Navigation – Plan du site

AccueilNuméros59Iris Murdoch’s Ethics of AttentionInteriority and Idiosyncrasy, Two...

Iris Murdoch’s Ethics of Attention

Interiority and Idiosyncrasy, Two Kinds of Privacy in Murdoch’s Conception of Mental Life

Intériorité et idiosyncrasie : deux types de valeurs privées dans la conception murdochienne de la vie mentale
Annalisa Paese


Il est de notoriété publique qu’Iris Murdoch a critiqué la psychologie morale dominante de son époque, la considérant exclusivement axée sur la dimension publique des concepts et incapable de rendre compte de l’existence et de l’importance de leur dimension privée. Selon Murdoch, ce type de psychologie morale ignorait deux sens distincts dans lesquels nos concepts peuvent s’entendre comme privés. Alors que le premier (c’est-à-dire leur intériorité) a été généralement reconnu, le second (c’est-à-dire leur idiosyncrasie)ne l’est pas encore. L’engagement de Murdoch à affirmer que certains concepts, en particulier les concepts mentaux, peuvent être idiosyncrasiques a été soit complètement oublié, soit mis de côté. Dans cet article, je tente de démontrer que l’attachement de Murdoch au caractère idiosyncrasique de certains aspects de notre vie mentale est au centre de sa conception de l’individu et qu’il enrichit cette dernière.

Haut de page

Texte intégral

  • 1 An example of this tendency is provided by Setiya 2013. Although he thinks that the tension between (...)

1In The Sovereignty of Good, Iris Murdoch targets a moral psychology that she characterises as ‘existentialist-behaviorist’ by attacking what she identifies as its ‘keystone’, namely, the argument according to which any picture connecting mental concepts to something inner is flawed and what is required for them instead is an analysis in terms of the publicly observable circumstances that allow us to acquire them in the first place. Murdoch maintains that this argument is a problematic development of a line of thought by Wittgenstein which rejects the idea that mental concepts can be understood in terms of objects to which the subject has infallible and exclusive access. This line of thought, she argues, has been mistakenly taken to support the view that inner life is non-existent or irrelevant and that overt action is the only thing that matters. The behaviorist reading of Wittgenstein’s work on privacy is no longer popular, but the import of Murdoch’s criticism of it has not been fully appreciated. Murdoch wanted to show that Wittgenstein’s considerations on privacy are compatible with two important senses in which our mental concepts are private but, while the first (i.e. their interiority) has been generally acknowledged, the second (i.e. their idiosyncrasy) has not. Murdoch’s commitment to the claim that mental concepts can be idiosyncratic has been either missed altogether or set aside as a slip generating doubts about the coherence of her professed realism.1 In this paper I argue that this commitment, far from being a marginal and problematic detail, lies at the very core of Murdoch’s conception of the life of the mind and is part of its attractiveness. I bring out the specificity and attractiveness of Murdoch’s understanding of privacy by comparing it with John McDowell’s. Both Murdoch and McDowell have defended the claim that we do have an inner life by holding that the dependence of the inner on the outer Wittgenstein draws attention to is more complex than some of his commentators have thought. Murdoch, though, goes farther than McDowell in placing limits on this dependence precisely by acknowledging and giving pride of place to the idiosyncrasy of certain aspects of our dealings with concepts. I argue that it is only with this move that we come to see a not merely incidental connection between privacy and morality. What Murdoch shows us with the sort of cases she focuses on is not just that some episodes can be private and morally salient, but that there can be episodes that are morally salient because of their privacy. This possibility, turns out, is a very important one.

Privacy and Moral Progress

2Murdoch schematically presents the general line of reasoning against inner objects through the following two claims, (a) such inner objects cannot be appealed to in deploying procedures for distinguishing good applications of a concept from bad ones (i.e. they are useless) and (b) they cannot be introspectively discovered (i.e. they are not there). The second claim has been defended by appeal to both empirical and logical considerations; it has been maintained that what introspection makes available is pretty scarce and hazy, and also that there are logical problems involved in the identification of such introspected materials.

3This general line of thought, as Murdoch points out, goes beyond Wittgenstein’s criticism of a certain way in which philosophy has unsuccessfully tried to make sense of mental life. There is, she argues, a significant difference between, on the one hand, an observation she attributes to Wittgenstein and, on the other, a problematic development of this observation various versions of which she finds in Hampshire, Hare, Ayer, Ryle. The observation is that it does not make sense to take the first-personal uses of mental concepts to be reports about inner objects. Its problematic development, which she calls the ‘genetic analysis of the meaning of mental concepts’, consists in drawing from it the conclusion that, by acquiring the capacity to apply such concepts (e.g. that of decision) in ordinary public contexts, ‘I learn the essence of the matter’ (Murdoch 1970, 12). As Murdoch immediately goes on to explain, this amounts to saying that mental concepts lack any structure besides their outer structure. There is no transition from a concept I acquire in learning to use a word in ordinary contexts to one having to do with inner experiences of a specific sort.

  • 2 The way of appropriating Wittgenstein’s work on privacy which I address in this paper is linked by (...)

4This characterisation of the genetic analysis already contains an indication of what motivates Murdoch to criticise the conclusions that have been drawn from Wittgenstein’s work, but it is easy to miss it. The reason is that, in articulating the contrast between the genetic analysis and her own conception of mental concepts, Murdoch exploits repeatedly the opposition between the outer (observable circumstances) and the inner (introspectable experiences) which might suggest that her exclusive target is behaviorism. What is not obvious is that this opposition matters for Murdoch for two connected but distinct reasons, that is, not just because our lives have a subjective character and it would be phenomenologically inaccurate to deny that they have it, but also because there is a link between the privacy of mental life and certain kind of moral progress. Shedding some light on this link is the aim of this paper as a whole, but I can say, in a preliminary fashion, that the inner structure of a concept is the dimension within which, according to Murdoch, any concept has the potential to undergo an indefinite process of complication which can (in the good cases) amount to this distinctive and important variety of moral progress. In her view, it is the characteristic tendency to engage in this sort of process, particularly with respect to mental concepts, that the genetic analysis most importantly misses. In other words, what’s wrong with the picture of concepts offered by the genetic analysis according to Murdoch is not just that it is behavioristic, but also that it is static. 2

Two Interpretative Difficulties

5In order to show that in Murdoch’s criticism there is much more at stake than behaviorism, we can start with two interpretative difficulties. Murdoch argues that the reliance on Wittgenstein’s treatment of privacy to justify the genetic analysis is misguided. She characterises the mistake as follows, ‘[b]ecause something is no use it has been too hastily assumed that something else isn’t there’ (Murdoch 1970, 10). This remark occurs at the beginning of a long discussion, and it is not immediately obvious what the ‘something else’ is whose existence has been denied on the ground that inner objects have no role in the practice of concept application. What we know is that the ‘something else’ is a phenomenon that Murdoch illustrates through an example about a man, to whom I will refer as R, who tries to establish, privately, whether what he feels is repentance (Murdoch 1970, 25) and through the widely discussed example of M and D which has to do with coming to be just and loving toward another individual after an initial unexpressed hostility (Murdoch 1970, 16–17). The most important feature of these examples for my purposes here is that each of them constitutes, in Murdoch’s intentions, an attempt at a special kind of moral progress connected with the idea of perfection, an attempt which is per hypothesis successful in M’s case.

6What are then the thing that has been demonstrated to be useless and the different thing which has been mistakenly taken to be not there? Here is a very natural way of answering this question. What we can come to recognise as useless thanks to Wittgenstein’s work in this area are, as we said, inner objects. What is meant here by ‘inner objects’ is items with certain specific attributes. They are supposed to be just like any other object but with the peculiarity of being cognitively available only to one subject (the one whose interiority they belong to). Furthermore, this subject is somehow infallible in this cognitive feat, as if she could look into a walled garden where such objects lie (according to the image used by Crispin Wright 1989). Finally, these objects are supposed to be the realities to which mental concepts correspond, i.e. their presence is identified as what makes the application of mental concepts correct. What has been too hastily concluded to be absent is inner life under any understanding of it. This entails the exclusion of what we ordinarily, i.e. without being driven by philosophical worries, would identify as our inner life.

7If we accept this reading, though, there is something that calls for an explanation. Inner life in the ordinary sense undoubtedly includes much more than the sort of circumstances that makes it the case, for instance, that a man might rightly judge, after long and difficult private reflection, that what he feels is repentance after all. It includes, for instance, completely mundane and uninteresting inner experiences, such as remembering, at one particular moment, that one has an errand to run. And yet Murdoch focuses on cases that she takes to exemplify a distinctive and important kind of moral progress. Why consider these cases if those about remembering that one has an errand to run would work just as well to reject the genetic analysis, i.e. a behaviorist reading of Wittgenstein’s treatment of privacy?

8A hypothesis is that Murdoch’s choice of cases depends on the particular themes of ‘The Idea of Perfection’, which is motivated by her interest in the philosophical understanding of moral personality. Questions about repentance are more typically connected with morality than the ordinary operations of our memory are. According to this way of understanding Murdoch’s chosen strategy, she might have just as well chosen to criticise the genetic argument by talking about any element of our mental life and not simply those that have moral relevance. In choosing the examples as she did, though, she showed not only that life which is, in a certain specified sense, private is there, but also that some of its episodes are morally salient.

9If this were the case, Murdoch could have rejected the behaviorist interpretation of Wittgenstein’s treatment of privacy by arguing that accepting it is compatible with the following claim:

It is not silent and dark within: We can understand instantiations of mental concepts as mental states brought about by mental processes which are not shadows of observable behavior in the sense that they obtain whether or not they find an expression in behavior.

10The sentence ‘it is not silent and dark within’ is used by Murdoch (Murdoch 1970, 13) in the course of her argument, and in what follows I will use it as a shorthand for this longer and less metaphorical claim. In order to establish this claim, one certainly does not need to rely on the intelligibility of cases in which the unfolding of the inner life of the subject counts as moral progress, let alone moral progress of a special kind which is connected with the idea of perfection. Yet, according to the interpretative hypothesis I am considering, once the claim that it is not silent and dark within is rescued from the behaviorist attack, the sort of episodes Murdoch is interested in, no more considered mythical, might be assessed for moral relevance. Their moral relevance (if they have any), though, would be quite independent from the fact that they are episodes in inner life. In other words, it would not be the fact that they are private that makes them morally relevant.

  • 3 Murdoch reiterates frequently that, in her view, the inadequacy of the genetic analysis is brought (...)

11Certainly, there is no reason to saddle Murdoch with the view that any application of a mental concept is necessarily governed by the effort to strive toward perfection. Her view is compatible not only with the claim that sometimes all there is to first-personal uses of mental verbs is the obtaining of certain publicly observable circumstances (as she explicitly concedes—see Murdoch 1970, 14), but also with the claim that pretty mundane inner experiences, such as that of suddenly remembering one has an errand to run, are part of one’s inner life too. And yet she repeatedly suggests that the sort of cases that she invites us to recognise as involving a striving toward perfection are needed for her attack to the genetic argument.3 She writes, for instance:

As soon as we begin to use words such as ‘love’ and ‘justice’ in characterising M, we introduce into the whole conceptual picture of her situation the idea of progress, that is the idea of perfection, and it is just the presence of this idea which demands an analysis of mental concepts which is different from the genetic one. (Murdoch 1970, 23)

12To summarise, the first interpretative difficulty about Murdoch’s criticism of the genetic analysis of mental concepts is that it seems to go beyond what is required. It seems that to refute this analysis all we need is vindicating the claim that it is not silent and dark within and, in order to do so, it is sufficient to show that episodes such as remembering that one has an errand to run (which are not morally relevant) can be understood as belonging to inner life. Nonetheless, Murdoch thinks that episodes of inner life that exemplify a kind of moral progress connected with the idea of perfection are essential to her argument.

13Of course, one might think that this is a mistake on Murdoch’s part. But what I take to be the case is that what can be imputed to her is just that she fails to clearly separate two different targets: (1) a behavioristic understanding of the circumstances that license the first-personal application of a mental concept and (2) a less specific and weaker idea of the identity of mental occurrences as dependent on public rules. While to reject (1) it is enough to bring out the perfectly non-mythical sense in which a first-personal application of a mental concept can be in order in the absence of publicly observable circumstances to license it, to destroy (2) it is necessary to bring back into view a further sense in which concept applications can be independent from public rules.

14This further independence from public rules marks the kind of privacy proper to the instances of moral progress that Murdoch wants to protect, which is different from the kind of privacy proper to episodes such as remembering that one has an errand to run. I am going to explain what this independence is and how it is illustrated by the examples chosen by Murdoch in the next section, but I will introduce the second interpretative difficulty first.

15This second difficulty consists in the fact that Murdoch repeatedly says that there is a connection between mental concepts and a kind of moral progress that involves the idea of perfection (this happens most explicitly at Murdoch 1970, 28) but she also says that perhaps all concepts are connected with the idea of perfection in this way (1970, 29) and, as I have already mentioned, concedes that there are uses of mental concepts that do not involve the idea of perfection at all (1970, 14–15, 35). Most importantly, Murdoch does not provide a characterisation of what counts as a mental concept.

16I think that the best way to understand what, according to Murdoch, is connected with the idea of perfection is in terms of what she calls ‘specialised uses of concepts’ (Murdoch 1970, 25) which are necessarily private. Though mental concepts paradigmatically lend themselves to such uses, any concept might end up being used in this way. Of course, it remains to be explained what specialised uses of concepts are, how (given her endorsement of Wittgenstein’s observation) Murdoch can take them to be meaningful despite their privacy, and what their distinctive privacy has to do with moral significance.

17I now move on to answer the questions raised by the two interpretative difficulties I have just identified. The solution is to be found in an understanding of Murdoch’s examples of R and of M and D. Murdoch’s discussion of them provides not just a vindication of the claim that it is not silent and dark within, but also an illustration of a more ambitious and controversial claim, namely, that certain uses of our concepts tend to become idiosyncratic in a way that can amount to a distinctive kind of progress. I will start with an explanation of how Murdoch vindicates the less ambitious claim in the next section before moving to the more ambitious in the following one.

Is Murdoch’s Conception of Privacy Mythical?4

  • 4 In this section and the following ones, I draw on McDowell 1998 and McDowell 2009 for the interpret (...)

18As I have anticipated, the two main examples that Murdoch uses to illustrate her conception of privacy are the example of M and D and the example of R, ‘a man trying privately to determine whether something which he “feels” is repentance or not’ (Murdoch 1970, 25). Here is the example of M and D: D is a young woman and M is her mother-in-law. M experiences a moral progress that consists in transforming her way of seeing D through an effort of self-criticism and attention toward her. M considers the possibility that she herself is ‘snobbish’, ‘conventional’, ‘narrow-minded’, and admits to herself that she is ‘jealous’. This self-criticism leads her to reconsider her idea of D and to engage in an effort of attention toward D, as a result of which she goes from thinking that D is ‘unpolished’, ‘lacking in dignity and refinement’, ‘tiresomely juvenile’ to thinking that she is ‘spontaneous’, ‘gay’, ‘delightfully youthful’ (Murdoch 1970, 16–17). Despite the radical change in her way of seeing D, M had behaved perfectly kindly toward D all along, so M’s transformation in the way she sees D does not result in any change in her behavior.

19The role of the example of M and D in Murdoch’s argumentation is complex because it serves many different purposes. Here I will only focus on three features of the activity which is described in it that are shared by the example about repentance. While the first two have to do with the claim that it is not silent and dark within, i.e. that nothing forces us to give up the very idea of interiority, the third one goes beyond that claim and its associated idea of privacy. It is this third feature that brings out the distinctively Murdochean concept of privacy which is connected with moral progress. I will discuss it separately in the next section.

20Murdoch characterises the kind of phenomenon that these two cases are meant to exemplify (1) as a specific kind of activity, (2) as something that is not hazy, but rather something that we find very familiar and (3) something that is also essentially one’s own, something that could not be done in conversation with someone else, but not for that reason infallible.

21In connection with the first of these three features, Murdoch writes:

I am now inclined to think that it is pointless, when confronted with the existentialist-behaviorist picture of the mind, to go on endlessly fretting about the identification of particular inner events, and attempting to defend an account of M as ‘active’ by producing, as it were, a series of indubitably objective little things. ‘Not a report’ need not entail ‘not an activity’. (Murdoch 1970, 23)

22Murdoch’s point here is that her example of M is not meant to recuperate the intelligibility of some version of the idea of inner data. But the fact that there are no data to report about does not exclude that there is activity inside. Her point is that inner life is something we engage in, not something we witness; what goes on inside does not go on independently of our awareness of it, but it is precisely this idea of ‘going on independently of our awareness’ that is built in the characterisation of the alleged inner data as ‘objective’, and as something one can ‘report’ about. Murdoch explicitly rejects an understanding of inner activity as reducible to a complex of such data by drawing attention to the subject’s will’s involvement in it. The example of M is helpful in this connection because in M’s case such involvement takes the shape of a deliberate effort, but what matters is the involvement itself and not the shape it takes. Consideration of more ordinary episodes of inner life can equally serve the purpose of showing the inadequacy of the imagery of the walled garden.

23It is worth noting that, at this stage of the discussion, the mental concept raising a problem for the genetic analysis of its meaning is ‘change of mind’. The concepts that M applies in the course of her activity are not yet relevant for the argumentation. It is the concept of change of mind that the example shows to be applicable in the absence of any observable circumstances without its instantiations’ turning into mythical inner objects. We do not fall into any mythical conception if we acknowledge that mental states and processes, as McDowell put it, ‘have no being independently of the fact that the concepts they instantiate figure in the content of consciousness’ (McDowell 1998, 311). All there is to M’s knowing that she has changed her mind is precisely her having done so, which is something that one does consciously. Parallel considerations hold for the example of R: it is the concept of self-examination (not the concepts deployed in such examination) that the contrast between reportables and activity shows to be applicable without the obtaining of any particular observable circumstance.

24Wittgenstein himself makes room for what might be broadly called the same sort of happenings in the mind in connection with applications of mental concepts that have nothing specifically moral about them. For instance, in Philosophical Investigation §662, he writes about someone who beckons to another person, called N., in the context of a series of passages that exploit the idea of appropriately used signals as a model for applications of mental concepts. The model is meant to emphasise the internal relations between the instantiation of the mental concept and its verbal expression—relations that might be obscured by thinking of such a verbal expression as a report. Wittgenstein writes, ‘One can now say that the words ‘I wanted N. to come to me’ describe the state of my mind at that time; and again one may not say so’. The reason why one might not say so is precisely the risk that these words encourage the idea of the problematic independence of the mental occurrence from one’s awareness of it. Of course, this raises the issue that, if the instantiations of mental concepts are not independent of the subject’s awareness of them, there is no room for the subject to make a distinction between being correct and seeming to be correct in her first-personal uses of these concepts. This is, for my purposes, the most important point and I will come back to it at the end of the next section, in my discussion of the third feature of the examples used by Murdoch.

25The second element of Murdoch’s characterisation of inner life (which is related to the first) is that if we drop the idea that the only thing inner life could be is some sort of parade of inner items for our inner eye to contemplate, the inner scene turns out to be way less elusive than that model leads us to think. There is nothing hazy (dubious) in our ordinary self-ascriptions of thoughts, intentions, wantings and the like, they simply are exercises of our capacity to apply these concepts. Murdoch brings this out by drawing our attention to how familiar and non-mysterious an activity like M’s is, and how similar it is to the sort of activity that is often described in novels and that we find completely intelligible (Murdoch 1970, 22; see also Murdoch 1951, 30–31).

26Once again, we find this point expressed by Wittgenstein in connection with more mundane instances of inner life, like remembering that one would have finished a sentence in a certain way, had she not been interrupted. I follow McDowell (McDowell 2009, 88–90) in reading PI §633-635 as showing that the task of completing what one was going to say before an interruption seems to be one of ‘reading the darkness’, in Wittgenstein’s view, only when one looks for elements that would explain her claim that she would have indeed finished the sentence in a particular way, rather than taking her to be exercising her competence in the relevant practice—that is, to be simply saying that such and so is, in fact, what she had meant to say. An appeal to that intention would not count as individuating a reality that justifies the way we talk; it would be just one more instance of talking that way, and therefore would not count as an explanation of the kind sought for in the rejected interpretation of the task. But the need for such an explanation will not be felt (or even thought of as intelligible) once the character of mental concepts is properly understood.

27What emerges from Wittgenstein’s discussion of the problems with the mythical conception of inner objects is not that there really are no mental states and processes after all, i.e., that so-called mental states are really nothing but shadows of publicly observable behaviors. Rather, the lesson to be drawn is that, when we use mental verbs in the first person, we are neither latching onto ‘indubitably objective little things’ nor surreptitiously referring to patterns of external circumstances. A proper understanding of what a mental concept is reveals that there need be nothing in the correctness of a first-personal use of a mental concept apart from its subject’s competence in the practice in which that concept has a role. This however, clearly does not imply that the subject has to wait and observe her own subsequent behavior in order to attribute to herself, for instance, a determinate intention (as the shadow view would have it). A subject knows her own intentions, in their full determinacy, by forming them. Her competence with the use of the concept of intention does not contribute to identifying the intention by being exercised in overt behavior, but in being presupposed in the very self-ascription of that intention.

28Here I need to note an important difference between the examples of inner life exploited by Murdoch and the just mentioned more mundane examples used by Wittgenstein and by McDowell in interpreting him. M’s activity, as Murdoch emphasises, is of a familiar kind, but it is difficult. Granted, it is not difficult in a way that has to do with the weird attempt to register, or retrieve from memory, ‘purely inner data’, and yet it involves more than basic competence in the linguistic practices involving the concepts she uses in it. As Murdoch says, ‘M’s activity is hard to characterise not because it is hazy, but precisely because it is moral’ (Murdoch 1970, 22). I’ll explain what I take Murdoch to mean by this claim in my discussion of the last of my three points.

Idiosyncrasy and Progress

29The third feature of the inner activity illustrated by Murdoch’s examples of M and D and of R is, as anticipated above, that it is something that is essentially one’s own, something that could not be done in conversation with someone else, but not for that reason infallible.

30It is with this feature that we arrive at a notion of privacy different from the one that the picture targeted by Wittgenstein tries and fails to make sense of, but also from that connected with the mere interiority of mental states, i.e. with the claim that it is not silent and dark within. The idea that this further kind of privacy is not only intelligible but also of great importance is a feature of Murdoch’s attack on the genetic analysis of mental concepts which differentiates it from McDowell’s position on privacy.

31In ‘Intentionality and Interiority in Wittgenstein’, as we have already recalled, McDowell articulates the thought that a non-mythical conception of inner life requires us to recognise that its episodes do not take place independently of their subject’s awareness of them. Immediately after he goes on to address the worry that acknowledging this dependence removes the possibility of a distinction between being right and merely seeming to be right with respect to whether such episodes occur, and, therefore, might be perceived as impugning the reality of mental states of affairs.

32McDowell retrieves the distinction between ‘being right’ and ‘seeming to be right’ with respect to the obtaining of mental states by appealing to (1) publicly accessible circumstances that help to fix the meaning of the concepts and (2) the thinkability of such mental states in the second- and third-person perspective. But Murdoch’s objective is to show the need to radically rethink the role of both these resources in the light of the existence of what she calls ‘specialised uses of concepts’.

33The sort of attack on the genetic analysis that we find in McDowell rejects the idea that our states of mind have a parasitical (or ‘shadowy’) existence. In his view, there is no need for any publicly observable performance to obtain for a mental state to be intelligibly and determinately present in our interiority. This possibility, however, according to McDowell, does depend on something public, namely, the subject of this state’s competence in the practice in which the corresponding concept has a role. This competence is all that is needed to make one’s self-ascriptions of this state authoritative. McDowell uses the example of the intention to type a period. Certainly, he argues, this intention could be fully formed in a subject and could be correctly self-ascribed independently of any subsequent performance. But the possibility of forming it depends on the fact that she is ‘party to the practices that are constitutive of the relevant concepts’ (McDowell 1998, 315). In this sense, the unactualised intention is a mental state whose existence is not parasitical and yet it is not mythical precisely because of its connection with certain public practices which include the ascription of intentions to others.

34This is all perfectly satisfying with respect to mental states such as the intention to type a period and can even be said to hold across the board. The link between a subject’s first-personal uses of certain mental concepts and her competence in the practice of applying them in other ways is required if her inner world is to be part of the one world there is. But Murdoch’s examples present us with reasons for thinking that the story can’t end here.

35She draws our attention to the fact that the activity of reassessing and redefining one’s mental concepts and the way one applies them ‘often suggests and demands a checking procedure which is a function of an individual history’ (Murdoch 1970, 25). In other words, this checking procedure involves precisely those aspects of our thought that go beyond what we share with most other people and therefore constitute our competence in using concepts. The criteria of correctness are a function of an individual history, unsurprisingly, for any use of a concept which has been shaped by our personal experiences and has become ‘specialised’ in Murdoch’s sense of the term.

36Certainly, the very idea of ‘specialisation’ presupposes a basis one builds upon. A specialised use of a concept, in Murdoch’s conception, can be highly personal but, if it makes sense to call it a specialised use of the concept of, say, repentance, such use must come from someone who is competent in its ordinary and conventional use. The independence from public criteria of meaning that the idiosyncrasy of specialised uses of concepts involves, then, does not entail that such criteria are dispensed with. It entails that they are insufficient to settle certain important questions about the proper application of such concepts on certain occasions and that second and third-personal uses of such concepts, even if thinkable, might not be really ever at stake. This means that, with respect to specialised uses of concepts the resources invoked by McDowell are not helpful.

37The examples of M and D and the example of R, then, exhibit a kind of privacy that does not derive from mere interiority (i.e. from a feature shared also by unexecuted and un-manifested intentions to type a period). The activity exemplified by M and by R exhibits a privacy that attaches not to the concepts of change of mind, in the case of M, and of self-scrutiny, in the case of R, but to the concepts that each of the two subjects is trying to apply properly in a sense that requires more than mere ordinary competence. Ordinary competence is part of what each character needs in order to even engage in his or her respective task, but it is far from sufficient to bring it about.

38That concepts lend themselves to specialised uses is a very familiar fact of life. Even if we do acquire our concepts in public contexts, some of them, and particularly mental ones, characteristically undergo a process of transformation. And here we finally reach the connection between the idiosyncrasy of specialised uses of concepts and moral progress involving the idea of perfection.

39The way in which we understand concepts—paradigmatically concepts that like courage and repentance articulate our understanding of ourselves and of other human beings—in fact, is not the same in different phases of life. We are prompted or pressured to change our concepts, or the way we apply them, by the specific objects of attention that are our own. In fact, there is something wrong with someone who does not go through this process of reassessing. Murdoch calls this process ‘the main characteristic of live personality’ (Murdoch 1970, 25, emphasis added) with the implication that there is something ‘dead’ in a person who is unwilling or unable to alter his conceptual repertoire in this way.

40Of course, the transformation of our concepts can be a case of degeneration and corruption, and I will come back to this problem. But, for now, the point is that the historical nature of human individuals brings in a perfectly intelligible and non-mysterious sense in which our inner lives, where these transformations take place, are private. The acknowledgment of the idiosyncrasy of some of our concepts entails no more than the thought that what one really means in applying them can only be understood under certain conditions.

Human beings are obscure to each other, in certain respects which are particularly relevant to morality, unless they are mutual objects of attention or have common objects of attention, since this affects the degree of elaboration of a common vocabulary. (Murdoch 1970, 32)

41Our inner lives, then, are often exclusively our own not because our mental states are inexpressible in principle. In the passage just quoted, Murdoch even gives a characterisation of the circumstances in which the elaboration of a common vocabulary is possible. But to say that this communicability is possible in principle should not, so to speak, domesticate the point. The fact that a shared world is only a remote ideal is a structural aspect of our condition. Murdoch writes:

that ‘reality’ that we are so naturally led to think of as what is revealed by just ‘attention’, can of course, given the variety of human personality and situation, only be thought as ‘one’, as a single object for all men, in some very remote and ideal sense. (Murdoch 1970, 37)

42In this passage, we find not only an important qualification to the realist idea that one’s inner world is part of everyone else’s world, but also (encapsulated in the appeal to ‘just attention’) an indication on the different standard for good application proper to specialised uses of concepts which is. Murdoch’s acknowledgement of the idiosyncrasy of some concepts, in fact, does not amount to saying that they don’t need a checking procedure for their application, but to reminding us that not all concept applications can sensibly be assessed in the same way.

Imagination and Fantasy

43The question, then, is what makes a specialised use of a concept correct. The short answer is: its capacity to disclose reality to the subject.

44This admittedly tautological answer becomes more informative if we consider, so to speak, two poles, two points that set opposite directions for the development of our concepts. Murdoch argues that our uses of concepts can become more personal and idiosyncratic both as a result of our letting ourselves get trapped in fantasies and as a result of our making strides. ‘Fantasy’ and ‘imagination’ name, respectively, the poles that set the directions of degeneration and the direction of improvement of our tools to access reality: concepts. An alternative conceptual pair that Murdoch uses to identify these two poles is constituted by selfishness and love or justice. Both poles in these pairs draw our concepts away from the shared core that makes ordinary everyday communication possible. In other words, both progress and degeneration constitute forms of specialisation.

45In this connection, it is very interesting that Murdoch chooses, in order to characterise M’s idea of D, at the beginning and the end of the story, respectively, concepts whose public criteria of application might easily end up being the same (‘tiresomely juvenile’/‘delightfully youthful’; ‘vulgar’/‘refreshingly simple’). What M needs in order to face her predicament is not a better command of these criteria, but an imaginative effort enabling her to elaborate concepts adequate to the uniqueness of her situation. Murdoch portrays M as able to identify specific obstacles in addressing this task. M considers the possibility that she is ‘conventional’ and ‘old-fashioned’ and admits to herself that she is jealous (a selfish emotion). The opportunity for change is provided, in the example, by M’s newly found capacity to consider the otherness of D. It is by turning her focus away from her own habits, needs, and wishes and to D that M grows and changes for the better.

46It might be said that R, whose situation Murdoch does not describe in detail, faces the same sort of task. Clearly R could not settle his question by revising the public criteria for the application of the concept of repentance, for example by consulting a dictionary. Like M, R needs to attend to the specificity of his situation (what he has done, his responsibilities, etc.) and to do so he has to turn away from his own obsessions and fears as well as engage in an effort of understanding of his own situation rather than falling into automatic responses. In other words, R’s task can also be described as that of overcoming selfishness and convention.

47The identification, on Murdoch’s part, of these two particular obstacles, selfishness and convention, as the two main sources of our failure to see reality for what it is depends on her conception of reality as infinitely elusive and on a broadly Freudian conception of human psyche. If reality is, as it seems to be, infinitely elusive, our conceptual repertoires need to be constantly altering if it is to conquer it even only locally and temporarily. Selfishness and convention, on the other hand, are, respectively, a force that derails this process of alteration by making it dominated by fantasies and a force that blocks it altogether. If we are, as a broadly Freudian picture has it, extremely vulnerable to them, these forces would be exactly what we need to fight in order to reach reality. This is what striving toward perfection would amount to. But should we accept this conception of our predicament or, as Murdoch calls it, this ‘picture of the soul’? Murdoch does not offer any argument to establish her favorite picture of the soul.

48This is not a matter of failing to bring the work to completion, but of challenging the idea that moral thought, including its philosophical variety, is in the business of establishing claims rather than empowering us in its characteristic way. She writes:

[T]he sketch which I have offered, a footnote in a great and familiar philosophical tradition, must be judged by its power to connect, to illuminate, to explain, and to make new and fruitful places for reflection. (Murdoch 1970, 44)

Haut de page


McDowell, John, ‘Intentionality and Interiority in Wittgenstein’, Mind, Value, and Reality, John McDowell, Cambridge, Mass: Harvard UP, 1998, 297–321.

McDowell, John, ‘Are Meaning, Understanding, etc., Definite States?’ The Engaged Intellect: Philosophical Essays, John McDowell, Cambridge, Mass: Harvard UP, 2009, 79–95.

Merritt, Melissa, ‘Love, Respect, and Individuals: Murdoch as a Guide to Kantian Ethics’, European Journal of Philosophy 25 (2017): 1844–1863.

Murdoch, Iris, ‘Thinking and Language’, Proceedings of the Aristotelian Society, Supplementary Volumes, Vol. 25, Freedom, Language, and Reality, 1951, 25–82

Murdoch, Iris, The Sovereignty of Good, London: Routledge, 1970.

Moore, George Edward, Principia Ethica, Cambridge: CUP, 1903.

Setiya, Kieran, ‘Murdoch on the Sovereignty of Good’, Philosophers Imprint 13.9 (2013): 1–21.

Wittgenstein, Ludwig, Philosophical Investigations, trans. G.E.M. Anscombe, Oxford: Blackwell, 1953.

Wright, Crispin, ‘Wittgenstein’s Later Philosophy of mind: Sensation, Privacy, and Intention’, Journal of Philosophy 86.11 (1989): 622–634.

Haut de page


1 An example of this tendency is provided by Setiya 2013. Although he thinks that the tension between the commitment to the idiosyncratic character of concepts and realism is only apparent, he also considers that commitment marginal and inessential to Murdoch’s picture of the mind. An exception to this general tendency is Merritt 2017, who explicitly acknowledges the distinctiveness and importance of the commitment in question.

2 The way of appropriating Wittgenstein’s work on privacy which I address in this paper is linked by Murdoch to the attitude of her contemporaries toward G.E. Moore’s views in the Principia Ethica. According to Murdoch, in fact, the generalised rejection of Moore’s conception of the good as an object of knowledge in favor of one that takes it to be a function of the will is the metaphysical counterpart of the moral psychology deriving from the genetic analysis (for this idea see Murdoch 1970, 4). In this paper I focus almost exclusively on the moral-psychological side of the matter, but the metaphysical side is implicitly touched whenever the endless perfectibility of our concepts is represented as a consequence of the infinite elusiveness of the reality we confront. For Murdoch the ideal limit of the process of improvement of our concepts does not inhabit a separated metaphysical dimension: it is nothing but that reality we inhabit, only recognised in its inexhaustible complexity. This conception of the ideal limit of conceptual progress is an important feature of Murdoch’s Platonism. Though very pervasive, this aspect of her thought will also remain mostly implicit in my paper given its moral-psychological focus.

3 Murdoch reiterates frequently that, in her view, the inadequacy of the genetic analysis is brought out by the idea of perfection and by the conception of human beings as historical individuals (see esp. Murdoch 1970, 23–28).

4 In this section and the following ones, I draw on McDowell 1998 and McDowell 2009 for the interpretation of Wittgenstein’s treatment of privacy in the Philosophical Investigations. I consider McDowell’s reading of this much debated theme in Wittgenstein’s later philosophy to be persuasive and well supported by textual evidence, but my goal here is not to assess its merits as an interpretation of Wittgenstein. What I am interested in is, rather, the difference between the understanding of the privacy of mental life that McDowell attributes to Wittgenstein and endorses himself and Murdoch’s understanding of topic.

Haut de page

Pour citer cet article

Référence électronique

Annalisa Paese, « Interiority and Idiosyncrasy, Two Kinds of Privacy in Murdoch’s Conception of Mental Life »Études britanniques contemporaines [En ligne], 59 | 2020, mis en ligne le 01 septembre 2020, consulté le 28 novembre 2020. URL :; DOI:

Haut de page


Annalisa Paese

Annalisa Paese obtained a PhD in Philosophy in 2018 from the University of Pittsburgh with the dissertation The Human, Love, and the Inner Life: Ethics after Murdoch. She arrived at the University of Pittsburgh after completing a Doctorate degree at Sapienza—University of Rome with the thesis Philosophy of Action and Practical Conceptual Content. She is currently a visiting scholar at the Higher School of Economics (Moscow). Her main research project, at the moment, is an investigation of the moral uses of the concept of a human being through the philosophical analysis of interpersonal conduct naturally described as inhumane.

Haut de page

Droits d’auteur

Licence Creative Commons
Études britanniques contemporaines est mise à disposition selon les termes de la Licence Creative Commons Attribution - Pas d'Utilisation Commerciale - Pas de Modification 4.0 International.

Haut de page
  • Logo Presses universitaires de la Méditerranée
  • Logo ERIH +
  • Logo DOAJ - Directory of Open Access Journals
  • OpenEdition Journals
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search