Emancipatory Attention

2.1 The Entanglement of Epistemology and Ethics

Because the status quo tends to be reinforced by mechanisms of wishful thinking and confirmation bias (Darley & Gross 1983; Gendler 2011), and because the deepest sources of oppression tend to conceal themselves (Fried 1979; Lugones 1990), we rarely have a clear view of the injustices that form the background of our lives. Even when these injustices come into the foreground, our perspective on them is subject to misleading psychological and social influences—some of which are malicious, and others inadvertent (Gilbert, Krull, & Malone 1990; Rapp & Salovich 2018). Attempts to promote justice are therefore at risk, unless they include epistemic work, owing to the possibility that matters may be seen through a distorting lens.

Such epistemic work is not a preliminary, to be completed before the moral work begins. The epistemic and moral parts of our emancipatory projects should instead be pursued in tandem. There are several ways in which they can then be mutually reinforcing. We can see a first example of this by considering a non-conforming teenager, whose lack of role models leaves them oblivious to the possibilities for flourishing that lie outside the norms of gender that are prevalent in their cultural context. The injustice of their position is thoroughly entangled with its epistemological disadvantages. In addressing either, one would thereby be addressing the other. For a more classic illustration of such entanglement, we might look to Adam Smith (1776) or to Karl Marx (1867): both attribute the poor epistemic position of the exploited laborer to the working conditions that “corrupt the courage of his mind” (Smith 1776, V.1.3.2). A more contemporary example could be found by considering the role played in projects of transitional justice by the establishment of an official public record, as in the Truth and Reconciliation Commissions of Australia, Canada, Chile, and South Africa. Here, the same quasi-legal processes perform both ethical and epistemic functions. The testimonials gathered in the course of those processes illustrate some of the cultural ways in which epistemology and ethics interact, by showing how extrication from a colonial mindset can depend on rejecting conceptions of knowledge-transmission that prevent indigenous peoples from sustaining their cultures, and from becoming knowledgeably aware of the brutality with which those cultures have been suppressed.

More will be said later about each of these three examples, but even the brief descriptions given above suggest that each is a case in which the epistemic consequences of an injustice play a role in the mechanisms by which that injustice becomes entrenched. Each is a case in which there is no practical distinction to be drawn between the epistemic project of education and the ethical project of emancipation. These projects are most closely entwined where injustices have been perpetrated by educational institutions—as in the imposition of oralism in the education of Deaf people (Ladd 2003), and the residential schooling of children from Canada’s First Nations (Cook 2018)—but the diversity of our examples suggests that the underlying point is a general one: even when an injustice is not primarily pedagogical, progress toward its mitigation may necessitate improvements to the epistemic position of oneself and others.

2.2 The Knowledge-centered Conception

It might seem natural to suppose that these epistemic improvements must be constituted by increases in the supply and distribution of knowledge, and so to suppose that the epistemic aspects of injustice are to be resolved by ensuring that all of the relevant people come to know all of the relevant facts. That supposition might seem to follow analytically from the characterization of these problems as epistemic ones. Their solution will then seem to be a matter of bringing certain true propositions to light, so as to ensure that they find a place on each person’s list of justified beliefs. This, however, should arouse our suspicions.

On any knowledge-centered approach to the epistemic aspects of injustice, the application of epistemology’s proprietary norms will seem more or less obligatory. If our problem is a lack of knowledge then its solution will require the establishment of justified beliefs in the minds of those who are, as yet, ignorant. The justifications that endow these beliefs with their epistemic credentials will need to be recognized as reason-giving by some as-yet-ignorant groups or individuals, and so the reasons that are invoked will need to be objective (or will, at least, need to be reasons with a sufficiently broad inter-subjective force). They will need to figure in reasoning that employs a logic the validity of which these as-yet-ignorant groups can recognize. Or, where the propositions needing to be known are not justified through sequential reasoning, but instead come from some immediate first-person experience, the ignorant groups who are being informed will need to defer to the authority of those with the expertise that comes from having such experiences. On this conception, efforts to address the epistemic aspects of injustice will therefore need to endorse the ideals of an objective, logical, and deferential dialectic.

Several complaints have been raised about the rhetoric surrounding those ideals. Feminist philosophers, post-colonial philosophers, and philosophers of disability have all identified ways in which oppression can be perpetuated through what purports to be the epistemic appraisal of objectivity, authority, or expertise (MacKinnon 1989; Code 1991; Anderson 1995; Fricker 1998; Zuckert 2019). Some have proposed that, if it is to avoid exacerbating oppression, the conceptual scheme of epistemology needs fundamental adjustment (Shotwell 2011; Fassio & McKenna 2015; Berenstain et al. 2021). The most radical of these proposals would abandon universalizable norms of rationality altogether (Barnes & Bloor 1982; Appleyard 1992), and would therefore be antithetical to our present enquiry, but we can acknowledge the absoluteness of our own commitment to rational ideals while recognizing that it would be a mistake to suppose that our philosophical discussions should be taken as a model for the epistemic struggles that are their subject matter. To adopt such a model would be to idealize in the way that has been deprecated as “ideal theory”—especially by Mills (2005): it would risk construing our epistemic problems as ones that should require nothing more than a good faith debate between well-intentioned inquirers who defer to one another’s expertise in matters beyond their own experience. This would obscure the conditions of injustice that prevent encounters across social divides from displaying the minimal features of such a dialectic.

2.3 Consciousness-raising

One might hope to respond to this worry about the dangers of “ideal theory”, not by giving up on the idea that the epistemic aspects of our emancipatory projects are to be addressed through the propagation of knowledge, but only by giving up on the idea that that propagation should proceed through rational argumentation. This would be to follow the lead of those twentieth-century activists who saw their epistemic problems as calling for something that is less like a dialectic, and more like a campaign of consciousness-raising. But if this response retains the idea that our epistemic problems are primarily problems of knowledge, then the position that results is one of unwelcome pessimism.

To see why that pessimism comes to seem unavoidable notice that, even when ideals of knowledge transfer through rational dialectic are suspended, this version of the knowledge-centered conception requires a distinction to be drawn between those cases in which the subjects of the to-be-addressed ignorance are subjugated, and those in which they are privileged. Rather than calling for a single dialogue between these groups, our emancipatory projects will now seem to call for two distinct tranches of epistemic work: one which raises the consciousness of the oppressed, and thereby alleviates the ignorance that serves to reinforce their oppression, and another which raises the consciousness of the oppressors, and thereby ensures that they come to know about the extent of their privileges, and the obligations created by them. The two phases of this work will be quite different.

Before considering the difficulties created by this difference, first notice its magnitude. Although there may be some overlap in the bodies of information that need to be communicated to these two groups, that overlap is unlikely to be large. Those in positions of privilege may need to be told about the plight of the oppressed, but (as standpoint epistemologists have often noted) the oppressed are themselves likely to be only too well aware of it. Oppressed groups may need to be informed about the means of exercising their legal rights, but there will be no need to communicate this to those who are sufficiently privileged to take such rights for granted. And even when the same information needs to be conveyed to both groups, it may need to be heard in different tones of voice. A reproof may be appropriate when reminding the privileged of their complicity in an oppressive system—without this there may be a tendency to gloat or to indulge in feelings of self-congratulation—but those who are exploited by a system are not to be reproached for the fact that their own labor is used in reinforcing it. Similarly, these oppressed groups may need to be exhorted to organize in solidarity, while the privileged need no such encouragement. The two tranches of consciousness-raising work therefore differ in their targets, content, and tone.

One of the problems created by this can be seen by noticing that the separation between these two phases of consciousness-raising work will tend to reinforce the social divisions that are created by the very forms of injustice that we are trying to resist. That problem is exacerbated by the differential burdens that tend to be created when this approach tries to avoid the risks of being patronizing and being presumptuous. To see why, it will be helpful to consider these risks in turn.

The first risk is sufficiently familiar to be popularly deprecated, in the context of race, under the trope of “white saviorism.” One might hope to avoid it by ensuring that oppressed groups are themselves empowered to take responsibility for identifying the knowledge that would be liberating for them, so that what is called for is not a top-down transfer of information from a patronizing elite to their patronized victims, but is instead the creation of opportunities for members of variously oppressed groups to share and synthesize the knowledge that they already have. Some exemplary consciousness-raising projects fit this model—their apotheosis being Paolo Freire’s “Pedagogy of the Oppressed” (Freire 1970)—but they avoid patronage only because the oppressed themselves must capitalize on these opportunities, with the result that it is they who must do the lion’s share of this phase of epistemic work.

By turning to consider the other phase of the required epistemic work—raising the consciousness of the privileged—we see that this burden is not offset elsewhere. The salient risk now is not that of being patronizing, but that of being presumptuous. Again it is familiar enough to have become the subject of a popular slogan, which can be heard when the representatives of marginalized groups insist that political decision-making should include “nothing about us without us.” Avoiding this risk again leads to a significant part of the epistemic burden being carried by those who suffer from the injustice in question. Just as it would be patronizing for those who are privileged to arrogate responsibility for informing the oppressed about their oppression, so it would be presumptuous for them to suppose that they are in a position to identify the information of which they themselves need to be apprised, unless when doing so they draw continually on the testimonial work of those who have first-hand experience of the inequities that this privilege creates. Efforts to raise the consciousness of the privileged therefore depend on work that can be done only by those who suffer as a result of their privilege. Again the burden of epistemic work shifts onto those who are already disadvantaged.

If we adopt a knowledge-focused conception of the epistemic work that is demanded by our emancipatory projects, while trying to avoid the idealization of construing this as work for a rational dialectic, then we have seen that we arrive at a position in which there are two separate tranches of epistemic work to be done, in both of which a disproportionate burden needs to be carried by those who are already at a disadvantage. The risks that create this situation are not hypothetical. We have seen that they are sufficiently conspicuous to be represented in slogans from the grassroots discourse. Their reality has been well documented in the literature of second wave feminism.

2.4 Frye on Consciousness-raising and Pessimism

Marilyn Frye gives a frank account of her own experiences in negotiating these risks. In The Politics of Reality (1983), Frye discusses her participation, in the 1970s, in groups that aimed to address white ignorance of ways in which sexism and racism intersect. She writes that “one can, and should, educate oneself and overcome the terrible limitations imposed by the abysmal ignorance inherent in racism” (Frye 1983, 118). But she also recounts the criticisms that were prompted when her attempts at such education included the formation of:

a white women’s consciousness raising group to identify and explore the racism in our lives with a view to dismantling the barriers that blocked our understanding and action in this matter. (Frye 1983, 111)

Frye becomes painfully aware of the presumptuousness that is risked in such an attempt. She tells us that “one Black woman criticized us very angrily for ever thinking that we could achieve our goals by working only with white women” (ibid.). She describes some of the ways in which attempts to avoid such risks led to the burden of epistemic responsibility being shifted insidiously onto women of color:

I was at a poetry reading by the Black lesbian feminist, Audre Lorde. In her poems she invoked African goddesses, naming several of them. After the reading a white woman rose to speak. She said first that she was very ignorant of African religious and cultural history, and then she asked the poet to spell the names of these goddesses and to tell her where she might look for their stories. The poet replied by telling her that there is a bibliography in the back of the book from which she was reading which would provide the relevant information. The white woman did not thank the poet and sit down. The white woman (who I know is literate) said, “I see, but will you spell their names for me?” What I saw was a white woman committed to her ignorance and being stubborn in its defense. She would convince herself that she cannot use this bibliography if the Black woman will not spell the names for her. She will say she tried to repair her ignorance but the poet would not cooperate. The poet. The Black woman poet who troubled herself to include a bibliography in her book of poems. (Frye 1983, 119)

Frye finally suggests that these epistemic-cum-ethical problems are, in part, intractable—“In a certain way it is true that being white-skinned means that everything I do will be wrong” (126)—and that our efforts to address them will run into paralyzing difficulties: “We began to lose hope; we felt bewildered and trapped” (112). Similar sentiments are expressed in the less scholarly feminist literature of the time (see, for example, Philpott 1980/1982, 585.)

When additional axes of injustice are taken into account, this sense of hopelessness threatens to become overwhelming. If we appreciate the magnitude of Frye’s own achievements, in her discussions of the invisibility of lesbianism, or of Mills’ achievement, in his demonstrations of the extent of white ignorance, we must realize that there are no grounds to suppose that such work will have been done so well elsewhere. Nor is this work that can ever be brought to completion. The ignorance that has been documented by these theorists has innumerable counterparts. We sometimes have an inkling of these, as when we consider the ignorance of gender-conforming people concerning those who are trans-gendered, when we consider neurotypical ignorance of those who are neurodivergent, or when we consider hearing ignorance of those who are deaf. But we must suppose that there are other occasions where our ignorance is so profound that we lack the vocabulary with which to identify it.

In each case it is likely to include ignorance of facts about the character of a first-person experience. Knowledge of such facts is hard to acquire for those who lack that experience (for reasons that have been much discussed by philosophers of mind, with reference to ethically uncomplicated cases, following Jackson 1982). Rectifying this ignorance requires something like an imaginative and sympathetic reading, rather than a third-person perspective on publicly accessible facts. Consciousness-raising will therefore need to wait until the experience in question has been made legible, through being documented with sufficient honesty and skill. And here we get our clearest view of the position that we are at risk of falling into, if we take the epistemic aspects of injustice to arise primarily from a lack of knowledge. Although we may be able to avoid the problems of presumptuousness and patronage that dogged late-twentieth-century projects of consciousness-raising, and although we may be able to avoid a culpably idealized view of what might be accomplished through rational dialectic, we thereby put ourselves at risk of arriving at a position where it is the oppressed who must carry the burden of making injustices legible, and where the epistemic responsibility of the privileged is merely to serve as an audience for memoirs of increasingly recherché victimhood. Even when this audience manages to avoid voyeurism, the appreciation of others’ suffering risks becoming an aesthetic taste, with no aptitude for being mobilized in emancipatory action.

2.5 From Knowledge-promotion to Attention-facilitation

Although Frye suggests that this position can appear hopeless, she is careful to indicate that this appearance might also be taken as indicating that our epistemic problems call for a more radical approach. Rather than attempting to alleviate ignorance by working to create knowledge, she suggests that we might approach the epistemic aspects of injustice more fundamentally, by addressing the complex of underlying conditions from which ignorance arises, and by which it is maintained:

Some of my experience has made me feel trapped and set up […]. Clearly, if one wants to extricate oneself from such a fate or (if the feeling is deceptive) from such a feeling of fatedness, the first rule for the procedure can only be: educate yourself.

While one is educating oneself […] one should also be studying one’s own ignorance. Ignorance is not something simple [… It] is a complex result of many acts and negligences. (Frye 1983, 118).

Among these complex ignorance-generating conditions, Frye suggests that attention features centrally (together with the ‘negligences’ that are its contrary):

If one wonders at the mechanisms of ignorance, at how a person can be right there and see and hear, and yet not know, one of the answers lies with the matter of attention. (Frye 1983, 120)

We do much of what we do with a great anxiety for how we will be received by men […]. With our attention focused on these men, or our imaginings of them, we cannot pay attention to the matter at hand and will wind up ignorant of things which were perfectly apparent. […] Attention has everything to do with knowledge. (121)

The last of these remarks makes it clear that Frye is not proposing that we turn away from knowledge in turning toward attention. Nor is that the proposal of the present paper. But if the turn toward attention is to provide the shift in perspective that enables a sense of hopelessness to be avoided, then the importance of attention cannot here be derived wholly from the knowledge that it facilitates. Although “attention has everything to do with knowledge,” it must also be recognized as a source of epistemic value in its own right. The risks of patronage and presumptuousness recede only if we take some of our epistemic problems to be constituted primarily by failures of attention, without reference to the knowledge that might result.

One who takes it that they have knowledge of whether P cannot sincerely attempt to learn whether P. Efforts to create knowledge are therefore implicated in presuppositions about ignorance (even when they concern very general questions, such as “How are things with X?”). When this ignorance is our own it puts us at risk of presumptuousness. But if our efforts are instead directed at rectifying failures of attention then this risk recedes: whereas the formulation of a question commits one to looking for an answer with a certain form, attempts to increase one’s attentiveness can be combined with a maximal degree of open-mindedness about the things that might be found. Such attempts need not entail a presumption about one’s prior epistemic standing.

We are, similarly, at risk of being patronizing when making assumptions about the ignorance of others, but the assumptions required by efforts to facilitate their attention are less insidious. If I inform someone of facts that they already know then I may be guilty of condescension (“’splaining”), but if I remove distractions from the foci on which their attention is already directed then I am not committing any similar offence. The person who is on the receiving end of attempts to provide them with knowledge they already have cannot join the conversation on its own terms. They must first cancel a presupposition. When a person’s attention is directed toward some topic on which they already have a standpoint, there need not be any such barrier to entering the conversation.

Shifting from a knowledge-centered conception of our epistemic work to an attention-centered conception also mitigates the risk of unhelpful idealization. The knowledge-centered conception risked “ideal theorizing” because the assessment of a person’s knowledge was a normative assessment, requiring us to assess the state of play in a game of giving and asking for justificatory reasons (Lewis 1979). Pervasive injustices have the consequence that this game is played in such a way that those without a position of privilege are at a disadvantage in it. Operating within its norms therefore opened us to the charge of ideal theorizing. The assessment of a person as attentive may still be normative, but the norms against which this assessment is made can avoid entrenching the advantages of the privileged because the question of whether one is attentive can instead be a question about how successfully one is performing by one’s own lights. This is emphasized in at least one theory of the psychological basis of attention, where it appears as a claim about the “cognitive resources” that must contribute to any given instance of attention:

The agent’s current understanding is what sets the standards of relevance to the task that a resource must meet in order for that resource to have a role in determining whether the task is attended. (Mole 2011, 61; original emphasis)

No agent’s understanding of her own tasks is immune from pernicious influences, and so the risk of enforcing oppressive norms is not completely avoided by a move from appraisals of knowledge to appraisals of attentiveness. That shift does nonetheless seem to be a step in the right direction. Whereas efforts to make the subjugated knowledgeable could seem to be insisting that they participate in a game that is played by someone else’s rules, efforts to promote attention will involve emancipation from those rules in cases where they have become a distraction from the pursuit of one’s own values.

Regarding systemic epistemic problems as problems of attention therefore promises to be a way in which we can avoid at least some of the difficulties that emerged in the course of attempts to treat those problems as if they were primarily problems of knowledge. But the move to an attention-based conception does need to be handled with care. Skeptical readers might accuse us of implying that injustices can be mitigated merely by staring at them. To avoid that implication, we need an account in which attention is not taken to be mere perceptual focalization, but is instead understood to be a locus of value.

In the hope of finding such an account, several philosophers have turned to the work of Iris Murdoch, and to the work of Simone Weil by which she was inspired (see, e.g., Bagnoli 2003, Mole 2007, Altorf 2008, Lovibond 2018, Driver 2020, Heiti 2021). The importance of that work is undeniable, but—because Murdoch was primarily concerned with particularized ethical demands, rather than with broader systemic injustices—her agenda is somewhat misaligned with the issues that we have been considering here. The next section gauges the extent of this misalignment. The concluding section suggests a remedy for it.

3.1 “The Idea of Perfection”

Although Murdoch’s The Sovereignty of Good (1970) is rich with insights about the value of attention, it should not be read as arguing for any claim about attention being particularly good. Had that been Murdoch’s intention, then she would have been guilty of a question-begging stipulation when writing:

I would like on the whole to use the word “attention” as a good word and use some more general term like “looking” as the neutral word. (Murdoch 1970, 36)

Instead, Murdoch was arguing for the moral importance of a broad class of mental phenomena: one that had become invisible to those of her contemporaries whose philosophy of mind was marked by a Wittgensteinian insistence on crediting the significance of our talk about the mind to the outward criteria by which mental phenomena are manifested. When Murdoch calls the most morally notable member of this class “attention,” she is using that word in a somewhat technical sense, which she tells us is “borrowed from Simone Weil” (Murdoch 1970, 33), but she is not aiming at the building of a technical theory. The Sovereignty of Good begins by suggesting that ethical inquiry calls for “a move back again towards the consideration of simple and obvious facts,” rather than “the building of elaborate theories” (1). Its first essay concludes by characterizing Murdoch’s task as:

the provision of rich and fertile conceptual schemes which help us to reflect upon and understand the nature of moral progress and moral failure and the reasons for the divergence of one moral temperament from another. (Murdoch 1970, 43)

Such remarks show Murdoch to have been conceiving of her project as an instance of what we would now call “conceptual engineering” (in the style of Haslanger 2000 or Cappelen 2018), rather than as a conceptual analysis. When we identify contexts in which her thinking does not apply, this should not be regarded as if it were the discovery of a counterexample to any such analysis. We are instead plotting the bounds within which Murdoch’s thinking operates, so as to show that our desired notion of emancipatory attention lies outside them.

Attention features most prominently in the first of the three essays that The Sovereignty of Good collects. In this essay—“The Idea of Perfection”—Murdoch’s argument is animated by a thought experiment that has been central to subsequent discussions (see, for example, Snow 2005, Vice 2007, Altorf 2008, Frederiksson & Panizza 2022). It concerns a mother-in-law who arrives, by an effort of “careful and just attention” (emphasis Murdoch’s), at a more favorable opinion of the daughter-in-law whom she had initially disliked, but who is now “discovered to be not vulgar but refreshingly simple, not undignified but spontaneous, not noisy but gay, not tiresomely juvenile but delightfully youthful, and so on” (Murdoch 1970, 17).

Murdoch intends this to be an example in which something of moral significance has clearly been accomplished. To remove any outward criteria to which her Wittgensteinian opponents could credit that significance, she stipulates that this mother-in-law “behaves beautifully to the girl throughout,” and that her daughter-in-law has emigrated or died, so that no future behaviors are in prospect (Murdoch 1970, 17). Murdoch therefore forces us to credit the mother-in-law’s moral accomplishment to her attention, just because there is nothing else to which it might be credited. She characterizes that accomplishment by relating her morally loaded notion of attention to “the idea of the individual and the idea of perfection” (27). The position that results is described as “suggesting that the central concept of morality is ‘the individual’ thought of as knowable by love, thought of in the light of the command, ‘Be ye therefore perfect’” (29). Murdoch tell us that she construes this command as requiring that “the gaze is directed upon the ideal” (30).

3.2 Murdoch’s Misalignment to the Social

We should be wary of applying Murdoch’s account to the forms of attention that are called for in the emancipatory projects that we have been considering. Whereas Murdoch focuses on “the idea of the individual” (Murdoch 1970, 27; my emphasis), those projects concern injustices that operate at an intergenerational level, are not perpetrated by or against individuals only, and which require more than individual action for their mitigation. These projects are cases in which the stakes are high, whereas the logic of Murdoch’s thought experiment requires her case to be one in which nothing other than attention is at stake. And we have seen that Murdoch gives a central place to “The Idea of Perfection,” whereas we have resolved to avoid indulging in “ideal theory.” By returning to our earlier examples, we can see why her thinking fails to provide the account of emancipatory attention that we have found ourselves to need.

The first of those examples concerned the social mores that conspire to prevent a non-conforming teenager from finding positive role models outside the norms of gender that are prevalent in their cultural context. Murdoch gives us the conceptual resources to see such mores as limiting our teenager’s scope for freedom. She writes, of her imagined mother-in-law, that “Freedom […] is a function of the progressive attempt to see a particular object clearly” (Murdoch 1970, 23). But our teenager’s problem is that social factors have hidden any of the “particular objects” at which such “progressive attempts” could be directed, so that the Murdochian advice to engage in these attempts finds no application. The sources of injustice operate more pervasively here than in the attitudes that lead the mother-in-law to her unjust assessment of her daughter-in-law. They cannot be addressed by an individual effort because our teenager’s problem is that they have been deprived of the opportunity to make such an effort.

When Murdoch’s account is applied to our other examples, further problems come to light. The second of our examples was the Marxist one of an exploited laborer. Again, this seems to be different in kind from any case to which one might fruitfully apply Murdoch’s ideas about the moral work that can be done by “a just and loving gaze” that is “directed upon the ideal,” and that is “thought of in the light of the command, ‘Be ye therefore perfect’” (30). All the naiveties of nineteenth-century utopian socialism would be embodied in advising the exploited laborer to fix their gaze on some ideal of perfection while lovingly attending to the individuals by whom they are exploited. And such advice would be just as inapposite on the other side of the class divide. A high-minded land-owner might be lovingly attentive to all the particular flaws and talents of his individual laborers, devoutly hoping to do what is ideally best for each of them, while nonetheless remaining oblivious to the systemic injustice by which those flaws, talents, and hopes have been shaped. Although the loving attention advocated by Murdoch gets us as far as a Dorothea-Brooke-like enthusiasm for providing workers with well-appointed cottages, we are looking for a concept of emancipatory attention that can play a role in contexts calling for a more revolutionary change.

This is not to deny Murdoch’s claim that there can often be something morally transformative about “efforts of attention directed upon individuals and of obedience to reality as an exercise of love” (Murdoch 1970, 41). We should agree that such efforts can be transformative of individual relationships, while insisting that in many contexts of injustice it is not an individual relationship that needs to be transformed. Murdoch recognizes this. She writes that “We are not always the individual in pursuit of the individual” (ibid.), and does not pretend to have given us “a formula which can be illuminatingly introduced into any and every moral act” (ibid.). But our complaint is not just that Murdoch’s advice is suited to situations other than the systemic ones that we are concerned with. It is that her advice may be counterproductive, because focusing attention on the individual may prevent systemic injustices from coming into view. Frye provides a memorable image for this:

Consider a birdcage. If you look very closely at just one wire in the cage, you cannot see the other wires. If your conception of what is before you is determined by this myopic focus, you could look at that one wire, up and down the length of it, and be unable to see why a bird would not just fly around the wire any time it wanted to go somewhere. Furthermore, even if, one day at a time, you myopically inspected each wire, you still could not see why a bird would have trouble going past the wires to get anywhere. […] One can study the elements of an oppressive structure with great care and some good will without seeing the structure as a whole, and hence without seeing or being able to understand that one is looking at a cage and that there are people there who are caged, whose motion and mobility are restricted, whose lives are shaped and reduced. (Frye 1983, 4f.)

An attempt to view the most pervasive forms of injustice from a perspective that is focused, like Murdoch’s, on “the idea of the individual” threatens to introduce the myopic focus of which Frye complains.

Our last example of ethical and epistemic interplay concerns the quasi-legal processes of Truth and Reconciliation, in which participants confront systems of value that differ fundamentally from their own. The incommensurability of such systems creates barriers to communication, which are one of the reasons why it would be unhelpfully idealistic to suppose that epistemic challenges can be addressed simply by opening a dialogue. Surmounting such barriers requires occasions where morally salient events can be made a topic of mutual attention for people on both sides. Truth and Reconciliation processes work to create these occasions. Murdoch sees the need that they thereby address:

If the common object is lacking, communication may break down and the same words may occasion different results in different hearers. […] 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. We develop language in the context of looking. (Murdoch 1970, 32)

If Murdoch’s point here were only about the development of language then it would be unobjectionable, but she is inclined to think of this “looking” not only as a requirement for the mutually intelligible representation of a problem, or of an historic outrage, but as sufficient for the discovery of an adequate response to it: “The love which brings the right answer is an exercise of justice and realism and really looking” (Murdoch 1970, 91; emphasis in original). Such looking may indeed suffice to “bring the right answer” to the predicaments that Murdoch considers, but it cannot be relied on to bring an answer in efforts toward political reconciliation. Such efforts fail when the work of scrutinizing facts is not followed by work that moves toward a future in which prevailing norms have been overthrown (Nagy 2013; Pityana 2018).

When contrasting her picture with that of her existentialist rivals, Murdoch emphasizes the relative unimportance of the role that she assigns to the resolute pursuit of a course of action that one has elected for oneself, in defiance of the norms that one finds already entrenched in the world. She inherits from Weil the idea that “will is obedience not resolution” (Murdoch 1970, 39):

The idea of a patient, loving regard, directed upon a person, a thing, a situation, presents the will not as unimpeded movement but as something very much more like “obedience.” (Ibid.)

Murdoch must here be thinking about individual injustices, rather than systemic ones. It may be true for the mother-in-law, whose prejudices originate in some attitude of individual jealousy or disdain, that “If I attend properly I will have no choices and this is the ultimate condition to be aimed at” (Murdoch 1970, 38), but such a lack of choices cannot always be the condition to be aimed at. An adequate response to the injustices that we encounter in Truth and Reconciliation processes calls for a will that is more than merely obedient. Almost all of the participants in a Truth and Reconciliation commission may sincerely claim that they have maintained obedience to some authority, secular or sacred. Attention will have a role to play in distinguishing between those authorities that make genuine claims on our obedience and those that make spurious ones, but genuine claims are unlikely to be the loudest, most persistent, or most pervasive, and so the attentiveness that facilitates this adjudication between claims cannot consist in mere thoroughness of normative uptake. In contexts of transitional justice, the attentive agent must choose, and not merely weigh. We might take this choosing to be an existentialist moment of unconstrained freedom, or a rationalist application of some more abstract principle. In either case, we would be incorporating elements of a moral psychology that Murdoch explicitly disavows.

Murdoch does recognize that moral attention will sometimes call for an “exploration” and a “moral discipline” that cannot be adequately construed on the model of concerted vision, but, when it comes to specifying these additional requirements, her project reaches the point at which it starts to treat concepts as “perfectly obvious,” and “perfectly familiar,” rather than as complexes that need to be philosophically engineered:

[A]t the level of serious common sense […] it is perfectly obvious that goodness is connected […] with a refined and honest perception of what is really the case, a patient and just discernment and exploration of what confronts one, which is the result not simply of opening one’s eyes but of a certainly perfectly familiar kind of moral discipline. (Murdoch 1970, 37)

If we are to explain the ways in which attention can be emancipatory in contexts of systemic injustice, we need a concept of moral attention that goes beyond that which Murdoch articulates.

4.1 The Orientation of Recent Theories

We have seen that, because it was intended to be corrective to a Wittgensteinian philosophy that disparaged private mental phenomena, Murdoch’s account of the value of attention was oriented toward individuals, in a way that compromises its usefulness when our goal is to articulate the role of attention in addressing injustices that are social. Although this shortcoming may sometimes be severe, it might not seem to put Murdoch at any disadvantage when compared with philosophical theories of attention that have been developed more recently. Those theories have been primarily concerned with issues arising from the research of cognitive scientists. The form of attention studied by these scientists might seem to be even less socially pertinent than the private attention that was exemplified by Murdoch’s imagined mother-in-law. This form of attention manifests itself in millisecond reductions to response times, during tasks that require no moral thinking and that are completed in laboratory conditions. One might expect to find nothing of social relevance in the outcomes of these studies, nor in the philosophical theories that attempt to address issues arising from them, but in the final part of this essay I shall argue that those theories do give us a vantage point from which the emancipatory role of attention can be seen, even if seeing it requires us to supplement them with ideas taken from an older tradition; a tradition concerning the forms of attention that are supported by artistic and critical practices.

In the Anglophone world this tradition has more often been literary than philosophical, but it finds philosophical expression in the aesthetic works of R. G. Collingwood, especially in his 1938 book, The Principles of Art. Collingwood often insists on an idiosyncratic interpretation of his aesthetic vocabulary, as when he tells us that, in his sense of “art,” “most of what generally goes by the name of art nowadays is not art at all” (Collingwood 1938, 278), or when he uses the word “consciousness” in such a way that there is a marked difference between “sentience” and “consciousness” (205)—a difference to which the presence of attention is said to make an essential contribution. The need to interpret these remarks in the context of Collingwood’s larger theory makes him difficult to explicate by means of quotation. A better route by which to approach his ideas is via consideration of some recent philosophical theories of attention.

These theories contrast with their psychological contemporaries in the emphasis that they place on the relationship between attention and action. Psychologists have tended to approach attention by treating it as a phenomenon of perceptual resource management: even when allowing that cognition and emotion make their own contributions to the way in which one’s attention is directed, they have tended to think that, on any occasion when attention is being paid, its work is complete once the answer has been settled to the question of how the relevant subject’s perceptual processing capacities are allocated. Philosophers have sometimes shared this emphasis on perception, but they have avoided focusing on perception exclusively. Instead, they have taken attention to play a role in the selective direction of our psychological resources more generally, so that our interest in attention is not exhaustively concerned with settling the question of which stimulus will be in the perceptual foreground, but also involves questions about the path that is to be taken by one’s thoughts, about the courses of action that one will embark upon, and about the means that one will take in pursuing them. As noted above, our call for an ethics of emancipatory attention needs to avoid implying that systemic injustices can be addressed by impertinently staring at them, but one way in which we can avoid such an implication is if attention is understood as determining not only where we look, but also what we do. Recent philosophical theories promise to provide such an understanding. In the theory developed by Wayne Wu, this role for attention in action is accorded a special importance.

4.2 Selection for Action

Wu revives a theory that once enjoyed currency in the psychological literature (see, e.g., Allport 1987, Neumann 1987, Pavese & Buxbaum 2002). It starts from an ethological premise, concerning the distinction between reflexes and non-reflexive acts. Whereas reflexes are one-track responses, elicited by stimuli of some predetermined type, it is characteristic of non-reflexive acts that they can be occasioned by various stimuli, and performed in various ways. Creatures therefore have a capacity for non-reflexive agency only if multiple responses fall within the range of their capabilities. The exercise of that capacity requires selection from among this multiplicity. Because it results in one of many potential actions being performed on one of many potential objects, Wu refers to the making of this selection as the finding of a solution to a “many-many problem.”

Wu’s claim is that the solving of such problems is always accomplished by attention (Wu 2011b, 2019). In determining that one will pick this apple, rather than that, with these fingers, rather than those, and at this pace, rather than a slower or a faster one, one is thereby paying attention to the apple. When this claim is combined with our previous paragraph’s ethological premise, we get the result that attentiveness is necessary for apple-picking to be something that one does as an exercise of one’s agency, rather than as a reflex, or as something that emerges from some uncoordinated mass of reflexive responding. Because this point is quite general, the distinctive claim of Wu’s theory is that all non-reflexive acts require attentive selectivity. This can be taken as telling us about something that is essential to attention (as in Wu 2011a, Wu 2011c), or as telling us about something that is essential to agency (as in Wu 2011b, Wu 2019). The central idea is that these two phenomena emerge together.

Wu applies this theory to the agency of individual organisms, but the form of selectivity that he discusses can also be seen in actions that are undertaken collectively. Just as non-reflexive agency is a possibility for individuals only if they have several actions available to them, and only if they are creatures for whom there might be several potential targets on which these actions might be performed, so agency will be possible on the part of a collective only if there are several things that its members might collectively do, and several targets toward which those collaborative efforts might be directed. Cases in which a collective needs to act non-reflexively therefore create their own version of the many-many problem, and so collective agency requires the selectivity that Wu identifies with attention.

When Wu’s many-many problem is transposed from individual actions to social ones, it starts to resemble the sort of problem that Collingwood took to be most readily soluble in societies where there is a culture of art. This raises the possibility that Collingwood’s ideas about the cultural role of what he calls “art” can be used to cast light on the question of what might be needed for the collective version of the many-many problem to be reliably solved, thereby casting light on the question of how the paying of attention to social problems might contribute to their mitigation.

4.3 The Constitution of the Collective as an Agent

The examples discussed by Wu tend to be of many-many problems in which there is a given repertoire of possible actions, and a given set of objects toward which those actions might be targeted. A solution can be found to these problems through a process of picking appropriately from the options that are known to be available. But cases that are susceptible to such a straightforward solution will not be the most basic of the many-many problems that agents face during the development of their agency, especially not when the agent is a collective. Mature agents may often find themselves in the position of being confronted with a situation that is seen to be one in which several familiar objects afford opportunities for any one of several possible skilled responses, but it is only through work that such a position can be achieved. Naive agents must do several sorts of psychological work before objects can be ready-to-hand for them as potential targets for action. They must do further work before some variety of behavioral responses is recognizable as constituting their repertoire of possible actions. If this work has not been done adequately then the most straightforward route to the solution of many-many problems will be unavailable. In such cases, the requirements of non-reflexive agency cannot be dependably met, and so a certain kind of free action may be ruled out. Collingwood takes the as-yet-inattentive child to be in a position of that sort:

The freedom of mere consciousness is thus an elementary kind of freedom; but it is a very real kind. […] A child feels pain and screams; fear, and cringes; anger, and howls or bites; each in perfectly automatic reaction to the emotion of the moment. […] When the child becomes conscious, he not only finds himself feeling in various ways, but attends to some of these feelings and not to others. If he now howls with rage, it is not because of the rage simply, but because of his attending to it. [… The howl is] not the automatic howl of sheer rage, but the self- conscious howl of a child who, attending to his own rage, seems anxious to draw the attention of others to it. As this consciousness of himself becomes firmer and more habitual, he finds that he can dominate the rage by the sheer act of attending to what he is doing, and thus stop howling, master his feelings instead of letting them master him. (Collingwood 1938, 209)

Collingwood must here be using “consciousness” (and “self-consciousness”) in a sense that differs from that which is now current among philosophers and psychologists (for he is allowing that the as-yet-inattentive child feels pain, in a phenomenologically rich way, while maintaining that, until this child exercises attention, it is neither “conscious” nor “self-conscious”). The clarification of Collingwood’s use here would require a detour from our present theme, but even without embarking on such a detour it should be clear that his point is to establish a role for attention in explaining the selectivity that non-reflexive action requires. His understanding of agency’s demand for attention is recognizably in agreement with Wu’s, although the work for attention that Collingwood emphasizes is that of configuring the option-space for a many-many problem, rather than that of solving such a problem by picking options from a repertoire of actions that is already given.

Collingwood’s account of the way in which this work is accomplished at the individual level was intended to establish a pattern that could be discerned again at the level of the collective (and that he operates with throughout the concluding chapters of his book, which concern “The Artist and the Community”). At the collective level, we can see the work that is needed to configure the action-space for a many-many problem by considering the organization of industrial labor. An established trade union might find itself in the position of deciding whether to address pay, holiday provisions, or workplace safety, and of deciding whether its action will take the form of legal pressure, work-to-rule, or an all-out strike. The many-many problem facing such a union would be analogous to that of the mature individual who is confronted with some number of clear options for familiar types of action, directed toward some given repertoire of targets. But if this union’s problem can therefore be addressed by a process of appropriately picking from a known repertoire, this is only because work has already been done to ensure that these options are collectively recognized as belonging to that repertoire. At both the collective and the individual levels, non-reflexive agency requires the completion of this work, so as to configure a many-many problem into a problem that the agent can deliberately address. Collectives that lack the long history of trade unions may confront many-many problems that put them in a position analogous to that of Collingwood’s not-yet-attentive infant. An “elementary kind of freedom” may be unavailable to them.

I take this analogy to be illuminating (notwithstanding the risk that it might also be construed as patronizing),¹ but one point of disanalogy is also worth noting. An individual agent will usually exist—like Collingwood’s infant—as an agent of uncontrolled or reflexive behaviors, before attention has enabled it to get its act together as an agent of behaviors that are dependably non-reflexive. In the collective case, however, the achievement of attention might play a more foundational role in making it the case that this particular collective is capable of any action at all, whether reflexive or not: A collective having various response capabilities might constitute itself as an agent through the selectivity in which it exercises its attention. The fact that we are able to act as a single collective agent, and are not just a mob of similarly motivated individuals, may be grounded in our capacity to formulate and address many-many problems, and to thereby instantiate a form of collective attention.

4.4 Addressing Injustice Through Selection for Collective Action

Many-many problems can be overcome in various ways. Psychology tells us about the ways these problems are overcome in the brain of an individual. There attentional selectivity can be achieved through some more or less flexible combination of rational control, canalized habit, and an orientation toward the things that happen to make themselves perceptually salient (Theeuwes 2019). The selectivity that makes agency possible at the social level can be achieved in analogous ways, some of which involve deference to top-down authority, some of which involve entrenched practices, and some of which depend on the tendency for urgent situations to attract resources to themselves.

Each of these routes to collective attention is prone to its own characteristic modes of failure, in which the many-many problems facing a collective agent come to be solved in ways that are not in the interests of, or that do not reflect the values of, the individuals by whom that collective is constituted. The problems created by such failures will be problems that call for a better solution to a many-many problem. According to the Selection for Action theory, they will be problems that can be mitigated by improvements to our collective attention. Each of the injustices that we considered above can be understood as problems that are, at least partly, of this sort.

When we say that systemic injustices have an epistemic aspect, I have argued that we should not be heard as saying that those injustices arise because people act poorly on account of a failure to know what’s what. Our emphasis should instead be on the fact that these injustices sustain themselves because they lead to conditions that impede a critical public engagement with ideas that would configure the space of possible actions. Such engagement conforms to the picture of attention given by the Selection for Action theory: it is necessary for a group to address itself to a many-many problem, and thereby to constitute itself as a collective agent that is capable of acting from its own values, and in its own interests. The cultural conditions that facilitate these forms of creative and critical engagement are those that Collingwood theorized in his Principles of Art. He provides an account of various culturally mediated ways in which affective responses can be brought within the scope of our individual and collective agency. That account applies to the several ways in which we configure the space of possible public actions by acknowledging and sanctioning one another’s responses (all of which Collingwood refers to as uses of “language”):

Just as the life of a community depends for its very existence on honest dealing between man and man, the guardianship of this honesty being vested not in any one class or section, but in all and sundry, so the effort towards expression of emotions, the effort to overcome corruption of consciousness, is an effort that has to be made not by specialists only but by every one who uses language, whenever he uses it. Every utterance and every gesture that each one of us makes is a work of art. (Collingwood 1938, 285)

If we adopt this as our conception of “art,” then it will be art that implements the attention that is required for uncoerced action at a social level, especially for those social actions that are motivated by value-laden emotions, as when we collectively commemorate a grief, celebrate a victory, or deplore some act of brutality. An approach centered on such artful attention would enjoy all of the expected advantages over a knowledge-centered approach. It avoids the idealizations of dialectic because artists do not purport to be providing their audiences with objective reasons for the formation of beliefs, and because those audiences need not defer to the epistemic authority of the artists to whom they are attentive. It avoids the risk of patronage because audiences who attend to art that underestimates their knowledge may find such art to be naive, or banal, but are not thereby patronized or silenced by it. And it avoids the risk of presumptuousness because the attention that is demanded by art may include an openness to the prospect of being shocked, and because the presumptuousness of an audience that underestimates its own ignorance is unlikely to survive such a shock.

The epistemic aspects of systemic injustices can be understood not simply as failures of knowledge, but also, and just as fundamentally, as disruptions to the circulation of information that would enable a socially distributed set of capabilities to be selectively coordinated around a coherent goal. For the forms of injustice that were discussed by Marx, this coordination may be impeded by an absence of information-gathering resources, or by the presence of misinformation. In cases involving the marginalization of a traditional culture, it may also be attributable to corruption of the mechanisms by which the circulation of information can be achieved: practices of commemoration and intergenerational knowledge-transmission may have been suppressed; testimonial authority may have been undermined; and languages or literatures may have been derogated. The processes of resource-coordination that would be emancipatory in these cases are processes of the agency-constituting sort that Wu describes in his analysis of attention, and Collingwood takes to be essential to “art properly so called.”

By increasing our capacity for attention, increasingly large sets of capabilities can be coordinated over increasingly long timescales, and in the service of increasingly complex ends. Our capacity for agency can thereby be expanded. When it facilitates this expansion in the scope of our capacity for individual and collective agency, attention can be emancipatory.²