Hostname: page-component-7b7567c8c-kwt9d Total loading time: 0 Render date: 2026-11-13T22:43:02.865Z Has data issue: false hasContentIssue false
  • English
  • Français

Training, Transformation and Education

Published online by Cambridge University Press:  07 July 2015

David Bakhurst
David Bakhurst*
Affiliation:
Queen's University
*
bakhurst@queensu.ca
Rights & Permissions [Opens in a new window]

Abstract

In Mind and World, John McDowell concludes that human beings ‘are born mere animals’ and ‘are transformed into thinkers and intentional agents’, principally by their initiation into language. Such ‘transformational views’ of human development typically represent first-language learning as a movement from a non-rationally secured conformity with correct practice, through increasing understanding, to a state of rational mastery of correct practice. Accordingly, they tend to invoke something like Wittgenstein's concept of training to explain the first stage of this process. This essay considers the cogency of this view of learning and development. I agree with Sebastian Rödl that the idea of training (as developed, say, by Meredith Williams) is inadequate to the nature of infancy and child-parent interaction, and I draw on the work of Lev Vygotsky and Michael Tomasello to offer McDowell a richer picture, which acknowledges the child's active role in fostering the second-personal relations that underlie the possibility of language learning. Such considerations force us to revise the transformational view, but do not refute it outright as Rödl believes. I conclude by considering the relevance of McDowell's view of second nature to two striking ideas: Ian Hacking's suggestion that the development of autistic children is ‘non-Vygotskian’ and Derek Parfit's claim that persons are not human beings.

Information

Type
Papers
Information
Royal Institute of Philosophy Supplements , Volume 76: Mind, Self and Person , May 2015, pp. 301 - 327
Copyright
Copyright © The Royal Institute of Philosophy and the contributors 2015 

1. Introduction

At the conclusion of Mind and World, John McDowell writes:

Now it is not even clearly intelligible to suppose that a creature might be born at home in the space of reasons. Human beings are not: they are born mere animals, and they are transformed into thinkers and intentional agents in the course of coming to maturity.Footnote 1

According to McDowell, during their upbringing, or Bildung, human beings acquire ‘a second nature’.Footnote 2 This occurs principally through the acquisition of language, which equips us with conceptual capacities that enable a form of life in which reason is manifest. To be initiated into this form of life, McDowell attests, ‘is the same thing as acquiring a mind, the capacity to think and act intentionally’.Footnote 3

In a recent paper, Sebastian Rödl rejects this ‘transformational view’ of human development, and criticises my sympathetic treatment of it in The Formation of Reason.Footnote 4 I am drawn to the transformational view by my admiration for McDowell's philosophy and because similar conceptions figure prominently in the work of Russian thinkers on whom I have written in the past, such as Vygotsky and Ilyenkov.Footnote 5 Ilyenkov, for example, maintains that at birth the human child has a merely animal mode of existence and her mind is a merely animal mind. Only through enculturation does ‘the individual for the first time become a person, a representative of the “human race”, while before this she was merely a representative of a biological species’.Footnote 6 Rödl thinks such talk incoherent. For him, the concepts animal and person signify an individual's ‘principle of being’; these are not statuses that can be acquired or lost in the course of a creature's existence so that a living individual could start out animal and become person. Moreover, Rödl argues it is intolerably paradoxical to portray the power of autonomy – which both McDowell and I think definitive of rational being – as a product of initiation or Bildung. Autonomy must be always-already present in the child, an aspect of her first nature, if she is to become a truly self-determining being.

In this essay, I want to explore whether McDowell's version of the transformational view might be developed and defended. I shall conclude by briefly considering the light McDowell's view casts on two striking claims advanced recently by distinguished philosophers: Ian Hacking's suggestion that the development of autistic children is ‘non-Vygotskian’, and Derek Parfit's contention that persons are not human beings.

2. The Status of McDowell's Position

McDowell himself resists the idea that his position stands in any need of development. He tells us his notion of second nature is to serve as what Wittgenstein called a reminder for a particular purpose;Footnote 7 namely, that we need not gift the concept of nature to natural science, so that anything that eludes natural-scientific modes of explanation is cast as supernatural or occult. If we heed that reminder, we can say that human responsiveness to reasons is a perfectly natural phenomenon even though it requires a form of understanding – ‘space of reasons intelligibility’ – distinct from and irreducible to natural-scientific styles of explanation. The reminder, however, is designed to bring to mind plain facts about the natural course of a normal human upbringing. It is not an invitation to build a philosophical theory of second nature.Footnote 8

McDowell suggests, then, that his talk of transformation is meant as a statement of the obvious, designed merely to draw attention to the undeniable difference in capacity between a newborn and a normal mature human being. But if this is all he intends, why does he formulate his position so provocatively, saying that human beings are ‘born mere animals’ and later ‘acquire’ minds? That makes the transformation sound metaphysically significant. And there are times when McDowell affirms just that. In a sympathetic discussion of Hegel, he writes:

We should not be frightened away from holding that initiation into the right sort of communal practice makes a metaphysical difference… Responsiveness to reasons, the very idea of which is inseparable from the idea of communal practices, marks out a fully-fledged human individual as no longer a merely biological particular, but a being of a metaphysically new kind.Footnote 9

Such remarks help explain why McDowell questions the intelligibility of a creature's being born into the space of reasons, but they make it hard to renounce philosophising about second nature.

What, then, does McDowell mean to preclude when he disavows theorising second nature? His usual target is constructive philosophical theories that aspire to answer traditional ‘How is x possible?’ questions.Footnote 10 For McDowell, the proper response to such questions is to interrogate the frame of mind that makes the possibility of x seem problematic, so that the question becomes no longer urgent. This is McDowell's way with ‘How is meaning possible?’ The question looks unavoidable only because we work with a scientistic conception of objective reality that makes it hard to accommodate meaning in the world. But if we reject scientism and operate with a richer conception of nature, the place of meaning in the world is no longer a puzzle to be solved by theory. McDowell's reluctance to theorise second nature is of a piece with this. The question of how something that starts a mere animal gets to enter the space of reasons is to be exorcised, not answered.

However, it is one thing to deny we need a constructive account of how the space of reasons and its inhabitants are possible, quite another to say that invoking ‘the bare idea of Bildung’ will lay all puzzlement to rest.Footnote 11 After all, the idea of second nature is bound up with substantive claims in McDowell's thought – such as the view that its acquisition is a precondition of experience proper, rather than mere sentience. Moreover, McDowell represents the concept of second nature as philosophically fertile; for example, he claims Kant was unable to make sense of the self as ‘a bodily presence in the world’ because he lacked ‘a seriously exploitable notion of second nature’.Footnote 12 But it is hard to see how such a notion could be seriously exploited unless we can say more about second nature's acquisition. And surely much could be said that would be less ambitious than the constructive project McDowell disavows, but nonetheless philosophically interesting.Footnote 13 It therefore seems worth trying to say more about the transformational view than McDowell is willing to.

3. Training

McDowell does tell us something about the acquisition of second nature. In Mind and World, he recommends generalising Aristotle's account of the formation of moral character to ‘arrive at the notion of having one's eyes opened to reasons at large by acquiring a second nature’,Footnote 14 and he invokes Gadamerian themes about the transformative effect of initiation into language, conceived as ‘a repository of tradition, a store of historically accumulated wisdom about what is a reason for what’.Footnote 15 Both discussions are brief, but suggestive. More recently, in response to commentators clamouring for more detail, McDowell appeals to the notion of training. Second nature, he tells us, is ‘no more than the idea of a way of being…that has been acquired by something on the lines of training’.Footnote 16 I propose to consider what might be made of this notion. McDowell's appeal to training is Wittgensteinian in two ways. First, it is deflationary and quietistic; it is supposed to be consistent with his aversion to theorising second nature. Second, training is a notion that figures prominently in Wittgenstein's own remarks on learning language and mathematics, so we might suppose McDowell means us to think of the child's entering the space of reasons along lines suggested by Wittgenstein.

In the Brown Book, Wittgenstein writes, ‘I am using the word “trained” in a way strictly analogous to that in which we talk of an animal being trained to do certain things. It is done by means of example, reward, punishment, and suchlike’.Footnote 17 This resonates with McDowell's claim that human children are born ‘mere animals’. But equally, it opens McDowell to the kind of criticisms provoked by Wittgenstein's remark. It has been noted that the term Wittgenstein uses, ‘Abrichtung’, is reserved for the training of animals and is not used of children.Footnote 18 Critics are quick to seize on this and lament the ‘brutal’ character of Wittgenstein's view of teaching.Footnote 19 Is this really what McDowell has in mind by training?

We should not be too quick, however, to jump on the critics' bandwagon. ‘Abrichtung’ is not the only expression Wittgenstein uses that is translated as ‘training’. There is also ‘Unterricht’ – meaning ‘teaching’, ‘tuition’ or ‘instruction’ – which has no such unfortunate connotations.Footnote 20 Moreover, Wittgenstein sometimes uses ‘abrichten’ in tandem with terms for teaching such as ‘lehren’ or ‘beibringen’.Footnote 21 The basic idea in Wittgenstein, and in Aristotle too, is that the teaching of fundamental concepts cannot presuppose the competence it is designed to instill. So when it comes to the child's acquiring rational capacities as such, learning must take the form of a movement from non-rationally secured conformity with correct practice, through a growing awareness of correct practice, to rational mastery of the grounds of correct practice.Footnote 22 It is training that is supposed to bring the child's behaviour into line with correct practice so that it becomes, as Jose Medina puts it, second nature to her.Footnote 23 Nothing suggests this process has to be ‘brutal’.Footnote 24

It remains to be seen, however, whether the concept of training can really be of use to McDowell. Meredith Williams develops the notion in the following way. The child learning her first language is normatively inept: ‘the novice is precisely one who cannot engage in first-order normative practices’. But she has ‘certain behavioural and perceptual capacities and abilities’ that permit her to be trained to act in ways that conform to the norms governing basic language games.Footnote 25 At first, the novice ‘is not able to see the meaning or significance of the words and actions of the master’, but adults nonetheless treat her words as meaningful, and her utterances as judgements and requests, before they are such.Footnote 26 There is thus a division of labour in which adults lend meaning to the child's words. To speak meaningfully is to participate in a practice or custom, and that requires an appropriate setting, which can be in place before the child is in command of the meaning of her utterances. This makes it possible for the child to find herself operating in a domain of meaning, and gradually to adhere to its norms so that she can deploy meanings to influence others' behaviour. Thus she steadily comes to act, not just in conformity with correct practice, but out of an appreciation of what is required, and with this she enters the space of reasons.Footnote 27

In this way, training grounds the acquisition of basic linguistic concepts and the bedrock practices and judgements constitutive of the background to our lives. It secures the common ground – the fundamental agreement – that is the precondition of our normative practices. As Williams puts it, ‘What we find natural is not a matter of our being attuned to certain metaphysical properties of the objects we are confronting, but rather is a matter of how we have been trained into the social practice of certain language-games’.Footnote 28 In this, the social dimension is essential – ‘our acquisition of language is our enculturation into a society’:

Failure to grasp this essential point makes it mysterious how human beings can ever change from being creatures subject to conditioning to adults displaying their mastery of language through [e.g.] engagement in the world as scientists using the epistemic and logical apparatus of hypothesis formation and testing.Footnote 29

For Quine, it's conditioning all the way up; for Fodor, it's hypothesis-testing all the way down. The truth is that it is conditioning up to a point at which, through enculturation, the child's behaviour acquires normative standing and, to invoke a famous phrase from On Certainty, light gradually dawns across the whole.Footnote 30

The predictable objection is that no appeal to training could possibly account for the child's emerging facility with meaning. As Michael Luntley sees it, the lesson of Wittgenstein's reflections on rule-following and family resemblances is that what the novice must learn is radically underdetermined by anything her teacher can say. There is a gap that needs to be bridged. But conditioning could not do this, for it is one thing to be caused to conform to a pattern, quite another to grasp what a norm requires and act in light of that understanding.Footnote 31 Luntley concludes that we have to credit initiate learners with rational capacities from the get-go. He argues that there are empirical grounds for this attribution. But it also stands to philosophical reason, for ‘[w]ithout a responsiveness to reasons, an acknowledgement of correctness and incorrectness of application… there are no first moves with words’.Footnote 32 Indeed, only if the child is already equipped with reasoning skills is training even possible.Footnote 33 This Wittgenstein himself recognises when he represents teachers as saying things like ‘This and similar things are called “games”’.Footnote 34 Such injunctions make no sense unless the learner has the capacity to work out the relevant notion of similarity. But if she has that capacity, then conditioning is unnecessary. Williams's (Vygotskian) idea of adults ‘lending’ capacity to the child is a fantasy. No amount of behaving as if a child has some competence is going to magic the competence into existence. So, Luntley concludes, we are better off reading Wittgenstein as friendly to a rationalist approach to learning.

4. A Plague on Both Houses

There is something unsatisfying about both sides of this controversy. Luntley complains that the kind of explanation Williams offers derives something out of nothing.Footnote 35 But his conception of explanation stacks the deck in favour of his rationalist alternative. The fact is that where there is learning and development, you do get – not something out of nothing – but more out of less. So you must embrace modes of explanation that capture becoming. Otherwise you will find no alternative but to build into the child the very capacities you are trying to explain. Luntley cannot appreciate the developmental significance of the proleptic attribution to the child of abilities she does not yet possess because he favours a highly individualistic conception of competence. For him, there is a huge contrast between the child (a) making a movement or sound as a response to a stimulus and (b) intentionally making a meaningful gesture or utterance; (a) cannot become (b) merely because others treat it as such – either the child acts with understanding or she does not. If the lights are off, no initiation is going to turn them on; if the lights are on already, initiation is mere setting. (I parody, but only a little.Footnote 36 ) Here Luntley ignores the constitutive role of context. The child's actions take on meaning because of the role they are accorded in the language game. This scaffolds the child's gradual mastery of those actions as acts of meaning, so that her moves in the game can become self-conscious acts (though as such they remain dependent on context). The development of the child's activity, of her mode of engagement with the world and with other people, is quite properly portrayed as the gradual mastery of techniques of language that enable the giving and taking of reasons (and much else besides). But this cannot be appreciated if, like Luntley, we focus on how it looks from the first-personal perspective of the child, which is after all in the process of taking shape.

Yet Luntley's scepticism about training is naturally provoked by Williams's talk of understanding how a creature subject only to conditioning changes into a fully-fledged inhabitant of the space of reasons. It is hard not to hear Williams as trying to answer the kind of ‘How is x possible?’ question McDowell would prefer to exorcise. Footnote 37 And if that is her aspiration, critics will feel entitled to complain that she lacks resources to discharge the explanatory burden she has taken on. Moreover, a Wittgensteinian should aspire to bring the natural history of human beings into view by means of ‘observations which no one has doubted, but which have escaped remark only because they are always before our eyes’.Footnote 38 But much that Williams says strikes me as insensitive to the reality of infancy and the interaction of child and parent (or other primary caregivers). The condition of the initiate learner is underdescribed in a way reminiscent of the behaviourist framework that predominated in Wittgenstein's time. We are told little specific about the child's capacities. She has ‘perceptual abilities’ and ‘natural reactions’ that training must modify and supplement with acquired responses. In this, the child is represented as a passive recipient, blindly obeying the adults who ‘calibrate’ her behaviour with the community's. The assumption seems to be that parents actively train their children into the normative structures of daily life, drill them in language norms by modifying their behaviour by reward and punishment. But, as Rödl's aforementioned paper nicely brings out, parents do no such thing. They care for their children in environments that are normatively structured, full of significant objects, and where there is a lot of language about. In time, children start to understand some of that language and later they start speaking themselves. But it is not clear that their parents do anything to make this happen.Footnote 39 So parents are much less active, and pre-linguistic children much less passive, than the position implies.

Philosophical accounts of early childhood can be rather artificial. Williams writes that the assertion ‘This is a hand’ is something ‘one can readily recognize’ as the sort of thing ‘that adults say to young children’.Footnote 40 But this strikes me as false. It is very unlikely that a child, even G.E. Moore's child, would be told any such thing. Of course, children are, for example, taught word-games and songs that introduce them to language (such as ‘Head and Shoulders, Knees and Toes’), but it is one-dimensional to see this as training, and contentious to assert that such practices are essential preconditions of learning language. The fact is that children are, from day one, included in activities expressive of our shared form of life, activities that constitute the common ground from which the rule-following paradox appears an empty intellectual game. From within this activity, the meaning of ‘cat’ is not an open question to be closed by training, except in the mundane sense that, if the child's use of the word is eccentric, she can be shown how to use it properly.

5. Vygotsky

If we are to say something on McDowell's behalf about the acquisition of second nature, we must look beyond Wittgensteinian conceptions of training, and aspire to a richer appreciation of the psychological capacities of the pre-linguistic child and a more nuanced view of child-adult interaction. This Vygotsky anticipated, and that is why I recommend his approach to McDowell in The Formation of Reason.Footnote 41 Vygotsky represents the human infant as endowed with what he calls ‘elementary mental functions’, which include non-linguistic thought (basic problem-solving activity), pre-intellectual speech, sensory awareness, involuntary or associative memory, and basic forms of attention, volition, desire, and emotion. These capacities are portrayed as independent modules and their development is a function of the child's biological maturation. The nature of the elementary functions – their deliverances, development and interrelations – can be explained in causal or biological terms. They represent a unity, the principle of which is to sustain an animal life – to fulfil needs for food, comfort, warmth and safety, and to cope with the dictates of the immediate environment.

The mature human mind, in contrast, forms a system of ‘higher mental functions’ that includes linguistic thought and reasoning, intelligent speech, voluntary memory and attention, rational perception, desire, and emotion, and intentional volition. These higher functions are unified in a number of ways. First, they are ‘interfunctionally related’, in that the functioning of any one informs the functioning of the others and the development of any one makes possible changes in the others. Second, their common currency is meaning; they deal in content-bearing states, which can be expressed in propositional form, so that what can be remembered can be thought of, spoken of, reasoned about, regretted, and so on. Third, the principle of their unity is their role in sustaining the perspective of a rational animal life, self-conscious and self-determining.

Vygotsky argues that the key point in the transition from elementary to higher mental functions lies in the convergence of the developmental trajectory of two hitherto independent functions – pre-linguistic thought and pre-intellectual speech. This occurs when the child begins to deploy articulate sounds or gestures in problem-solving activity. With this, intellectual speech and verbal thought become possible, as do logical memory and forms of attention, perception, emotion and volition mediated by meaning, and as the child learns language so those possibilities are made real. As a result, the child's mind is restructured by her emerging facility with meaning.

It is central to Vygotsky's position that the higher mental functions are social in nature and origin. Meaning's home is communication, and communication originates in collaborative activities between individuals. Vygotsky illustrates this with his famous account of pointing.Footnote 42 He argues that the origins of pointing lie in the child's grasping for desired objects that are out of reach. Her grasping is interpreted by adults as indicating her desires and they respond accordingly. At first, then, the meaning of the child's act is established by others. Gradually she becomes attuned to how others respond to her grasping and she starts to deploy the movement to direct others' attention to objects she wants. In this way, the child's movement comes to have meaning for her, not just for others, and taking a physically simplified form, becomes a general means to direct another's attention.

So here is a pattern similar to the one Williams celebrates. The child's activity is first given meaning by surrounding adults, meaning that is only later ‘internalised’ by the child as she begins to point so that others should understand what she is indicating. With this, a movement that began in the space of stimulus-and-response becomes a true gesture, a performance in the space of reasons.Footnote 43 Although the case of pointing illustrates only one aspect of the child's entrance into meaning, Vygotsky takes it to be indicative of deep facts about the social nature of mind:

The pointing gesture most likely begins to indicate by movement what is understood by others and only later becomes a direction for the child himself.

Thus we might say that through others we become ourselves, and this rule refers not only to the individual (lichnost') as a whole, but also to the history of each separate function. This also comprises the essence of the process of cultural development expressed in a purely logical form. The individual becomes for himself what he is in himself through what he manifests for others. This is also the process of the forming of the individual.Footnote 44

This idea Vygotsky later expressed in his ‘general genetic law of psychological development’: each higher mental function originates in the interrelation of child and adult engaged in joint activity, and is only later mastered by the child.

Vygotsky clearly subscribes to something like the transformational view. Human elementary mental functioning is analogous to the psychology of apes, but the transition to the higher functions represents a qualitative change in the nature and structure of the mind. The higher mental functions are not simply more developed forms of their elementary counterparts, and there is no chance of a reductive reconstruction of the emergence of the higher functions out of the elementary. Instead, we must trace their emergence by attention to facilitating processes of biological maturation and enculturation. Vygotsky's position is consistent with the idea that a human individual acquires a second nature, one that is essentially cultural and historical in form. He might even have agreed that a human being ‘acquires a mind’, if by ‘mind’ we mean the higher psychological functions characteristic of rational thought, experience and action. But at the same time Vygotsky was no crass environmentalist. His view is compatible with a sophisticated nativism about the elementary psychological functions.

Interestingly, Vygotsky explicitly denies that training, in the sense of habit formation, plays a significant role in the child's entrance into meaning, while at the same time rejecting the rationalist alternative that grasping meanings should be understood as a kind of intellectual discovery on the child's part.Footnote 45 Neither training nor discovery will do. What is at issue is the transition from one mode of life, informed by a certain developing style of psychological activity, to another differently-organised mode of being. There is a sense in which this transition brings into being the child's capacity to think for herself; so we cannot represent her as exercising that capacity from the outset. The motor of development is the child's entering certain kinds of social relation with caregivers, but this process is complex and multifaceted and cannot be subsumed under a single concept like training. Consider again the pointing example. On Vygotsky's account, the child's development could not happen without the contribution of adults tending to the child. But nothing like training is going on.

6. Tomasello

Vygotsky was writing more than 80 years ago – that's ancient history from the perspective of contemporary psychology. I recommend Vygotsky's position to McDowell not on grounds of empirical accuracy, but for its general shape. Vygotsky shows us how we might think the emergence of second nature without resorting to conceptions of training that have their home in the kind of behaviourist paradigm Vygotsky was keen to transcend. But of course it would be nice to offer McDowell some compelling contemporary empirical work, broadly consistent with Vygotsky's vision. In this regard, the obvious resource is Michael Tomasello's path-breaking reflections on language acquisition and the natural history of human thinking.

Tomasello's view is that human communication is fundamentally cooperative. Cooperative communication first emerged in evolution, as it emerges in human individuals, with natural gestures of pointing and pantomiming. He argues that such communicative acts presuppose intentionality on the part of the communicators: not just individual, but shared intentionality. That is, human communication requires that speakers have an understanding of a plural subject, a ‘we’, with joint goals and shared knowledge. Tomasello speaks of a ‘shared intentionality infrastructure’ that (i) establishes common conceptual ground among human communicators (making possible joint attention and the formation of joint intentions), and (ii) involves pro-social motivations to request help from and offer help to other people, and to share information, attitudes, and emotions with them.Footnote 46

Tomasello argues that ‘language acquisition is possible only when young children have available to them something resembling the full shared intentionality infrastructure’.Footnote 47 This emerges around 12 months of age, when children start to engage in joint attentional activities and to point communicatively.Footnote 48 Tomasello and his collaborators represent its origins in a way reminiscent of Vygotsky: it issues from the convergence of two lines of development. The first is ‘a general primate (or perhaps great ape) line of development for understanding intentional action and perception, which evolved in the context of primates’ crucially important competitive interactions with one another for food, mates, and other resources'. The second is ‘a uniquely human line of development for sharing psychological states with others’.Footnote 49 Shortly after the convergence of these trajectories, children begin to construct shared goals and intentions, and to volunteer help when they discern that another's goals have been frustrated. They also begin to produce iconic gestures to indicate activities or absent objects. As the child develops language, these gestures are gradually displaced, taking up refuge in pretend play. This developing facility with shared intentionality, Tomasello argues, provides the foundation for the internalisation of social norms, shared beliefs, and cultural institutions.

Tomasello does much to establish this position experimentally and to defend its compatibility with evolutionary theory. For present purposes, however, the crucial point is his view of shared intentionality. Tomasello takes the shared intentionality infrastructure to have been forged by evolution. Though a facility for shared intentionality is ‘not simply innate, or maturational’, requiring for its emergence ‘constant interaction with the environment, especially the social environment’,Footnote 50 it does not issue from training, and once it is in place, it makes possible modes of teaching and learning that differ dramatically from training (at least if training is understood as akin to conditioning).

Shared intentionality has been a hot topic in philosophy for a quarter century or so. Where there is shared intentionality, you and I share the intention to φ, not just when you intend to φ and I intend to φ, but when you and I intend jointly to φ: I intend that you and I φ together, and vice-versa. That is, we intend that we φ together. Much of the philosophical discussion of the nature and possibility of such ‘we-intentionality’ considers the attitudes of mature subjects that have self-conscious understanding of their own mental states. It is important that Tomasello is also concerned with something more primitive. This is the basic intersubjectivity at the heart of joint attention and joint activity: where the focus of child and adult on a shared object of activity involves an element of mutual recognition of a common endeavour. This, Tomasello suggests, ‘seems to be present in nascent form very early in human ontogeny as infants share emotional states with others in turn-taking sequences’.Footnote 51

A facility for shared intentionality is not merely cognitive. It rests upon a fundamental desire to ‘join in’. Sebastian Rödl (with Tomasello in mind) brings this out nicely. He writes that, at around nine-months, the child's activity displays a significant motivational tendency to be like others. She not just imitates the behaviour of others, she imitates in order that they should recognise her as like them and that she should recognise this in herself. This she does for no other end. Knowing oneself as one of a kind, and sharing in this knowledge with others, is its own reward. Thus children engage in games the sole point of which is to establish mutual recognition. Rödl gives a lovely example, no doubt from life, where a father suddenly and for no apparent reason puts his forehead on the table in front of him, and then his child follows suit and they both roar with laughter. The game continues in the same vein. This, as he notes, is characteristically human behaviour. It is all about the first person plural – about us – and perhaps that's all it's about.

For Rödl, children's desire to establish their identity as one of kind is not mere instinct or the result of conditioning. Nor is it in the space of reasons, as this is usually understood. But it does manifest ‘an incipient consciousness of the general’, as Rödl puts it, and its deliverances are self-conscious acts, in that the child acts in light of an end, knows when she has achieved it, and in achieving it recognises herself through the recognition of others.

Rödl takes these ideas to undermine, not just the notion of training, but the whole framework of the transformational view. For him, from very early on – certainly before she begins to speak – the human child manifests her rational being. She does not start a ‘mere animal’, since the she is born with an orientation towards conspecifics that no other animal seems to possess, and she does not become a rational being. The Bildungsprozess is therefore better conceived as awakening and cultivating ‘habits of reason’. This is education's role. But it is the formation of something always-already there, not its creation. Hence education is best seen as facilitating the child's ‘growing into itself’. Now we can without paradox acknowledge autonomy as the end of education, but the price is to forsake the transformational view. Rödl writes:

If children were born animals, the first act of education would have to be abrichten. Indeed, the metaphysical conviction that the child is an animal that becomes a person through education seems to be the only possible source of the notion that abrichten plays an essential role in the development of children. Observation surely does not suggest it… Education begins with the shared consciousness of doing the same, which arises around nine months of age and in incipient consciousness of the general. From then on, abrichten is no longer possible; the form of the child's consciousness excludes it. Before that time, there is a logical possibility of abrichten of a human child. In point of empirical fact, it seems very difficult, if not impossible. In any case, such abrichten has no inner relation to the education that comes later.

This is a powerful passage. I don't think, however, that Rödl's insights wholly undermine the transformational view. Admittedly, what we have learnt from Vygotsky and Tomasello suggests it can be misleading speak of a transition from the ‘merely animal’ to the rational, or in terms of children ‘acquiring a mind’. But what Tomasello and Carpenter describe as ‘the big Vygotskian idea’ can remain up front and centre: ‘what makes human cognition different is not more individual brainpower, but rather the ability of humans to learn through other persons and their artifacts, and to collaborate with others in collective activities’.Footnote 52 In this sense, culture transforms the mode of life of the human animal. This is true of the species, but also of the individual: the child's entrance into culture, which is one with its learning language, transforms its life into that of an animal who is not just in the world, but who has a world, who can ‘hold the world in view’, as McDowell likes to put it. She becomes an animal responsive to reasons in the full sense: an animal that can, not just engage in ‘the game of giving and asking for reasons’, but can entertain the infinite and contemplate its own finitude. This is rational life, which finds expression in creatures who seek to make sense of themselves and who can take responsibility for their lives. We do not start out that way – except in the sense that we are thrown into culture from the outset. We become such through education, in the broadest sense of the word. In my view, this is the spirit of the position that emerges in Mind and World. So whatever one thinks of McDowell's formulations at the end of that work, we should continue to insist on the transformative role of culture, an idea that, when seen for what it is, ought to be uncontentious, even if it also possesses a certain metaphysical gravitas.

7. Hacking on Autism, Parfit on Persons and Human Beings

My reader may wonder why I am keen to preserve McDowell's view of the acquisition of second nature, casting Vygotsky and Tomasello as means of developing and defending McDowell, rather than simply setting McDowell aside and concentrating on them. The reasons are two. First, I think elements of McDowell's view of second nature illuminate psychological development. And second, I think McDowell's overall project, of which his view of second nature is part, is profound. To illustrate this, I will consider the relevance of McDowell's position to the two claims I mentioned earlier: Hacking's on autism and Parfit's on persons and human beings.Footnote 53

Hacking on Autism

Autism is a developmental disorder characterised by difficulties with social interaction and communication, narrow interests and repetitive actions. It is considered a spectrum condition, with ‘low-functioning’ individuals at one end, who have usually experienced significant delay or impairment in language development and have a low IQ, and ‘high-functioning’ individuals at the other, who have an average or better IQ and who have experienced no substantial language problems, save perhaps difficulty with figurative language. There are many theories of autism on the contemporary scene. One prominent view is that autistic individuals have problems with the concept of mind. They have difficulties ‘reading’ other people's mental states and hence problems predicting their behaviour, and this accounts for their social ineptitude, problems making friends, seeming lack of empathy, difficulties understanding jokes, and for the loneliness, anxiety and misery that can issue from the resulting social isolation. This is sometimes known as the ‘mindblindness’ theory of autism, and it is usually understood in terms of the paradigm ‘theory of mind’.Footnote 54 This paradigm represents our vocabulary for psychological description and explanation (our ‘folk psychology’) as, or as akin to, a theoretical vocabulary. Mental states and phenomena are unobservables; they are ‘internal states’ we posit to explain and predict behaviour. Autistic individuals, it is argued, lack a theory of mind, or are delayed in developing such a theory and poor at deploying it.

In a series of recent articles, Hacking offers an interesting variant on mindblindness.Footnote 55 He cites Wittgenstein's remark that ‘The human body is the best picture of the human soul’,Footnote 56 together with reflections from the psychologist Wolfgang Köhler (1887–1967) on human beings' ability to judge what others are thinking and feeling from their demeanour – i.e. their facial expressions and comportment (their ‘body-language’, as we might now say). In conversation, I can discern that one of my interlocutors doubts what I have just said and is impatient to speak, another is credulous, another bored, another uncomfortable with the topic. Köhler represents this ability as ubiquitous, as ‘the common property and practice of mankind’.Footnote 57 Psychology's failure to acknowledge the phenomenon is, he claims, a case of being blinded by the obvious.

Hacking contends that although ‘Köhler's phenomenon’ accurately describes the way in which neurotypical individuals can be aware of others' states of mind, this is not true of the autistic. They precisely lack this ability immediately to discern the mindedness of others. In contrast to theory of mind, Hacking denies that this is a deficit in theoretical knowledge. It is rather an inability to see (or perhaps better ‘intuit’) what others are thinking or feeling. It is not a matter of inference but of direct awareness akin to perception (hence ‘mindblindness’ is less of a metaphor than may first appear). The autistic can learn to work out what others are thinking and feeling by theorising. They can develop a theory of mind, but theoretical knowledge of other minds is knowledge of a different order from immediate awareness. To respond ‘naturally’ to a loving look, or a grief-stricken expression, one cannot be working out what is going on with the other person. One has to ‘see’ it.

Hacking describes the development of the autistic child as ‘non-Vygotskian’. Because autistic children lack Köhler's phenomenon they are not ‘in society’ in the full sense and are unable to ‘internalize social relationships to form concepts of the mental’.Footnote 58 I think this is insightful, but it gets things backwards. McDowell's concept of second nature can help us see this.

The notion of second nature contains an ambiguity. There is the idea, central to the above discussion, of a way of being. Thus McDowell speaks of acquiring a second nature. But there is also the rather different idea of forms of activity (habits) that, though acquired, are internalised to the degree that they are exercised as fluently and spontaneously as innate or instinctual responses. This idea of second nature turns up in everyday discourse, when, e.g., a music teacher tells her student to keep practising so that the finger movements become ‘second nature’. McDowell is sometimes criticised for conflating these notions,Footnote 59 and it is true that he does not distinguish them. But they come together in his thinking in the following way. In acquiring a second nature, the individual acquires conceptual capacities that become second nature to her in that they are exercised spontaneously (we deploy concepts without having to work out how) and in that the responsiveness to reasons they enable is, in at least some cases, immediate and non-inferential in character. For example, when we possess the relevant concepts we are able to acquire non-inferential knowledge of the world through perception and testimony; of morally and aesthetically relevant properties of situations and of how to act; of what others mean by what they say, and thereby of their mental lives. On the latter issue, McDowell writes:

[S]hared membership in a linguistic community is not just a matter of matching in aspects of an exterior that we present to anyone whatever, but equips us to make our minds available to one another, by confronting one another with a different exterior from that which we present to outsiders… [S]hared command of a language equips us to know one another's meaning without needing to arrive at that knowledge by interpretation, because it equips us to hear someone else's meaning in his words… [A] linguistic community is … bound together, not by a match in mere externals (facts accessible to just anyone), but by a capacity for a meeting of minds.Footnote 60

It is a natural complement to this view to hold, not just that we can know what someone is thinking by hearing meaning in her words, but that we can sometimes see what she is thinking or feeling by observing her actions or demeanour.Footnote 61 And these two dimensions to perceiving mind are interrelated, for to understand another's words is to see her as the subject of a mental life, which can find expression in ways other than speech. So we can say that it is a feature of psychological concepts that, in the normal case, their acquisition makes it second nature to us to discern the disposition of a person's mindedness in her demeanour, expression and actions, as well as in her words. This is a specific form of responsiveness to reasons, responsiveness to reasons for the ascription of mental states to others.

This McDowellian view equips us to say, with Hacking, that autistic people have trouble with this: they do not find it second nature to perceive the mental lives of others with such immediacy. But I think Köhler's phenomenon is not best seen as a mode of pre-conceptual awareness, the absence of which inhibits the internalisation of psychological concepts. It is an ability enabled by the internalisation of psychological concepts. If the development of the autistic child does not fit Vygotsky's model, this is not explained by impairment in Köhler's phenomenon. On the contrary, that impairment is precisely what needs explaining and its roots may lie in the autistic child's difficulties with shared intentionality, which may go back to autistic infants' problems with joint attention.Footnote 62

All this is speculative. My point is that Hacking is on to something, but McDowell offers a better way to understand the predicament of the autistic person. Better not to describe autistic children's development as ‘non-Vygotskian’. Most do internalise psychological concepts, but not in a way that enables lived experience of the minds of neurotypicals. (Of course, the reverse impairment also exists: resonating with autistic minds is not second nature to neurotypicals.) The McDowellian view suggests that there may be more than one way in which to compensate for this deficiency. The autistic can acquire theoretical knowledge of other minds – sometimes they are explicitly taught how to infer what others are thinking or feeling from their expression. But it might also be possible to cultivate a more immediate responsiveness to psychological reasons, rather in the way that someone who finds musical relations difficult to perceive may, by hard work, develop relevant perceptual competence (here the notion of training does get a grip). Perhaps the aurally-challenged musician may never hear intervals or harmonies as someone with perfect pitch does, and she will always sight-read with difficulty, but her developing capacities will, when exercised successfully, yield her genuine musical knowledge and not merely a theoretical surrogate. Where people have difficulties understanding other minds, we can hope for improvement along similar lines.

Parfit on Persons and Human Beings

An important aspiration of McDowell's philosophy is to foster ‘a firm and integrated conception of ourselves as rational animals’.Footnote 63 There are many elements to this: understanding the continuity of consciousness, so crucial to our conception of ourselves as enduring persons, as an aspect of the life of a living animal, and not as some kind of self-sufficient domain only contingently attached to a particular body; understanding action as bodily movement imbued with intelligence, so that our doings appear, not merely as causal consequences of mental goings-on, or as bodily signs of the mental, but as mind-in-movement; and, picking up the theme just discussed, understanding thought and emotion as there to be discerned in the deportment of a living human body.

There are many tendencies in philosophy, and in the culture at large, that make such a ‘firm and integrated conception’ hard to attain. McDowell's remark just quoted appears in the conclusion of an article on Parfit, who has recently published a paper no more friendly to McDowell's aspirations than the work with which McDowell originally took issue.Footnote 64 In this paper, Parfit denies that persons are animals: we are not human beings. A person is rather a part of an animal, the thinking part. Parfit argues his case principally by thought experiments of the kind for which he is famous. Suppose Parfit's head were transplanted onto Williams's body (presumably Parfit has Bernard in mind, rather than Meredith), the resulting person would be Parfit, but not the same animal as Parfit was once a part of. The same is true if we transplant, not the whole head, but just enough of Parfit's cerebrum to secure continuity of consciousness. So long as that is functioning, the resulting person will be Parfit, whether the brain-matter is housed in a human body or maintained by an artificial support system. But if this is true, then Parfit cannot be an animal; at best he's an ‘embodied part’ of one.

The argument is alluring, but (if you will forgive the pun) wrongheaded. I am uncertain that we know what we are imagining when we discuss such cases, and whether, so far as we do, our intuitions about them are sound. That we can enjoy the movie Freaky Friday does not show that its plot is metaphysically possible. Confronting Parfit's gruesome scenario, we might conclude that the resulting creature is a kind of chimera, a Parfit/Williams hybrid. But even if we grant the surviving person is Parfit, I am inclined to say that what his essay describes are events in the life of an animal, Parfit, who undergoes a dramatic ‘body transplant’ or is sustained by life-support system. This is a way for a human being to survive, admittedly a strange and no doubt vastly impoverished way, but it is still a kind of animal life. And if we want to say that only part of the animal survives, then why not say that only part of the person survives? This is surely no more counter-intuitive than the move to which Parfit must resort, namely, to portray the first-person pronoun as ambiguous, referring to the Inner-I (the true person) and the Outer-I (the body of which the person is strictly speaking only part). I, in contrast, believe that when we speak of educating ‘the whole person’, we mean it.

In my view, persons are animals of a kind that possess certain characteristic rational powers. Human beings are such animals. That conviction is firmer, I think, than any intuition prompted by Parfit's imaginary cases. I do not believe, however, that this matter will ever be settled by philosophical arguments that address them head-on. To overcome Parfit's view, we have to learn to stop thinking along lines that make the position attractive. In The Formation of Reason, I suggest that reflecting on the acquisition of second nature can help us develop a sense of the unity of our rationality and our animality, so that we do not cut loose the thinking side of our nature from the animal being that is its cause and unity. The developing mindedness of young children is not a history that we can separate from the character of their bodily being. It cannot be cast principally as a series of events in inner space. And the interrelations with adults that are essential to its development are not plausibly seen as a relation between two ‘inner selves’, but as an engagement in which mindedness is on the surface, present in the character of their joint activity. Where are we to find the emerging mindedness of our children if not in the manner of their bodily engagement with the world? This, I believe, is an antidote to a certain kind of philosophical confusion about the kind of beings that we are. It is not a powerful antidote, one guaranteed to be efficacious. But I recommend it notwithstanding. The transformational view may celebrate our rationality, but it does not forget that we remain what we always were – animals – even though our rational powers make it possible for us to lose sight of this in spectacular fashion. This is a lesson of McDowell's philosophy really worth heeding, and it motivates my enduring interest in the landscape his ideas place before us.Footnote 65

Footnotes

1 McDowell, J., Mind and World, 2nd edition (1st edition, 1994) (Cambridge, MA: Harvard University Press, 1996), 125.

2 Footnote Ibid., 84.

3 Footnote Ibid., 126.

4 S. Rödl, ‘Education and Autonomy’, paper presented at the launch of the Centre for Philosophy of Education at the Institute of Education, London, May 2012, and to appear in Bakhurst, D. and Fairfield, P., Education and Conversation (London: Bloomsbury Academic, forthcoming); Bakhurst, D., The Formation of Reason (Oxford: Wiley-Blackwell, 2011).

5 See, e.g., Bakhurst, D., Consciousness and Revolution in Soviet Philosophy (Cambridge: Cambridge University Press, 1991) and ‘Meaning, Normativity and the Life of the Mind’, Language and Communication 17 (1997), 33–51.

6 Ilyenkov, E.V., ‘Dialektika ideal'nogo’ (‘The Dialectic of the Ideal’), in his Isskustvo i kommunisticheckii ideal (Art and the Communist Ideal) (Moscow: Isskustvo, 1984), 68. An extract from this paper appears in English translation as ‘The Concept of the Ideal’, in Ilyenkov, E.V., The Ideal in Human Activity (Pacifica, CA: Marxist Internet Archive, 2009).

7 McDowell, J., ‘Comment on Hans-Peter Krüger's Paper’, Philosophical Explorations, 1 (1998), 122 , citing Wittgenstein, L., Philosophical Investigations, trans. Anscombe, G.E.M. (Oxford: Basil Blackwell), § 127. See also the conclusion of Lecture 4 of Mind and World.

8 McDowell explains: ‘What we are reminded of should be something that we knew all along, but were intelligibly induced to forget under the peculiar stress of philosophical reflection. What we are reminded of should be in itself – that is, considered in abstraction from the feeling of being confronted by deep and difficult intellectual problems that it is supposed to liberate us from – thin and obvious. I would be quite happy if someone responded to my diagnosis of some characteristic ailments of modern epistemology by saying, “Of course! How can we have been enticed into forgetting how obviously right it is to say that a repertoire of conceptual capacities belongs to the acquired nature of a mature human being?” And it is another way of making the same point to say that I feel misunderstood if someone responds rather as follows: “Second nature, that's an interesting idea; but don't we now need a philosophical theory of second nature?” This would be a refusal to take the reminder as I intend it.’ (‘Comment on Hans-Peter Krüger's Paper’, 123).

9 McDowell, , ‘Towards a Reading of Hegel on Action’, in his Having the World in View: Essays on Kant, Hegel, and Sellars (Cambridge, MA: Harvard University Press), 172.

10 McDowell, Mind and World, xxiv and ‘Postscript to Lecture V’, and ‘Wittgensteinian “Quietism”’, Common Knowledge 15, 365–372.

11 McDowell, Mind and World, 95.

12 Footnote Ibid., 104.

13 One might, for example, explore how in the course of their maturation and enculturation, and their acquisition of language, human children come by a conception of the world; that is, concepts, beliefs, forms of thought and reasoning, modes of self-understanding, and so on. To this end, we might emphasise the skills of narrative understanding that Alasdair MacIntyre in ch. 15 of After Virtue (London: Duckworth, 1981) deems central to the idea the unity of a human life, or the rich conception of folk psychology that Bruner, Jerome in Acts of Meaning (Cambridge, MA: Harvard University Press, 1990) portrays as enabling our understanding of self and others. MacIntyre and Bruner, in Vygotskian fashion, describe cognitive tools that are part of the cultural inheritance that makes us what we are. And the significance of such tools is not only cognitive, since they influence many aspects of our mental lives, including the character of our emotions, feelings, and desires. Their acquisition is therefore genuinely transformative, since they enable distinctively human awareness and understanding. But neither MacIntyre nor Bruner are doing ‘constructive philosophy’ in McDowell's pejorative sense.

14 McDowell, Mind and World, 84.

15 Footnote Ibid., 126.

16 McDowell, , ‘Responses’, in Willaschek, M. (ed.), John McDowell: Reason and Nature. Lecture and Colloquium in Münster 1999 (Münster, Hamburg, Loon: LIT Verlag), 98. See also McDowell, , ‘Responses’, in Lingaard, J. (ed.), John McDowell: Experience, Norm and Nature (Oxford: Blackwell), 220, where he offers a more expansive view, referring to ‘education, habituation, or training’.

17 Wittgenstein, L., The Blue and Brown Books (Oxford: Basil Blackwell), 77.

18 Huemer, W., ‘The Transition from Causes to Norms: Wittgenstein on Training’, Grazer Philosophische Studien 71 (2006), 205–225.

19 See, e.g., Luntley, M., ‘Training and Learning’, Educational Philosophy and Theory 40 (2008), 697.

20 Wittgenstein, Philosophical Investigations, §§ 6, 9.

21 Wittgenstein, L., Last Writings on the Philosophy of Psychology, Volume 1. Edited by von Wright, G.H. and Nyman, H., translated by Luckhardt, G. and Aue, M. (Oxford: Basil Blackwell, 1982), §§ 203 (see note), 206.

22 In The Formation of Reason (136–141), I call this ‘learning by initiation’. Meredith Williams notes the parallel between Wittgenstein and Aristotle in ‘The Significance of Learning in Wittgenstein's Later Philosophy’, Canadian Journal of Philosophy 24 (1994), 173 . I have more to say about McDowell's use of Aristotle and Wittgenstein in ‘Freedom and Second Nature in The Formation of Reason ’, Mind, Culture, and Activity 19 (2012), 172–189.

23 Medina, J., The Unity of Wittgenstein's Philosophy: Necessity, Intelligibility, and Normativity (Albany NY: SUNY, 2002), 167.

24 Though tales of Wittgenstein's own practice as a schoolteacher do give one pause. As is well known, after completing the Tractatus, Wittgenstein worked for several years in an elementary school. He eventually gave up the profession after he struck a boy so hard as to cause him to collapse. Later that day, one parent told Wittgenstein ‘he wasn't a teacher, he was an animal-trainer!’ (reported in Monk, R., Wittgenstein: The Duty of Genius (London: Vintage, 1990), 233).

25 Williams, M., Blind Obedience (London: Routledge, 2010), 105.

26 Footnote Ibid., 238. See also 106: ‘The status of the naïve learner's utterances (that, for example, they are taken as judgments or requests) is a function of the status extended to those utterances by masters of that practice. The initiate learner speaks, makes judgments, requests, and the like only by virtue of a courtesy extended to the learner by those who have already mastered the practice.’

27 Williams, like McDowell, uses this expression, which of course originates with Wilfrid Sellars.

28 Williams, Blind Obedience, 100.

29 Footnote Ibid., 97.

30 Wittgenstein, L., On Certainty, edited by Anscome, G.E.M. and von Wright, G.H., translated by Paul, Denis and Anscombe, G.E.M. (Oxford: Blackwell), §141.

31 Some Wittgensteinians recognise that training must be distinguished from conditioning. For example, although in ‘The Significance of Learning in Wittgenstein's Later Philosophy’, Williams admits that ‘[o]stensive training or teaching, for Wittgenstein, has affinities with the behaviorist notion of conditioning’ (178), she later claims that training is ‘importantly different’ because it aims to instill conformity with a normative practice rather than mere dispositions to respond (Medina takes a similar position, The Unity of Wittgenstein's Philosophy, 158–159). But this merely recognises the problem to which Luntley is drawing attention without solving it.

32 Luntley, ‘Training and Learning’, 709–10 n9.

33 Footnote Ibid., 703.

34 Wittgenstein, Philosophical Investigations, §69 (Wittgenstein's emphasis).

35 Luntley, ‘Training and Learning’, 698.

36 Luntley (‘Training and Learning’, 710 n16) quotes Fodor and Lepore with approval: ‘Wittgenstein suggests that first language learning is somehow a matter of “training”; but he says nothing intelligible about how training could lead to learning in a creature that doesn't already have a mind.’ ( Fodor, J. and Lepore, E., ‘Brandom Beleagured’, Philosophy and Phenomenological Research LXXIV (2007), 677–691 ).

37 Someone might protest that Williams poses this question as one a Quinean must confront. It is not obvious that she thinks the same question is a live issue for the Wittgensteinian. Nevertheless, her text gives the impression that the appeal to training into social practices is a way to answer the question rather than to exorcise it.

38 Wittgenstein, Philosophical Investigations, §415.

39 Of course it is important that they not do things that might inhibit their child's development, such as radically isolating or otherwise abusing them.

40 Williams, Blind Obedience, 255.

41 Bakhurst, Formation of Reason, 152–157. For more detail, see my Consciousness and Revolution in Soviet Philosophy, ch.3; ‘Social Memory in Soviet Thought’, in Middleton, D. and Edwards, D., Collective Remembering (London: Sage, 1990), 203–226 , and ‘Vygotsky's Demons’, in Cole, M., Daniels, H. and Wertsch, J. (eds), The Cambridge Companion to Vygotsky (Cambridge: Cambridge University Press, 2007), 50–76.

42 Vygotsky, L.S., History of the Development of the Higher Mental Functions. The Collected Works of L.S. Vygotsky, Volume 4. Translated by Hall, M.. (New York: Plenum Press, 1997), 104–5.

43 Note that although Vygotsky here discusses an isolated gesture, he would be happy with the holistic view that, when it comes to the child's emerging facility with meaning, ‘light dawns gradually over the whole’.

44 Vygotsky, History of the Development of the Higher Mental Functions, 105.

45 Vygotsky, L.S. and Luria, A., ‘Tool and Symbol in Child Development’, in van der Veer, R. and Valsiner, J., The Vygotsky Reader (Oxford: Blackwell, 1994), 113.

46 Tomasello, M., Origins of Human Communication (Cambridge, MA: MIT Press, 2008), 11–12.

47 Footnote Ibid., 104.

48 Footnote Ibid., 139 ff.

49 Tomasello, M. and Carpenter, M., ‘Shared Intentionality’, Developmental Science 10 (2007), 124.

50 Tomasello, M., A Natural History of Human Thinking (Cambridge, MA: Harvard University Press, 2014), 145–147.

51 Tomasello and Carpenter, ‘Shared Intentionality’, 124. In his recent book, A Natural History of Human Thinking, Tomasello distinguishes two kinds of shared intentionality. First, there is joint intentionality, which is essentially second-personal, involving collaboration and communication between ad hoc individuals in the moment. This presupposes ‘personal’ common ground between the participants, but need not rest on any conventional modes of communication or socially-established norms. Second, there is collective intentionality, which involves a relation between an individual and a transpersonal ‘we’, mediated by cultural common ground, group norms, and conventionalised modes of communication. Tomasello thinks that both forms of shared intentionality are distinctively human. Joint intentionality is a developmental prerequisite of collective intentionality, both phylogenetically and ontogenetically (children's facility with joint intentionality emerges around their first birthday, but they are not at home with collective intentionality until age 3), and collective intentionality is a precondition of many of the characteristic forms of human mindedness. This picture further undermines the usefulness of the concept of training. It now looks very implausible to think of training, as Wittgensteinians do, as the initiation of a normatively inept creature into forms of collective intentionality. Once we countenance the significance of joint intentionality (the hallmark of which is negotiation rather than training), we see that training is only one among many avenues of the cultivation of collective intentionality. To all this, I would add one cautionary note. It can be misleading to portray joint intentionality as essentially second-personal, for what is at issue is not just I-Thou relations, but primitive first-person plural conceptions of ourselves as ‘we’ (as Rödl's reflections (below) bring out). Lurking here are vexed questions of whether there is a fundamentally irreducible ‘we-mode’ of intentionality, an issue informatively addressed in Gallotti, M., ‘A Naturalistic Argument for the Irreducibility of Collective Intentionality’, Philosophy of the Social Sciences 42 (2012), 3–30 , and Gallotti, M. and Frith, C., ‘Social Cognition in the We-Mode’, Trends in Cognitive Sciences 17 (2013), 160–165.

52 Tomasello and Carpenter, ‘Shared Intentionality’, 121.

53 I recognise, of course, that the idea of an alliance between McDowell, Vygotsky, and Tomasello (let alone Rödl) raises many questions about the compatibility of their respective views. In The Formation of Reason I take this up with respect to Vygotsky and McDowell, though there I largely overlook the following issue. McDowell's notion of second nature draws on a distinction between two modes of intelligibility – natural-scientific and space-of-reasons intelligibility – which he sometimes glosses in terms of the distinction between explanation and understanding (‘Comment on Hans-Peter Krüger's Paper’, 121). But Vygotsky explicitly rejects the dichotomy between psychological theories that aspire to causal explanations and those that maintain that psychological phenomena demand a different order of explanation – hermeneutical understanding (see History of the Development of the Higher Mental Functions, 7–8). Where does this leave the supposed union of Vygotsky and McDowell? I do not think Vygotsky would have denied that there is an important difference between explaining particular instances of belief, action, and other intentional phenomena, and explaining natural goings-on in which reason does not figure. Moreover, Vygotsky is clear that psychology cannot explain the higher mental functions by principles of biological or physical science. We need to employ explanatory principles appropriate to the subject matter under study, at each stage of its development, and this involves understanding the mind as a cultural-historical phenomenon. However, Vygotsky aspired to a systematic and comprehensive account of psychological development worthy of the name ‘scientific’. It all depends, of course, on what counts as scientific, and he would likely have agreed with those who complain that Mind and World operates with too narrow a conception of science (such as Macdonald, G., ‘The Two Natures: Another Dogma’, in Macdonald, C. and Macdonald, G. (eds), McDowell and his Critics (Oxford: Blackwell, 2006), 224–225). McDowell subsequently broadened his view, but he continues to hold that responsiveness to reasons lies outside the reach of natural-scientific explanation. Vygotsky would have agreed that ‘human beings are unique among living things – outside the reach of the sort of understanding achievable by a scientific biology – in virtue of the freedom that belongs with our responsiveness to reasons as such’ (McDowell, ‘Response to Graham Macdonald, in McDowell and his Critics, 237), but recommended that McDowell broaden his conception of science still further (rather as he proposes to do with the concept of nature). I'm inclined to think that Tomasello would concur – his work is clearly scientific, but hardly displays the kind of narrow naturalism that is McDowell's target.

Discussion of the compatibility of McDowell's position and Tomasello's, which would need to take proper account of the latter's recent Natural History of Human Thinking, must await another occasion, but I must address one obvious issue. McDowell asserts that children acquire the ability to think and act intentionally as they enter the space of reasons, while Tomasello attributes individual intentionality to infants and apes. This looks like an irreconcilable difference, but it need not be. McDowell sets the bar for intentionality very high: it requires the ability to form propositional thoughts presupposing a network of concepts and a repertoire of conceptual capacities, including self-consciousness. For Tomasello, in contrast, individual intentionality presupposes only the ability to solve problems about how to act by representing situations and actions ‘off line’, operating on such representations, and monitoring the plausibility and desirability of possible outcomes. There is no reason why McDowell cannot credit infants and non-human animals with such abilities.

54 Baron-Cohen, S., Mindblindness: An Essay on Autism and Theory of Mind (Cambridge, MA: MIT, 1995). Baron-Cohen recognises that the mindblindness theory needs significant supplementation by other approaches. For a helpful review see his Autism and Asperger Syndrome (Oxford: Oxford University Press, 2008).

55 Hacking, I., ‘Kinds of People: Moving Targets’, Proceedings of the British Academy 151 (2007), 285–318 ; ‘Humans, Aliens, and Autism’, Daedalus 138 (July 2009), 44–59 ; ‘How We Have Been Learning to Talk About Autism: A Role for Stories’, Metaphilosophy 40 (2009), 499–516.

56 Wittgenstein, Philosophical Investigations, II.iv, 178.

57 Köhler, W., Gestalt Psychology (New York: Horace Liveright, 1929), 250–251.

58 Hacking, ‘How We Have Been Learning to Talk About Autism’, 504.

59 See Forman, D., ‘Autonomy as Second Nature: On McDowell's Aristotelian Naturalism’, Inquiry 51 (2008), 571.

60 McDowell, J., ‘Wittgenstein on Following a Rule’, in his Mind, Value, and Reality (Cambridge, MA: Harvard University Press, 1998), 253.

61 As McDowell maintains in ‘Criteria, Defeasibility, and Knowledge’ (in his Meaning, Knowledge, and Reality (Cambridge, MA: Harvard University Press, 1998)). Here he suggests an important qualification. In a case where, say, we discern someone is angry from their demeanour, we do not need to insist that their emotional state is itself an object of direct perception. We need only say that their anger is expressed in their demeanour in such a way that awareness of their demeanour constitutes non-inferential knowledge of their emotional state (see in ‘Criteria, Defeasibility, and Knowledge’, 387).

62 See Leekam, S., ‘Why Do Children With Autism Have a Joint Attention Impairment?’, in Eilen, N., Hoerl, C., McCormack, T., and Roessler, J. (eds.), Joint Attention: Communication and Other Minds: Issues in Philosophy and Psychology (Oxford: Oxford University Press, 2005), 205–229.

63 J. McDowell, ‘Reductionism and the First Person’, in his Mind, Value, and Reality, 382; see also his ‘Criteria, Defeasibility, and Knowledge’, 470.

64 Parfit, D., ‘We Are Not Human Beings’, Philosophy 87 (2012), 5–28.

65 An earlier version of this paper was presented at the 36th Annual Wittgenstein Conference on Mind, Language, Action, held in Kirchberg, Austria, in August 2013, and appears in the conference proceedings. I am grateful to the Oesterreichische Ludwig Wittgenstein Gesellschaft for the invitation to speak at that meeting and for permission to reproduce parts of my presentation in this longer essay, and to the audience for helpful comments and criticisms, which led me to revise my views into the form they take here.

Ibid
Ibid
Ibid
Ibid
Ibid
Ibid
Ibid
Ibid
Ibid