individuate human actions, and can have a grasp of the difference between a human action and an involuntary gesture, or between a human carrying potatoes and a potato-carrying automaton, prior to and independently of having a grasp on what it is to give a reason for action.Footnote ⁹ But these entities and distinctions are not things to which he is entitled. His error is akin to supposing that one can identify an individual mental-item as a thought prior to identifying it as something that stands in inferential relations to other thoughts. For these physical happenings to be a human action (and not merely the movement of potatoes from one place to another by means of some organic matter) is for the happening to be characterisable in a way that locates it within a framework of reason, prediction, and action – in a way, that is, that locates it within the ongoing bustle of human affairs. So, in taking pre-philosophical intuition to entitle him to the category of human action, Davidson not only renders his investigation illegitimate, he – worse – precisely excludes the sort of enquiry that might yield a proper understanding of his topic. He excludes the enquiry that Anscombe undertakes in Intention: an enquiry into the formal order in which anything, if it is a human action, will participate, and which characterises the field of human affairs. If we return to the form of investigation of human action that I contrasted with Anscombe's above – the one that takes as its object of enquiry the individual human subject, her psychological states and processes and her intentional actions – we can add this: the individual ‘agent’ is also not something that can be taken as ‘given’ in the way that the Big Head investigation supposes.Footnote ¹⁰ This is one of the deep insights of Anscombe's ‘The First Person’.Footnote ¹¹

3. Return to action

To return to my original deliberations as a grocer. We can now frame a new question: can we describe the background against which this sort of thinking and acting is possible, and by doing so can we start to reveal what is essential to practical thought? By ‘possible’ I do not of course mean physically or psychologically possible – though reflecting on the empirical conditions for the possibility of an instance of practical thought may help us to understand something about its essential character. By ‘possible’ I mean rather: what context and background are required for my thinking to so much as to be the kind of thinking that it is? This new question rejects the pressure to locate what is essential to this little scene inside my head and eschews the search for a connection between thought and action, inner and outer.

The answer begins, as Anscombe's Intention begins, with a grounding observation designed to situation our vignette in the context of an ongoing human history into which I – one grocer among many, past, present and future – will, for a short time, participate. We have grown to the age of reason in a shared world.Footnote ¹² It is the fact that I have grown to the age of reason in this world, a world shaped by my fellow humans, that I am able both to pose this question – ‘How shall I supply you with potatoes?’ – and to draw on the tools of the language I have learnt to answer the practical problem I set myself. As Anscombe points out in Intention, it is also this fact that would allow me to say ‘straight off’ that you were supplying potatoes to me if I observed you carrying potatoes to my house and leaving them there.Footnote ¹³

It is important to see that it is the shared world, the history of mankind (as Wittgenstein puts it), that carries the burden of holding together the conceptual relations that I draw upon when I pose to myself the question: what should I do, given that I want to supply you with potatoes? The connection between carrying potatoes to a place and leaving them there, and supplying potatoes, is located in the meaning of those descriptions, in the language that I use and we share, and in the shared world we inhabit.

The capacities that my mastery of our language encodes were shaped and reshaped by my kind: human animals with needs and interests that are characteristic of our species. Those needs and interests show up in the working of our language. ‘Supplying’ is not merely leaving stuff somewhere for another to find. A community of people who carried potatoes to houses and left them there, but who violently despised potatoes or had no institutions of commerce, exchange, nor any practice of the mutual provision of necessities, would not thereby be supplying potatoes. This absence of background would show up in the wider context – in the way that individuals reacted to the potatoes that appeared on their doorsteps or to those who delivered them or in the wider rituals in which this behaviour was embedded. ‘Supplying’ is a way in which we make resources available to each other, in order that we may each have that which we need to live or want for pleasure. So, the description ‘being supplied with X’ is completed with an ‘X’ that is wanted, needed or requested – only in very special circumstances would it be appropriate to say that I supplied you with a toenail clipping or a black eye. A cat does not supply her owner with a mouse and nor does a child supply his mother with her slippers because to act under the description ‘supplying’ is to stand in a relation to another that involves the sort of complexity and mutual recognition that animal-to-human or child-to-adult relations necessarily lack. This is not, of course, to say that a cat or a child cannot bring or even give a mouse or a pair of slippers.Footnote ¹⁴

To learn the meaning of ‘supply’ – to learn to recognise when the concept applies and to be able to follow the order ‘Supply X!’ – is to learn, among other things, similarities and distinctions between supplying X and other activities like giving X, delivering X, providing X, making X available, and so forth.Footnote ¹⁵ Part of what someone who learns the meaning of ‘supply’ must know is what matters to humans – what sort of things they need and want. So a person who can employ this concept in practical thought and in judgment is someone who has at least a partial grasp on what sort of things are good or desirable for humans.

What one learns when one learns a concept, acquires a capacity, has of necessity a sort of incompleteness – this is an insight pursued to startling effect by Wittgenstein.Footnote ¹⁶ We acquire concepts, capacities, rules, by being shown examples. After a number of examples we confront: ‘and so on’. This ‘and so on’ is not shorthand for a finite list that is occluded for pragmatic or epistemic reasons. Rather, it marks gesture toward a standard of ‘sameness’ or ‘similarity’ that a normal learner will catch on to, but which cannot be articulated independently of the concept itself. This ‘and so on’ is tied to learning situations, so a conception of ordinary procedure, and relatedly of what is abnormal and exceptional is built into the grasp of the rule. But talk of what is ‘exceptional’ needs care. It would be an exceptional if I suddenly discovered I had won the lottery or if the roof fell in or if I received a phone call from the Queen. But none of these exceptional circumstances touches the relationship between the bringing and leaving of the potatoes and the supplying of the potatoes. What counts as ‘exceptional circumstances’ is internal to the meaning and significance of the description involved. It is here that disagreements can occur about what a person really did, and about whether the circumstances in which she acted are relevant not just our evaluation of her action, but to the descriptions under which it fell.

The connection between the description ‘supplying you with potatoes’ and ‘carrying potatoes and leaving them at your house’ cannot, then, be synonymity because exceptional circumstances can always mean that an activity that falls under the former description does not fall under the latter. Yet, the connection between the descriptions is not merely constant conjunction. Rather, carrying potatoes to a house and leaving them there is an exemplary case of supplying potatoes – it is one of the many cases of human action we might point to if we wanted to give the meaning of ‘supplying’ or to indicate what is normal and ordinary with respect to it. Even in cases of potato-carrying and leaving where the description ‘supplying potatoes’ is false, we might say: it has application here. The question: ‘Is she supplying potatoes?’ is salient and relevant as the question ‘Is she organising a wedding?’ is not. It is here that concepts like pretence, fraud and sham have their home.

We can represent the relation between ‘carrying and leaving potatoes’ and ‘supplying potatoes’ using a hypothetical: If a then b. Such hypotheticals encode our shared knowledge of what is and what can be done, and I, a human grown to the age of reason in this shared world, can make use of that knowledge when I use them to reason about how to do what I want to do.Footnote ¹⁷

Let me pause to say something about the logical form of these hypotheticals. The connection between the descriptions is not synonymity but neither is it constant conjunction. The hypotheticals that I make use of in a piece of practical reasoning cannot have the logical form of existential generalisations. The relation is one that holds between descriptions of human action and it is a formal relation that tells us what those actions are and what are those actions. We must not try to represent the knowledge that these hypotheticals encode in the form:

Such existential generalisations are false, which is bad enough, but to attempt to force our hypotheticals into this form is both a symptom of and an encouragement toward the sort of philosophical treatment of action that involves having one's attention in the wrong place. As a symptom, it reflects a misconstrual of the nature of the relata that the hypothetical concerns. In the quantified version the hypothetical statement ranges over entities that are specified prior to the application of the concepts that are our topic. It is as if you could pick out some event e prior to the specification of it under an action description, e.g., carrying potatoes. But this is precisely wrong and is a version of the error Anscombe finds in Davidson. This is to turn one's philosophical attention away from The Mind and onto minds.

Before I move on, I want to make one final comment about the institutions and social and cultural practices that provide the background to many of our concepts. This is an aside here, but it makes a point I want to have in place later, when we come to consider the subject-matter of ethics. What I want to add is this: just as many of our concepts have their point in the context of our institutions and social and cultural practices, so too our institutions and social and cultural practices have their point in the context of our needs and desires, and take their shape from the material conditions of our lives. These are things that in some aspects remain constant. Our instinct for the so-called ‘four F's (flight, fight, food and reproduction) is characteristic of our species. The atmospheric pressure, the boiling point of water and the weight of this pen are the same now as in the past, and the same here as in America. But our lives – human lives – and the conditions in which they are lived do change, sometimes dramatically, not least because as our use of tools – and in particular, the tools of language – become ever more sophisticated, our interests and habitat can be altered and changed. Shifts across time, culture and geography in what we need and want and in the material conditions of our lives are to be expected for language-using creatures like us. This is one reason why – as Anscombe, Mary Midgley, Iris Murdoch, Alasdair MacIntyre and Bernard Williams each in different ways warn – once useful concepts can change from genuine capacities into dysfunctional habits or pathologies and once necessary institutions can become structures of injustice and oppression.

4. From psychology to ethics

What I have been saying is, as you may have spotted, an exploration and elaboration of Anscombe's discussion of ‘brute facts’ in her rightly celebrated paper ‘Modern Moral Philosophy’.Footnote ¹⁸ You will remember that in that paper she says: ‘It is not profitable at present for us to do moral philosophy. It should be laid aside at any rate until we have an adequate philosophy of psychology, in which we are conspicuously lacking’.Footnote ¹⁹ It is in consideration of this that Anscombe appeals to the relation of ‘brute-relative-to’. In the remainder of this paper I want to start to think though how the reorientation I have described in philosophy of action provides the sort of philosophy of psychology that can provide a starting point for ethics.

Anscombe introduces the topic of brute facts in relation to Hume's puzzle.Footnote ²⁰ She says that the troubling transition from ‘is’ to ‘ought’ makes an interesting comparison with the transition from ‘is’ to ‘owes’. Hume's worry about the legitimacy of the former transition, she says, points us toward a kind of relation that is of significance to ethics. The relation is between levels of description of human action. Hume's worry is how a description of human affairs that employs only non-normative concepts can stand in this relation to a description that employs normative concepts, like ‘ought’, ‘should’ and ‘must’. But Anscombe observes that the relation that Hume finds unintelligible is one that exists between descriptions of human affairs at many levels, not just between descriptions that employ only non-normative concepts and those that involve ‘ought’, ‘must’, ‘should’, and so forth. How does the fact that I carried potatoes to your house and left them there relate to the description ‘supplying you with potatoes’? She writes:

What Hume puts his finger on is this: if we attempt to do moral philosophy without having understood this kind of relation we will be left mystified by the transition from ‘is’ to ‘ought’. Anscombe's insight is that – at least part of – Hume's worry is not about the transition from ‘is’ to ‘ought’ but about the transition between levels of description of human action.

Returning to my moonlight job as a grocer. The fact that I carry potatoes to your house and leave them there is brute-relative-to the description ‘supplying you with potatoes’. And the fact that I supply you with potatoes you ordered is brute-relative to the description ‘You owe me for the potatoes’. So, here we have a fact is brute-relative-to a description that seems to be of interest to ethics.

Anscombe's reorientation in philosophy of action has given us the tools to articulate ‘merely “factual”’ descriptions that employ normative concepts like ought and should, must.Footnote ²² The hope was that this, in turn, would give the ethical a foothold in the human. It can seem, however, that this hope is short-lived because though we have introduced normative concepts it is not clear that the ‘merely “factual”’ ‘ought’, ‘should’ and ‘must’ can do the work that we want our normative concepts to do in ethics. First, the norms we have in play here do not transcend the institutions or practices in which they have their home. As Foot puts it: ‘owes’ is the sort of ‘evaluative’ concept, that ‘belong[s] within an institution’, and this means that if the institutions is a bad institution – an unjust one, we might say – then it may not follow that debts accrued in the context of that institution are debts one has a duty to pay – even though the description of the debt generates an ‘ought’ within the institutional context.Footnote ²³ This recalls an observation we made earlier: institutions and practices that were once useful or sound can become instruments of oppression and injustice if facts on the ground change. Given this, norms that are internal to those institutions and practices cannot, it seems, be verdictive in the way that the norms of ethics seem to be.

The second difficulty is that it seems that the ‘merely “factual”’ conceptions we have so far would provide me with reasons for action only if I there is something that I want that makes relevant to me those descriptions and gives me a reason to consider how I might make them true through my activity. Why should I use the hypothetical – if I hand over this money I will pay what I owe – if I am not interested in whether or not I pay what I owe? Why should I use it any more than I would use my knowledge – if I carry these potatoes to NN's house I will supply her with potatoes – when I have no interest or desire to supply NN with potatoes. But now we seem to have lost the foothold that ethics had. If these ‘merely “factual”’ descriptions have their point only within the context of man-made institutions – institutions that can be unjust, bad, or useless – and if the opportunity to make these descriptions true or false neither gives me a reason to act nor to refrain from acting – it seems we still have work to do before the philosophy of psychology that Anscombe gives us connects with the subject-matter of moral philosophy.

I wrote above of Anscombe's first warning at the beginning of ‘Modern Moral Philosophy’: ‘It is not profitable at present for us to do moral philosophy. It should be laid aside at any rate until we have an adequate philosophy of psychology, in which we are conspicuously lacking’. Now is the time to remind ourselves of her second warning: ‘the concepts of obligation and duty – moral obligation and moral duty, that is to say – and of what is morally right and wrong, and of the moral sense of “ought”, ought to be jettisoned if it is psychologically possible’.Footnote ²⁴ Now, the thought I want to explore in this final part of this paper is the following: that our feeling that we do not have enough in the ‘merely “factual”’ descriptions we already have is a symptom of the psychological hold that the idea of a ‘moral sense of “ought”’ has over us. We have enough already, I want tentatively to suggest, to ‘do moral philosophy’. Though, of course, what ‘doing moral philosophy’ looks like now is going to be rather different to what we have been doing – or trying to do – before.

Here is the idea. Despite appearances, the framework we have – one which describes Mind not minds – is sufficient to generate the sorts of generic ‘oughts’ that are of interest to ethics. By ‘generic’ I mean: they are not relative to the particular desires or plans of the individual whose actions they concern. My idea – which I hope is close to Anscombe's idea – is that these generic ‘oughts’ are enough to characterise the subject-matter of ethics without us needing to appeal to a quasi-legal ‘Moral Ought’. Let me try to bring this out.

Suppose someone asks, ‘Why must I pay what I owe?’ and consider two possible replies:

1. (1) You must pay what you owe because if you don't the grocer will never supply you with potatoes again

2. (2) You must pay what you owe because it would be unjust not to do so

These two replies look very similar, but Anscombe observes that despite their shared sentential structure, they have quite different logical forms. The first reply gives the individual who asks it a reason for acting. This reason will only interest the individual if the description of what will happen as a consequence of her failing to pay is one that she is interested in avoiding. Replies of this form offer a reason for acting that is specific to the individual who asks her question ‘Why should I do it?’. As we might put it, the ‘I’ in this question and the ‘You’ in the reply introduce the particular individual who is deliberating about what to do.

Now contrast the second reply: You must pay what you owe because it would be unjust not to do so. In this answer the ‘I’ in the question is treated by the responder as generic, akin to ‘one’ or ‘we’. ‘Why should one pay?’ The answer contains the generic ‘you’: ‘One shouldn't because it is unjust’. The phrase that follows the ‘because’ does not give an independent reason that is sensitive to the individual's idiosyncratic desires and beliefs, but rather characterises the action described in the first part of the statement as one that falls under a further description containing a vice term. The fact that the act would be unjust is not given as a further independent consideration, a reason for action, but is put forward as a characterisation of the action itself. The answer reminds the asker of a connection in the formal order that characterises human thought and action. Anscombe reserves the label ‘logos’ for an answer to the question ‘Why?’ that has this sort of generality.Footnote ²⁵

Recall Frege's project to describe the formal order in which anything, if it is a thought, will participate. In so doing, we noted, he traced from the inside the limits of intelligibility in thought. To answer a ‘Why?’ question by giving a logos is to draw attention to action's location at the limits of intelligibility in human life. This sounds rather grand, but it need not be; it is commonplace to say ‘I can't make sense of what she did’, ‘I don't understand his life’, ‘She seemed hardly human’. To employ the terminology that has been our topic today: the answer provides a reminder that the proposed action would be brute-relative-to a description of an action kind that we do not go in for: ‘breaking a promise’, ‘contravening a rule’, ‘infringing a right’, ‘cheating’, ‘steeling’, ‘bilking’. These descriptions are ones whose meaning and point in our language is to mark out classes of activity that we, collectively in our language and our lives, recognise as of serious importance to human life going well – or badly – and they encode our shared knowledge of that fact. Here is an echo of Mary Midgley's brilliant insight: ‘moral’ is the superlative of serious.Footnote ²⁶

The idea, then, is that while an answer to the question ‘Why?’ which gives a reason for action can intelligibly be answered ‘I don't care about that!’, when an answer that gives a logos attracts this response the intelligibility of that answer is not to be taken for granted. When the ‘you’ in ‘You mustn't because …’ is generic, rather than individual, it makes an appeal to a standard of evaluation of the action in question that transcends the particular aims of the individual and makes reference instead to the logos of human action as such. An act of carrying can be evaluated as good or bad by reference to the individual's aims and goals in performing that act, and by reference to a standard case of carrying. Was it too slow, too ostentatious? Was the load carried to the right place, at the right time? Was it efficient and safe? But in evoking these descriptions in terms of virtue and vice – just, unfair, cowardly, and so forth – there is an appeal to an evaluation of the action by reference to a more general standard: the standard of good human action. We don't do that, remember? is the form of reply. This is where the ethical gets its foothold.

We may still raise, of course, a version of the worry that we had before, and I don't want to say that this worry can be easily dismissed. The worry from before was: ‘shoulds’ that are internal to institutions cannot be verdictive in the way that the oughts of morality are. The worry now is: ‘shoulds’ that are internal to the human form of life cannot be verdictive in the way that the oughts of morality are. What is one to say to that?

I don't quite know, but I want to try to give a sense of what is odd about that thought by recalling the depth of the connection between our form of life, the concepts that we have, and the things that we, humans, want and need. Recall that to learn the meaning of ‘supply’ – to be able to employ that concept in thought and deed – is to know something about the sorts of things that humans might need or want. Our concepts have a point for us – they reflect our interests and goals – and learning a concept is, in part, learning that point. So, the limit against which a person comes here is not a limit in an arbitrary system of rules or ‘mere’ convention but the limit of our form of life. The concepts that we have record ‘the history of mankind’ and are tools that we have fashioned to serve our collective ends, our natural instincts and desires, in response to our environment. They are also the only means we have to navigate that world because what is to participate in the history of mankind is to grow to the age of reason in a world that is shaped by those concepts. Moral pioneers, individuals who live at the edges of intelligibility, do exist. We might call them the nihilists or saints or radicals. But their ability to pioneer, and through their actions to refashion our concepts, depends on the vast majority of us sustaining, in thought, word and action, the form of life into which we were born. On the whole, in the ordinary bustle of our day-to-day lives, we do keep our promises, honour our contracts, treat each other with compassion and justice. When we do something vicious, we often seek to demonstrate that exceptional circumstances undermine the description of what we did as such. Attempts to do so are often attempts to render our actions intelligible by rejecting the description of it as unjust, unfair or cruel in favour of one that reconnects it to human virtue: I was cruel to be kind, we say (meaning: what I did does not really fall under the description ‘cruel’). If we – collectively, for a sustained period – ceased to operate broadly within the limits marked by these evaluative concepts, the result would not be a worse state of affairs but the chaos of a world in which there were no ‘states of affairs’ that we could recognise or describe, and nothing that we would recognise as human life and action.Footnote ²⁷