R.G. Collingwood’s Principles of Art1 is a complex work: part treatise on aesthetic theory, part contribution to the philosophy of mind. In one of its aspects, it is a critique of certain theories of aesthetic value which I propose to call psychologistic. Such theories attempt to explain the value of art works by appeal to psychology, conceived of as a non-normative discipline that does not itself presuppose judgements of value.

In particular, I shall argue, Collingwood criticises the theory of value put forward by I.A. Richards in his Principles of Literary Criticism.2 Principles of Art is not explicitly framed as a response to Richards, and I do not claim that this is its main focus. Nevertheless, there are some grounds for thinking that a critique of Richards formed part of Collingwood’s purpose in writing the book.

Collingwood mentions Richards by name several times in the course of his book.3 Most of these references mark crucial points of disagreement with Richards. Principles of Art, like Collingwood’s other books, is relatively sparse in citations, especially of contemporary authors. Collingwood makes little effort to cite relevant literature, even literature that had in one way or another influenced him. With this in mind, that he chooses to make several explicit references to Richards seems significant.

Richards’s book, which was first published in 1926, was a major theoretical statement by a leading literary critic. Richards can, indeed, be considered a founding figure of academic literary criticism. His theory was a salient feature of the intellectual landscape in 1938, when Principles of Art was first published. It was, further, a prominent contemporary representative of some of the trends in aesthetic theory that Collingwood most trenchantly criticises. For all these reasons, a comparison of the two books can shed light on Collingwood’s purposes in Principles of Art. It is this comparison which I seek to pursue here.

Collingwood’s criticism of Richards does not come from an entirely alien perspective. The two books are directed towards a similar overall project, which they approach in notably similar ways. Further, there are certain important presuppositions which they share. In section one, I outline some of these similarities. In section two, I discuss Richards’s psychologism. In section three, I address Collingwood’s critique of psychologism, and point to some of its implications.

1. Collingwood and Richards: a shared project

Both Collingwood’s and Richards’s books attempt to construct theories of aesthetic value, in the sense of theories that explain how and in what sense art works are of value. The theme of Principles of Literary Criticism is broader than the title suggests. Richards’s topic is not limited to literature; instead, he puts forward a theory of the value of art in general.

As well as sharing the same general project, Collingwood and Richards approach the problem in, broadly speaking, the same way. Both attempt to ground their theories of value in terms of an account of the psychology of experience. For both, the value of art works lies in the articulation, by the artist, of experiences. In order to clarify the nature of art, both engage in investigations of the nature of experience. As a result, such notions as feeling, emotion, and imagination play a key role in both theories. Further, both deny that there is a domain of specifically aesthetic experience. That is, what distinguishes art from other forms of human endeavour is not that art consists in the articulation of experiences which are themselves specifically aesthetic in character. The experiences with which the artist is concerned are those of ordinary human life and are not particular to the life of the artist.

In this respect, Collingwood’s Principles contrasts with his earlier work in aesthetics. His first major work to treat at length of the philosophy of art, Speculum Mentis (1924), is similar to the Principles in identifying art works with acts of the imagination.4 But the acts in question are imaginative acts of a particular sort: they are defined as having beauty as a constitutive norm. Similarly, in his Outlines of a Philosophy of Art (1925), Collingwood says that “that which art is the attempt to achieve is beauty.”5

For the early Collingwood, then, aesthetic experience is by its nature governed by the standard of beauty. It is to be considered valuable insofar as it counts as beautiful. What distinguishes aesthetic imagination from other forms of imagination is that it aims at beauty. Ugliness is defined as that which occurs when the aesthetic imagination fails to achieve what, by its nature, is its aim. On this view, if we are to understand the value of art work, we need to examine not experience in general, but beautiful experience in particular. A grasp of the concept of beauty will provide insight into the nature of the experience for which beauty provides a standard; more specifically, it will provide insight into the nature of the imaginative acts for which beauty provides an aim.

Collingwood’s later aesthetic theory comes into direct contact with Richards’s through the abandonment of this concern with the beautiful as a norm. Like Richards, Collingwood now appeals to the nature of experience in general. A striking fact about both books is how small a role there is for specifically aesthetic notions such as the beautiful and sublime, or indeed the dainty and dumpy.6 This shift from the examination of aesthetic notions to investigation of the nature of experience holds out the prospect, for both authors, of anchoring aesthetic theory in their broader views about the nature of the mind.

To see how, despite these similarities, Collingwood’s book functions as a criticism of Richards, we must look to some of the details of Richards’s theory of value. This is the concern of section two.

2. Richards’s psychologism

Richards aims to escape what he calls “the chaos of critical theories”.7 Literary criticism as practised down to his own time suffered from a lack of agreed standards and procedures. Different critics had proceeded by different methods. At times they used no clear method whatever. As a result there are no clear criteria for the evaluation of works of literature, and no means of settling critical disagreements. Richards’s project is the characteristically modernist one of placing literary criticism, for the first time, on firm foundations.

In particular, the foundations which Richards wishes to lay are scientific foundations. He appeals to psychology because he conceives of psychology as having a scientific status which literary criticism had previously failed to achieve.8 Specifically, he appeals to the psychology of experience. The scientific study of experience, he hopes, will provide us with clearer criteria for evaluating works of art, including works of literature. Richards appeals to a view of the mind which he takes to be the product of a new intellectual development: the application of the scientific method to the domain of the mental. A scientific psychology will, he hopes, bring in its train a scientific approach to aesthetic value. On the basis of such psychological insight into the nature of aesethetic value, literary criticism will turn into a rigorous discipline.

How does Richards hope to derive a theory of value from a science of experience? His conception of the link between the two can be summarised in the following three key claims:

(i) The value of art lies in the communication of experiences from the artist to the audience.9

(ii) The value of such communication derives from the value of the experiences themselves. Thus, it is experiences which ultimately do or do not have value.10

(iii) The value of experiences can be described in terms which are not themselves value-laden, and which are entirely psychological, in a restricted sense of this term.11

The value of art works lies in their capacity to communicate experiences which, Richards argues, are themselves particularly valuable. Their value lies in what Richards considers to be their purely psychological features. He points particularly to the richness of detail of certain experiences, and the degree of consciousness with which that detail is felt. Artists are prone to enjoy experiences which are inherently more ordered than the experiences of others, which tend to be more confused and indeterminate.12 Such experiences are worth having and worth communicating. Richards makes no appeal, in his account of the value of experience, to considerations pertaining to what the experiences are about, or which environmental factors have occasioned them. Aesthetic value is to be understood in terms of purely qualitative features which are intrinsic to experiences themselves.

The domain of the psychological is to be understood entirely in causal and not in intentional terms. Richards is explicit that this restriction defines his project of founding a theory of value in scientific psychology. It is only insofar as experience is understood causally and not intentionally that it can play the role that Richards envisages for it.

Psychologism, as I understand it here, is not merely the view that theories of aesthetic value must be founded in an account of the psychology of experience. Additionally, it relies on a particular conception of psychology. Psychology, on this conception, is a scientific discipline in the sense that it understands its subject matter in purely descriptive and not value-laden terms. Further, it does its explanatory work by appeal merely to the causal and not to the intentional features of experience.

To see what is meant by causal rather than intentional explanation, we can look to Richards’s account of sensation. Sensation is the passive aspect of experience.13 Sensations are prompted by stimuli, and the character of a sensation is largely dependent on the nature of its stimulus. Richards concedes that the disposition of the sensing subject is partly responsible for the character of a sensation. But he asserts that the role of the stimulus is greater.

Sensation is ‘cognitive’. It yields knowledge of objects by virtue of its stimulus-dependent character. It is because sensations are caused by external world objects that we gain knowledge of objects through them. In particular, the attribution of representational content to sensations plays no role in Richards’s account of their cognitive nature, nor does the attribution to them of meaning or aboutness. The cognitive content of a mental event is entirely dependent on its causal history.

Richards’s failure to appeal to such intentional notions is not a mere oversight, nor is it the result of a lack of philosophical sophistication on the part of a literary critic. In Richards’s own words14, he substitutes “the causation, the character and the consequences of a mental event as its fundamental aspects in place of its knowing, feeling and willing aspects.” In other words, Richards’s approach to sensation is an instance of his programme of replacing intentional with causal explanation.

In doing so, he self-consciously distances himself from traditional psychology. The advantage he sees in this move is that thinking, feeling and willing are specifically mental notions. Analysing

sensations and other mental events entirely in terms of their causal and qualitative properties, properties also possessed by non-mental events, holds out the prospect of an assimilation of psychology to the rest of science.

By contrast with sensation, emotions and feelings of pleasure and pain relate rather to our “attitude or behaviour” towards stimuli.15 Insofar as they are themselves cognitive, it is in virtue of their yielding knowledge of ourselves, rather than of objects. That is, their cognitive content is again dependent on their causal history as originating in the experiencing subject. Sensations, on the one hand, and emotions and pain or pleasure, on the other, are distinct experience types. “A sound which is pleasant for a while may become very unpleasant if it continues and does not lapse from consciousness. And indisputably it may remain qua sensation the same.”16 The distinction is drawn on the basis of the causal history of the experiences themselves. Thus it is a distinction which is internal to psychology even in Richards’s restrictive sense of that term.

Built on the dichotomy between sensation and emotion is a dichotomy between the “scientific” and “emotive” uses of language: uses which are entirely exclusive.17 Scientific uses of language function to communicate sensation. Emotive uses function rather to communicate emotional experience. The distinction between scientific and emotive language tends to be masked by grammatical terminology, which marks distinctions that run orthogonally to it. Thus we are inclined to suppose that a statement, or a question, is a speech act of a particular kind, independent of whether it belongs to scientific or emotive usage. This is misleading. One who utters an indicative sentence in the context of scientific usage is doing a different sort of thing from one who utters such a sentence emotively.

Art is distinguished from other forms of communication by the specifically emotive nature of the experiences which art works convey. Indeed, poetry is described simply as the supremely emotive form of language. It is appropriate to assess a sentence of the former kind for its truth or falsity, but point-missing to do so when reading poetry, which is “the supreme form of emotive language.”18 Poetry is thus valuable, when it is valuable, for its capacity to communicate valuable emotional experiences. But the value of the experiences themselves derives from those of their features which fall within the domain of psychology, where psychology is conceived as a purely causal study of experience. In the next section, we will see how Collingwood dissents from Richards’s view.

3. Collingwood’s anti-psychologism

Richards’s thesis that art is an emotive form of language may remind readers of Collingwood’s identification of art with language in Principles of Art. Collingwood indeed shares with Richards the notion that art expresses emotion, in a way that is at least analogous if not equivalent to linguistic expressions of emotion. But Collingwood rejects Richards’s distinction between emotional and scientific uses of language. All uses of language are, as such, emotional. In this section, I wish to show, in the first instance, how Collingwood arrives at this conclusion. Secondly, I show how it entails a rejection of psychologism in the theory of aesthetic value.

Collingwood denies Richards’s dichotomy of sensation and emotion. But he does not deny that any distinction whatever can be drawn between them. These are factors in experience which, for the purpose of analysis, must be distinguished.19 That is, in order to understand experience we must understand that it involves both sensation and emotion. But this is merely an analytical distinction. Sensations and emotions are not distinct experience types. On the contrary, in any given experience, sensations and emotions are inextricably bound together. Collingwood expresses this by saying that sensations have emotional charges20:

When an infant is terrified at the sight of a scarlet curtain blazing in the sunlight, there are not two distinct

experiences in its mind, one a sensation of red and the other an emotion of fear: there is only one experience,

a terrifying red.

An adequate account of the infant’s experience would have to include mention of its fear as well as its sensation of red.

The point is quite general. For Collingwood, no experience lacks this emotional element, though it may be more or less salient in the way in which we think about or describe experiences. For example, he describes a case of hallucination in this way21: “Dusk is coming on; I am writing in the window; and glancing back over my shoulder I see something in a darkened corner of the room which looks like a black animal crouching there.” Associated with the appearance of this animal is “a slight but perceptible feeling of fear” in “a person who, as a little boy, was frightened of the dark.”

When we consider experiences of this nature, we are often inclined to suppose that the experiencing subject has first visualised an animal, which afterwards becomes an object of fear for that subject. On this view, the hallucination itself is a visual and not an emotional experience, in the sense that the intentional object of the hallucination is a purely visual object. It is this visual object of which the subject is, having hallucinated it, afraid. To say, as Collingwood does, that experience is as such partly emotional, is to undermine this way of looking at the hallucination. The subject’s visualisation of the animal is, as such, an experience of fear. It is no more the case that the subject fears an animal which it has already visualised than that the subject sees as an animal that it already fears.

Similarly, thoughts, understood as episodes of thinking, carry their own emotional charges. Collingwood emphasises that to abstract the purely intellectual aspect of thinking from its emotional charge is to give a one-sided characterisation of the experience of thinking. But he also brings out the corollary. In order to understand a given emotional experience, one must understand the thought on which it is a charge.22

In an art work that concerns human life, the artist will seek to express the emotions felt by human beings in concrete situations. These emotions are not, by Collingwood’s theory, independent of the ways in which human beings think about the world. In King Lear, for example23:

Lear is envisaged, by Shakespeare and by ourselves, not simply as an old man suffering cold and hunger, but as

a father suffering these things at the hands of his daughters. Without the idea of the family, intellectually

conceived as a principle of social morality, the tragedy of Lear would not exist. The emotions expressed … are

thus emotions arising out of a situation which could not generate them unless it were intellectually


In order to understand the emotional content of (at least many) art works, then, we must also understand what thoughts they express. The two are not distinguishable. Collingwood’s anti-psychologism can now be expressed as follows: he contends that thought cannot be understood in purely psychological terms.

Thought is intrinsically intentional. It is distinguished from feeling by virtue of the fact that it has success conditions: truth conditions, in the case of propositional thought.24 In order to understand an agent’s thought, therefore, we must understand what counts for that agent as a success in thinking. The study of thought involves discovering what standards the thinker applies to his or her own activity of thinking. In order to understand an agent as a believer, for example, we need to understand her as attempting to think something true, and thus as failing if she does not. This constraint is involved in the very notion of attributing beliefs to agents.

It is important that the distinction between success and failure is internal to thought. We may, in addition, wish to assess from our own perspective the success or failure of a thinking activity. For example, we may find we agree or disagree with a given thinker’s beliefs. But to assess an agent’s beliefs in this way is to apply a standard which is external to those beliefs themselves. What matters for the study of thought as such is a grasp of the standards which are applied by the thinker.

One way to express this thought is to say that a genuine science of thought would be normative. But Collingwood’s prefers the term “criteriological”, precisely in order to avoid any implication that the norms that are to be applied to thought are those of the student of thought and not of the thinker herself25:

The characteristic of thought in virtue of which a science of thought is called normative consists not in the possibility that one man’s thought may be judged successful or unsuccessful by another, real though that possibility is; but in the necessity that in every act of thought the thinker himself should judge the success of his own act.

A science of thought will tell us, not what we ought to think, but what criteria are applied by agents in acts of thinking.

The normative or criteriological aspect of a science of thought arises from the fact that thought is intentional. Any attempt to study thought that ignores the question of criteria is thus inadequate to account for the intentionality of thought. In particular, since not all mental phenomena are, for Collingwood, intentional, not just any study of the mental will be apt to deal with the issue of thought. It is Collingwood’s contention that psychology, as he understands that discipline, though adequate for the study of some domains of the mental, is not adequate for the study of thought.

Collingwood defines psychology as the science of feeling, not of thought.26 It studies feelings by examining their causal relations both with each other and with the external world. This is legitimate, since for Collingwood feeling is not intentional. Collingwood emphasises that it is not cognitive27: “in feeling a coldness or seeing a redness or hearing a shrillness we [are] not cognizing an object but simply having a feeling, due no doubt to things in our environment but not itself constituting knowledge of these things.” But the non-cognitive nature of feeling is merely a consequence of the fact that in order to understand a feeling we do not have to understand it as having any object at all.28 Feeling as such does not have success conditions: in feeling we are not striving to do something at which we might fail. The purely causal science of psychology is thus adequate to deal with the topic of feeling. But just because it is a merely causal science it is not adequate to deal with the topic of thought.

Our immediate concern here is not so much with the merits of this definition as with its implications for the project of understanding aesthetic value. In particular, we may note that Collingwood’s conception of the domain of the psychological coincides with Richards’s. For both, psychology is a study of sensation and emotion in causal terms. This agreement may be obscured by Richards’s claim, which we saw in section two above, that sensation is cognitive. But, as was made plain there, in calling sensations ‘cognitive’ Richards is merely making a claim about their aetiology. He means that they originate externally to the sensing subject. Thus, apart from a point about usage of the word “cognitive”, there is nothing for Collingwood to disagree with in this area of Richards’s thought. Their disagreement arises over the question whether psychology is adequate to ground a theory of value.

Richards’s psychologistic theory of aesthetic value fails, for Collingwood, because it leaves no room for the normative aspect of experience. In noting this, however, we must be wary of a certain misunderstanding. In section one above, it was noted that the theory of art put forward in Speculum Mentis takes aesthetic experience to be governed by a certain norm, that of beauty. Acts of the aesthetic imagination, on this view, are by their nature directed towards the achievement of beauty. For the Collingwood of Principles of Art, by contrast, there is no particular norm governing aesthetic experience as such.

Instead, the normativity of the notion of aesthetic value arises from the fact that the emotions expressed by artists are typically the emotions which attend thought, and not those which arise at the level of mere feeling. Because of the intentionality of thought, to understand an art work is to understand the ways in which the artist sees the world, which in turn involves understanding the standards which he applies to it. Richards’s psychologism does not fail because he associates aesthetic value with the expression of emotion. On the contrary, Richards and Collingwood agree that art works are expressions of emotional experience. Rather, it fails because psychology is inadequate for the understanding of the emotions which artists typically express.

4. Conclusion

It hardly needs to be said that the material presented here falls far short of a satisfactory interpretation of the Principles of Art as a whole. I have concentrated on a single strand of that complex book. But that thread is a significant one: it represents Collingwood’s argument for the thesis that aesthetic value is not reducible to the domain of the psychology of experience.

Collingwood’s reason for holding this thesis of non-reducibility is also of interest. His point is not that art as such aims at some particular value – that of beauty, say – of which psychology is ill-equipped to provide an account. His thought is rather that a proper understanding of art works as expressive of human emotion requires more than psychology can provide. Psychology, as Collingwood and Richards both conceive it, is inadequate to provide a theory of aesthetic value because it is inadequate for the understanding of human beings qua thinking and desiring creatures.

Readers of Collingwood’s other work will recognise here a dominant theme of his writings, and in particular of his later writings, roughly from Principles of Art on. In his later work, he was concerned to argue that the methods of natural science are inadequate to the task of understanding human thought and action. The discipline that studies these things, using its own methods and presuppositions, is history. Collingwood’s anti-psychologism about aesthetic value is of a piece with his insistence that history is an autonomous discipline, irreducible to natural science.


AUSTIN, J.L. (1961), Philosophical Papers, ed. J.O. Urmson and G.J. Warnock (Oxford, Clarendon Press).

COLLINGWOOD, R.G. (1924), Speculum Mentis or the Map of Knowledge (Oxford, Clarendon Press, 1924)

__________ (1925), Outlines of a Philosophy of Art (London, Oxford University Press)

__________ (1938), Principles of Art (Oxford, Clarendon Press)

__________ (1998), An Essay on Metaphysics, revised edition, ed. Rex Martin (Oxford, Clarendon Press)

RICHARDS, I.A. (2001), Principles of Literary Criticism, second edition (London, Routledge)

1R.G. Collingwood, Principles of Art (Oxford, Clarendon Press, 1938).

2I.A. Richards, Principles of Literary Criticism, second edition (London, Routledge, 2001).

3Collingwood describes Richards as “at present the most distinguished advocate” of the theory of art as craft (Principles of Art, p. 35n) and criticises his distinction between scientific and emotive uses of language (pp. 262ff), while citing with approval Richards’s disdain for psychoanalytic approaches to criticism (p. 237n).

4 R.G. Collingwood, Speculum Mentis (Oxford, Clarendon Press, 1924), pp. 63-4.

5R.G. Collingwood, Outlines of a Philosophy of Art (London, Oxford University Press, 1925), p . 19.

6J.L. Austin, Philosophical Papers (Oxford, Clarendon Press, 1961) p. 131. Austin’s complaint is that philosophers had concentrated on too narrow a range of aesthetic concepts. Collingwood and Richards agree on a more radical thesis: philosophers of art should not be fundamentally concerned with specifically aesthetic concepts at all.

7Richards, Principles of Literary Criticism, pp. 1ff

8A caveat should be entered here. There is, in fact, little or no explicit discussion of actual empirical psychological research in Richards’s Principles. The psychology it contains is largely drawn from common sense, tempered by Richards’s own conception of what a scientific psychology must look like.

9Richards, Principles of Literary Criticism, pp. 20ff

10Ibid, pp. 39ff

11Ibid, p. 46

12Ibid, pp. 55-6

13Ibid, p. 90

14Ibid, p. 80

15Ibid, p. 90

16Ibid, p. 85

17Ibid, p. 244

18Ibid, p. 256

19Collingwood, Principles of Art, pp. 161-2

20Ibid, p. 161

21Ibid, p. 181

22Thus, Collingwood supposes that a genuine understanding of the emotional content of Archimedes’ exclamation ‘Eureka!’ would involve an appreciation of the discovery that prompted it. Principles of Art, p. 267.

23Ibid, pp. 294-5

24Collingwood, An Essay on Metaphysics, revised edition, ed. Rex Martin (Oxford, Clarendon Press, 1998) pp. 109-111

25Ibid, p. 109

26Ibid, pp. 106ff

27Ibid, p. 110

28On this point see Collingwood, Principles of Art, pp. 198-9. G.E. Moore’s concerns about the nature of the objects of sensation are dismissed as irrelevant to the project of understanding sensory experience.