Charles Taylor is, by his own admission, a hedgehog. Though the essays in these two volumes range over a variety of topics – the concept of a person, meaning, the value of cognitive psychology, sexuality as a mode of political control – they all argue for one basic idea: that the conceptions of objectivity and scientific method which we have inherited from the 17th century are unable to give us an account of ourselves.
The nub of Taylor’s position is the claim that man is a ‘self-interpreting animal’. One way in which a human agent differs from other animals is by the possession of a conception of himself, an understanding of who he is and what he is doing. An agent’s self-understanding is not merely a distinguishing property: as Taylor puts it, an agent ‘is partly constituted by that understanding’. Self-interpretation is part of the essence of human life, so any theory which cannot account for it cannot be an adequate theory of man. According to Taylor, there is no way in which a theory based on the model of the natural sciences can capture this fundamental feature of human life. The reason is that it is the goal of the natural sciences to abstract from ‘anthropocentric properties’: that is, from those properties of objects which must ultimately be understood by reference to the experience of human agents. Secondary properties, like an object’s colour, smell and taste, provide paradigms of anthropocentric properties. It is generally agreed that if one were to give a scientific description of what a mindless physical object is like in itself, independent of the experience of observers, that account would not mention an object’s colour. And even a scientific explanation of our experience of colour when we look at the object would not mention the colour. It would, rather, mention the reflection of light rays off the surface of the object and the structure of our neuro-physiological make-up.
Man, according to Taylor, cannot be accounted for in non-anthropocentric terms. To understand what it is to be a man, one must understand what it is like to be a man. And to understand what it is like to be a man, one must ascribe both to the world and to man what Taylor calls ‘subject-referring properties’. Subject-referring properties are a special type of anthropocentric property: they ‘are properties which can only exist in a world in which there are subjects of experience, because they concern in some way the life of the subject qua subject’. Take, for example, shame.
Taylor’s argument is double-edged, with repercussions both for the contents of our minds and for the contents of the world. On the one hand, for an agent to be capable of feeling shame, he must live in a world in which there are shameful situations. Of course, not every experience of shame need be a response to a shameful situation. Given any experience of shame, we can always ask, ‘Was the reaction appropriate?’, ‘Was it really a shameful situation?’, and the answer may be ‘no’. However, the very content of an experience of shame is established by its general relation to shameful situations. In a world which entirely lacked shameful situations, there is no way an occupant of that world could feel shame, for there would be no way an emotion could draw into itself the content required for being a feeling of shame. On the other hand, ‘shameful situation’ is a subject-referring property. In a world which lacked subjects capable of experiencing shame, there would be no shameful situations.
Taylor claims that our contemporary conception of objectivity cannot cope with a subject-referring property like ‘shameful’. For this conception, Taylor says, divides up the world into objective properties and subjective reactions to them. But ‘shameful’ does not fit into the supposed dichotomy. It is not an objective property of the world, because ultimately it cannot be understood independently of the experience of subjects capable of feeling shame. Neither is it subjective, for to judge that a situation is shameful is not equivalent to feeling shame. A situation may be shameful, even if the agent does not experience it as such; and an agent may feel shame even though the situation is not really shameful.
But if our contemporary conception of objectivity cannot cope with a subject-referring property like ‘shameful’, neither can it cope with human beings. For the ability to feel emotions such as shame, dignity, pride, remorse, and to conceive of ourselves as experiencing the world in these ways, lies at the heart of human existence. The point is not merely that these experiences are important to us. That might be true even if they were ultimately to be understood as subjective responses to the world. The point is rather that these experiences help to constitute man’s fundamental nature, so any theory which purports to be a theory of man must account for these constitutive experiences. Yet there is no way to understand what it is for a being to be capable of having such an emotional and cognitive life other than from the standpoint of a being capable of living that life: that is, from an anthropocentric perspective.
Taylor diagnoses what, in Hegelian terms, might be called a contradiction in our conception of objectivity. In viewing the rest of nature objectively, we are able to promote a flattering self-image – an image of ourselves as capable of achieving a type of dispassionate disengagement from nature. Nature ceases to be that within which we must discover our role, that which determines our purposes for us, and becomes instead that whose purpose is determined by us. However, when we try to extend this dignifying outlook by including ourselves within it, we seem to rob ourselves of all dignity. We turn ourselves into objects, and what emerges are theories of human life – such as behaviourism in psychology – that are implausible and inadequate.
The issues which Taylor raises in these essays are, I believe, of fundamental importance both for the development of a philosophical account of the world and for the more tangible problem of how to proceed in the social sciences. The essays have an engaged, committed, polemical style: they are written by a man who really does believe that ‘there can be no dispassionate awareness of the human good.’ And I believe that many of these essays may legitimately be regarded as pioneering. If today behaviourism in psychology seems a crude and dated theory of man, it is because it has been subjected to a sustained philosophical critique in which Taylor’s essays have played a significant role. However, in this virtue lies the greatest weakness of these two volumes. Because Taylor’s essays tend to be directed against specific theories, his criticisms have an ad hominem quality. What I wanted to know was: given the very best theory of objectivity that could be devised which was still recognisably a descended of the 17th-century conception, would it necessarily be incapable of giving us an account of ourselves? Taylor suggests that we should answer this question affirmatively, but he has not given us a satisfying reason for doing so, for he never confronts a sufficiently sophisticated formulation of the position he is criticising.
I do not believe that Taylor succeeds in showing that the contemporary conception of objectivity cannot account for a subject-referring property like ‘shameful’. Consider, by way of analogy, an ordinary colour like green. On the one hand, green is not an objective property of the world, in the sense that a description of the world which only mentioned properties which involved no reference to the experience of agents would not mention green. On the other hand, green is not subjective, in the sense that green cannot be reduced to or explained in terms of the experiences someone has when he has the experience of something looking green. (Indeed, it is hard to imagine how we can explain what it is for something to look green other than by reference to the normal visual experience of observers when they see things which are green.) Yet one cannot validly infer that the contemporary conception of objectivity cannot cope with colours, for the objective-subjective distinction, as here presented, is not a genuine dichotomy. There is room for anthropocentric properties, like colour, which we may variously call ‘subjective’ – if we wish to contrast them with non-anthropocentric properties – or ‘objective’ – if we wish to contrast them with properties exclusively ascribed to the phenomenology of experience.
There is a prevalent picture of objectivity which leads us prematurely to belittle it. The picture is of our scientists talking to Alpha Centauran scientists who, though intelligent, lack anything like the human sensory apparatus. But there are no Alpha Centaurans. And the point of constructing an objective account of the world does not lie in the hope that we may some day communicate with them. Even when we give a scientific account of colours which makes no mention of colours, this explanation is being constructed by and for beings who do live in a world they experience as coloured: namely, ourselves. It is doubtful whether this explanation of colour would have any value for, or even be intelligible to, beings who experienced the world as colourless. One value of a scientific account of colours, for we who do live in a coloured world, is that it enables us to have the following reflective thought: the fact that the world contains colours depends in some way (difficult to set out) on the fact that it contains conscious, sentient beings.
If one wishes to call ‘reality’ those properties which the world possesses when one has abstracted from there being any observers, and ‘appearance’, those properties which in any way depend on the human perspective, then a case may be made that colours are appearances. Since at least the fifth century BC there have been philosophers and scientists who have wished to make such a case. However, with the world carved up in this way, it is difficult to avoid the further reflective thought that the world does contain observers. A satisfying account of the world cannot rest content with dismissing colours and colour-experience as appearances: it must extend its conception of reality sufficiently to explain such appearances. This extension of the scientific conception of the world so as to include and explain the various perspectives on the world, including itself, is what Bernard Williams has called ‘the absolute conception of reality’. Now it may be that the absolute conception is ultimately incoherent; and it may be that in trying to flesh out the absolute conception we find that our conception of objectivity does have to be radically revised. Though I am sympathetic to both suggestions, I do not think that Taylor or anyone else has shown that they are more than suggestions.
Taylor is, I believe, right to claim that shame cannot be explained as a subjective response (a feeling) to an objective, morally inert world. But that is only one way in which the scientific account of the world might be extended: there are others. Taylor is also correct that no explanation of shame could be fully intelligible to beings who were insensitive to shameful situations. But the point of an objective account is not to explain to Mr Spock what a shameful situation is. Nor need it be to explain shame away. The aim is to give beings who do live in a world of shameful situations a perspective on shame which is not the perspective of shame itself.
It is far from clear what the perspective on shame might be like. And it is (non-absolutely) fascinating that it would be available only to beings who were also sensitive to shameful situations. Taylor is, however, mistaken in his belief that this constraint alone rules out the possibility of an objective account. In fact, I doubt whether this constraint even restricts the class of those who would find the perspective on shame intelligible.