‘The Ethics of Attention to Language’ Introducing Conceptual Injustice

2.1 Wittgenstein and Concepts

The first point I want to make in this chapter is that it is no easy matter defining what a concept is. This is something that Herman Cappelen and David Plunkett discuss in the introduction of Conceptual Engineering and Conceptual Ethics (cf. 2020: 1 – 26). On the one hand, I base my understanding of concepts on their work when they define conceptual engineering as: “(i) The assessment of representational devices, (ii) reflections on and proposal for how to improve representational devices, and (iii) efforts to implement the proposed improvements” (Cappelen & Plunkett 2020: 3); consequently, concepts appear as “representational devices”. On the other hand, I draw on Wittgenstein, for whom conceptual activity has more to do with playing language-games and following rules of usage—something that can be better understood in the light of Gestalt theory, as conceptual activity appears as both rule-following and productive thinking (cf. Campbell 2020). If we follow Wittgenstein’s idea that in “a large class of cases—though not […] all—in which we employ the word ‘meaning’ […] the meaning of a word is its use in the language” (PI 1958: 43), then a concept—as it is defined in linguistics as the meaning of the word—appears to be close to use. It thus means that it is something fluent, dynamic, but at the same time it has to be permanent in some sense in order to enable communication; this stability is explained by rule-following and agreement in language. His discussion of rule-following is an investigation of what it means to share a sense of when (i. e. under what circumstances) to say what. “‘[O]beying a rule’ is a practice” (PI 1958: 202): we are trained to follow rules.

As Sandra Laugier explains in Du réel à l’ordinaire, the question is actually one of the criteria: our agreements in language determine themselves and are determined by criteria; these criteria regulate the conditions of our discussion and our application of concepts. They govern what we say: to agree in language means that language—our form of life—produces an agreement insofar as it is the product of an agreement, i. e. insofar as it seems natural for us in some sense (cf. Laugier 1999: 119). “[T]o imagine a language means to imagine a form of life.” (PI 1958: 19) These criteria are shared, they govern the use, and reveal that language arises from a “background of pervasive and systematic agreements among us which we had not realized, or had not known we realize,” and we discover that “we do agree in judgement—eliciting criteria goes to show that our judgements are public, that is, shared” (Cavell 1999: 30 – 31). Wittgenstein’s point is: there is no meaning behind the use, there is only meaning in use. The “meaning of a word is its use in the language” (PI 1958: 43).^[2]

As Toril Moi writes, use appears as some kind of relationship. We do not pay enough attention to use: we keep looking for something deeper, something that can ground use, like the ultimate explanation of meaning as such.^[3] Wittgenstein’s understanding of meaning as use unfolds into a original vision of the intertwinement of world and word, and of our lives in language. By “language-game” he means something like “language-practice” (Moi 2017: 44). The point of the term is to draw attention to the intertwinement of words and practices in order to show that we cannot understand a word or an utterance unless we understand the practice it is a part of. The concept of “language-game” tells us that to learn a language is not to learn a set of names but to be trained in—to learn to recognize and participate in—a vast number of human practices. Wittgenstein does not set up a systematic inventory of language-games: it would be impossible, as use is open-ended, in constant transformation, always responding to new circumstances. The concept of language-games helps us to see that there are many different tools in the tool-box of language (cf. PI 1958: 11). “To speak or write is never just to represent something, but to do something.” (Moi 2017: 47)

As Jasmin Trächtler writes in “Speaking in Monster Tongues: Wittgenstein and Haraway on Nature, Meaning and the ‘We’ of Feminism”, Wittgenstein is concerned with questions regarding how words can relate to things in the world, what the meaning of a concept is, and why we have the concepts we have in the first place. He presents this issue as a question about the extent to which there is a “correspondence between our grammar and general (seldom mentioned) facts of nature” (RPP I 1998a: 46). By this, he does not mean a causal explanation of our grammar from facts of nature (cf. PPF 2009: 365). Rather, he is interested in the significance of facts of nature for our grammar and the extent to which our concepts are thus part of our life (cf. LWPP II 1999: 72). As Trächtler recalls: there can surely be connections between our concepts and certain natural factors but mediated by what is important to people in their practice and lives, i. e. what they are interested in. Nevertheless, this does not mean that our concepts are justified by nature. Already in the early 1930s, Wittgenstein thought that grammar, that is, the logic of use of concepts, is not accountable to any reality. The rules according to which we use terms constitute meaning and are, in a sense, arbitrary (cf. BT 2005: 233r; see Trächtler 2023a). Conceptualizing is, therefore, paradoxically, a cultural and natural activity at the same time. As Trächtler explains, forms and patterns of life mean a recurring and recognizable order of characteristics of actions, situations, and linguistic expressions, which the speakers understand as a structuring regularity within their lives and refer to them by a concept. This underlines the rootedness of our concepts in human social life. Concepts by which we refer to patterns of life are “elastic which, however, ‘does not mean they can be deformed randomly and without offering resistance’ [LWPP II 1999: 24], but rather that their boundaries are not sharp. […] Wittgenstein does not want to reduce this conceptual unsharpness ‘to sharpness; but to capture unsharpness conceptually’ [cf. Ms 137: 64]” (Trächtler 2023b: 6).

2.2 Iris Murdoch and Concepts

My analysis of Murdoch’s thoughts on concepts moves from Wittgenstein’s influence on Murdoch’s philosophy of language to her singular understanding of moral concepts. Challenging the Oxfordian philosophy of the 1950s and 1960s, she is a pioneer at championing the post-Wittgensteinian moral current (cf. Mac Cumhaill & Wiseman 2022; Krishnan 2023). She witnesses a 1940s-Wittgenstein, approaching language as a social phenomenon, basically saying: language is not merely a picturing activity, it is not simply about naming, but it is an overlapping array of human practices. “Here the term ‘language-game’ is meant to bring into prominence the fact that the speaking of language is part of an activity, or of a form of life.” (PI 1958: 23) Although she opposes Wittgenstein on some points, e. g. on the importance of inner life for moral life or the fact/value distinction (cf. Murdoch 1956: 49; Murdoch 1992: 156), she is indebted to Wittgenstein on many points, such as large parts of her philosophy of language, embodying his “detestation of (what he felt to be) any careless use of words” (Murdoch 1992: 280). From her talks with Elizabeth Anscombe, who was attending Wittgenstein’s classes in the 1940s, she understood the importance of looking at (language-games) as a specific method, i. e. paying attention to language and grasping what is at play in it, what it says about us when we speak with it and through it. Murdoch and Wittgenstein share the view that discourse happens with a background “against which we see an action, and it determines our judgment, our concepts, and our reactions” (RPP II 1998: 629). The two philosophers are aligned when it comes to the openness they see in conceptual activity and the limits of language.

Against a form of logical positivism that is flourishing at Oxford in the 1950s, Murdoch writes in her 1956 article “Vision and Choice in Morality” that there are some situations where the moral agent may find themselves unable to describe something which they, in some sense, apprehend. This is precisely when we must go to the limits of our language. She endorses the Wittgensteinian argument that private language is logically impossible, but this does not mean for her that there is no intimacy in language: this is what she sets out to show. She defends a moral life in language that is open to permanent reinterpretations and rehabilitations, and her parables echo Wittgenstein’s background, pictured as “a very complicated filigree pattern” (RPP II: 624): “How ambiguous a parable appears to be will depend on the coherence of the moral world in which it is being used. Certain parables or stories undoubtedly owe their power to the fact that they incarnate a moral truth which is paradoxical, infinitely suggestive and open to continual reinterpretation.” (Murdoch 1956: 50).

But Murdoch never fully acknowledges the influence of Wittgenstein in her work. What has been most remembered about Murdoch and Wittgenstein is her opposition to the distinction of fact and value for which she reproached Wittgenstein, as it seemed to her to be inadequate to offer a realistic view of conceptual activity. She sees ordinary conceptual activity as a “continuous detailed conceptual pictorial activity whereby (for better or worse) we make and remake the ‘world’ within which our desires and reflections move, and out of which our actions arise”, and argues that in everyday speech we deploy “a complex densely textured network of values round an intuited centre of ‘good’. We imagine hierarchies and concentric circles, we are forced by experience to make distinctions, to elaborate moral ‘pictures’ and a moral vocabulary. We work with value concepts, value words.” (Murdoch 1992: 325)

While we talk and think, Murdoch says, we constantly examine and alter our sense of the order and the interdependence of our values: “The study of this interweaving is moral reflection and at a theoretical level makes intelligible places for defining and understanding central concepts [such as] happiness, freedom, or love.” (Murdoch 1992: 326^[4]) She criticizes the view of moral concepts whose meanings are in line with elements of recommendation and specification in which the moral agent appears as rational and responsible, moves unhindered against a background of facts, and can alter the descriptive meaning of his moral words at will (cf. Murdoch 1956: 35). By redefining the ethical importance of vision as a counter-ideal of attention and combining it with the Gestalt perspective to be found in Wittgenstein’s thought, she comes to the understanding that such an approach to human life implies a redefinition of the traditional conception of concepts and meaning:

This picture seems plausible if we take as the centre of ‘the moral’ the situation of a man making a definite choice (such as whether to join a political party) and defending it by reasons containing reference to facts. It seems less plausible when we attend to the notion of ‘moral being’ as self-reflection or complex attitudes to life which are continuously displayed and elaborated in overt and inward speech but are not separable temporally into situations. Here moral differences look less like differences of choice, given the same facts, and more like differences of vision. In other words, a moral concept seems less like a movable and extensible ring laid down to cover a certain area of fact, and more like a total difference of Gestalt. We differ not only because we select different objects out of the same world but because we see different worlds. (Murdoch 1956: 40)

What interests Murdoch first is what people see through their “schemes of concepts” and especially through the moral character of these concepts (cf. Broackes 2012: 10). In The Sovereignty of Good, she argues against moral theories that make moral disagreement simply a matter of difference in the way people choose between alternatives and instead insists on the dependence of moral thought upon conceptual scheme. The point is that a person’s conceptual apparatus can limit or broaden the range of options that she is able to recognize and choose from. The conceptual scheme may be said to determine and reveal the character of the moral world in which a person lives. She characterizes these conceptual schemes by foregrounding the differences that distinguish people within a moral perspective: a difference not just in the implementation of shared concepts but in the repertoire of concepts that people comprehend and use (cf. Broackes 2012: 12). Therefore, she is not simply interested in the phenomenon of one person’s change of thought about a particular case but more in the process of revision, evolution, and deepening of our vocabulary and conceptual scheme—which leads to a better mutual understanding (cf. Diamond 2010). For our differences of perspective are above all differences of the ethical concepts possessed:

[I]f we attend to the more complex regions which lie outside ‘actions’ and ‘choices’ we see moral differences as differences of understanding […] which may show openly or privately as differences of story or metaphor or as differences of moral vocabulary betokening different ranges and ramifications of moral concept. Here communication of a new moral concept cannot necessarily be achieved by specification of factual criteria open to any observer (‘Approve of this area!’) but may involve the communication of a completely new, possibly far-reaching and coherent, vision; and it is surely true that we cannot always understand other people’s moral concepts. If we take the view that moral differences are in this sense ‘conceptual’ and not exclusively behaviouristic, we shall also be able to see moral philosophy itself as a more systematic and reflective extension of what ordinary moral agents are continually doing, and as able in its turn to influence morality. (Murdoch 1956: 41 – 42)

As Murdoch herself recognizes, the idea that moral differences are conceptual—in the sense of being differences of vision—suggests that morality must, to some extent, be studied historically. This is linked to the view of conceptual activity as what counts for us (with all the ambiguity of this “us”). Hence the idea of “day-to-day” human “fables” constituting moral worlds: “If my argument […] is accepted so far, then it will be conceded that a considerable area of personal reflection is morally important in the sense of constituting a person’s general conceptual attitude and day-to-day ‘being’, which will in turn connect in complex ways with his more obviously moral ‘acts’.” (Murdoch 1956: 44 – 45) By referring to the background of lives into which concepts inscribe themselves, she emphasizes the irreducibly social aspect of language, pointing, without referring explicitly to it, to language-games and the rule-following aspect of conceptual activity (cf. Murdoch 1956: 48).

But where Wittgenstein and Murdoch differ is in what we might call the “sites” of conceptual change: whereas Wittgenstein seems to endorse the concept only in its public shareability (the idea of following a rule accounts for this social and public criterion), Murdoch explores more inward avenues of conceptual reform, drawing attention to how inner life is an ordinary site of moral life (cf. Moi 2017; Laugier 2022).

2.3 Framing Conceptual Injustice

In “Facts, Concepts and Patterns of Life—How to Change Things with Words”, Trächtler refers to Wittgenstein’s last writings where he characterizes concepts and conceptual differences that we have with the role they play in our lives, as connected “with what interests us, what matters to us”. But “[w]hat if certain concepts—or their absence—are exclusionary, discriminatory, or otherwise unjust to those who are not ‘us’?” (2023b: 1) How can we perceive conceptual injustices that language can carry? Trying to render significant what she feels as conceptual injustices from Wittgenstein, Trächtler writes:

Wittgenstein never explicitly explained this ‘us’ or ‘we’ and he seems even less interested in the moral and political implications of such a ‘we’—especially as conceived opposite to a ‘they’ […]. However, this is of particular relevance in the case of concepts that do not refer to ‘very general facts of nature’, such as color and number words, but to patterns of life or, more generally, ‘social facts’: i. e., concepts that refer to social experiences, identity or to social life. […] Language can thus be rendered a political instrument of power whose conceptual distinctions draw boundaries that include and exclude and where the presence or absence of certain concepts can stigmatize, discriminate or reinforce social injustices. (Trächtler 2023b: 6)

Drawing on Fricker’s insights, Trächtler notes that “the extent to which such exclusion and injustice become manifest at such a conceptual level [shows itself] in the case of the absence of certain concepts” (2023b: 6). To understand the delicate shift between, on the one hand, conceptual activity as a matter of language-games, practices, customs, and uses and, on the other, a certain unfairness felt as inherent in this process of conceptual activity, she defines conceptual injustice as “an umbrella term” for injustices arising through concepts, conceptual distinctions and conceptual boundaries, or their absence.

Cases of conceptual injustice that go beyond hermeneutical (and epistemic) injustice include, e. g., the wrongful exclusion of a person, group, or phenomenon from a concept referring to social facts, such as the exclusion of transwomen from the concept ‘women’, the use of the generic masculine for groups of different genders, the deliberate use of inaccurate gender pronouns or first names (deadnaming), the conceptual misrecognition or ignorance of gender diversity in bureaucratic contexts or forms of address, and offensive, derogatory or otherwise discriminatory terms used to refer to certain marginalized groups. Looking at these various forms of conceptual injustice, it is not surprising that the struggle against social injustice is often related to language politics, changing discriminatory expressions and introducing new concepts—to finally change relationships, states of affairs, matters of fact, or more generally, ‘things’ of social life. (Trächtler 2023b: 7)

How is conceptual change possible? Trächtler calls on us to “exercise [our] imagination and shifts of perspective to see the limitations or the exclusionary boundaries of our concepts and to find new conceptual possibilities, and new patterns in the weave of life” (2023b: 7). From her, I perceive a form of conceptual injustice as one in which a concept is not shared by all the members composing the society, or is missing, which leads to an unequal distribution of cognitive resources within the society, a gap in the collective hermeneutical resource leading to unequal hermeneutical participation, or hermeneutical marginalization.^[5]

In Murdoch’s philosophy, another implicit form of conceptual injustice is highly perceptible, one in which we use words whose meaning for us has been lost—our words no longer ‘speak’ to us (cf. Diamond 1988; Forsberg 2013; Cavell 2015). She discusses this loss in two essays, although she does not define the problems she refers to as conceptual injustices: “House of Theory” (1958) and “Against Dryness” (1961). In both articles, she offers an analysis of the political and moral situation of Great Britain in the 1950s (in particular, socialism and liberalism as the two major political and intellectual movements shaping British political life), but the novelty lies in the linguistic and conceptual approach she takes. In “House of Theory”, she demonstrates that if socialism can no longer bring people together, it is because the concepts it uses no longer speak to people, bring them together, or even tell them anything (cf. Murdoch 1958). In “Against Dryness” she goes further as she deplores a general state of conceptual impoverishment in moral philosophy and political philosophy (cf. Murdoch 1961: 18).

What Murdoch does illustrate is the ethical dimension of conceptual engineering: what a concept implies in terms of norms and values. As Cappelen and Plunkett ask: “What are the dimensions of assessment for concepts? Which philosophical concepts are defective, and how can we improve them? How important are facts about the history (or ‘genealogy’) of our use of concepts to the assessment of our current concepts? These are questions at the heart of the fields that we call ‘conceptual engineering’ and ‘conceptual ethics’.” (Cappelen & Plunkett 2020: 2) Actually, this is what Murdoch does with “freedom”, “love” and “good”, and her philosophy is about reconceptualizing and repotentializing language from a moral perfectionist perspective (cf. Murdoch 1961: 19 – 20).

Conceptual engineering is one response to conceptual injustice. However, to understand the most important dimension of conceptual injustice, we need to understand how complex it is, and how, within conceptual engineering itself, it can inevitably and perpetually reproduce itself, i. e. situations of injustice and harm can be reintroduced. One example is a concept like wokism. Such a concept reflects a state of affairs already explored by Judith Butler in Excitable Speech: A Politics of the Performative, where she speaks of a “clash of uses” over the term “queer”, whereby it becomes either an offense directed towards homosexuals or a pride-claiming use, i. e. whether it is a stigma or a force to be reckoned with. Similarly, “wokism” is either a claim to social awareness of historical and institutional injustices against minorities or an expression of the will to keep these phenomena invisible or not really mattering (“You wokist!”). The risk that the term “queer” may become an insult again is never far away which reveals that conceptual engineering is not just about introducing a new concept to replacing a dysfunctional one, but it is about “struggles for usage” in ordinary, everyday language. Some people still use the concept “queer” in a pejorative sense, so, in this sense, one could confirm Butler’s view that what could be called a “clash of uses” ultimately saw the positive identification use winning—but only partly, and for how long remains to be seen.^[6]

Based on these insights, I have defined in the introduction conceptual injustice as an interweaving of three structuring dimensions: it is not necessary for all these three dimensions to be involved for there to be conceptual injustice; one of them can sufficiently account for conceptual injustice. In the next section, I use my definition of conceptual injustice to offer new insights on the experience of oppression and on how an ethics of attention to language can help counteract oppression in its linguistic aspects.

3.1 Language and the Experience of Oppression

This section considers the experience of oppression primarily in terms of the structuring relationships between it and the language that the person experiencing oppression possesses. In other words, it is about understanding the fabric of oppression in its linguistic, existential and epistemic manifestations. This part is mostly indebted to the work of Mickaëlle Provost: L’expérience de l’oppression (2023). The point is to grasp how an experience of oppression is not solely weaved around a language of oppression (e. g. the racist address) but is consolidated by mechanisms of conceptual injustice that enclose subjects in situations of oppression.

Scholars might argue that naming a situation of injustice or oppression is not all there is to do—I can only agree.^[7] But we misjudge the fact that, as language is a social phenomenon, our concepts are necessarily related not just to what matters but also to what counts for dominant groups. What must be understood is how concepts can be to the advantage of some and to the disadvantage of others and how to get rid of a pervasive false naivety about language. This means two things: on the one hand, when a person lacks a concept that would be of benefit, the situation is unjust and necessarily advantages a group of people who benefit from that person not having the concept (e. g. Carmita Wood’s sexual harassment story); on the other hand, a concept that exists but is exclusionary can also be disadvantageous to someone or a group and advantageous to some dominant group. One might say: but what is not fair about that?

Oppression marks and permeates lived experience but does not always constitute a central dimension of experience, which means that it is not always acknowledged or named as such: oppression deprives one of precisely the resources that would enable it to be perceived, thought or expressed (cf. Provost 2023). In other words, living under oppression does not entail knowingly experiencing oppression (consciously or attentively) nor is it necessarily accompanied by a feeling of oppression and the emotions this feeling arouses. Experience therefore has political preconditions: what counts as experience depends on historical, political and social, state, or medical practices that define what is and is not supposed to count, for whom and how.

These practices help to give certain aspects of experience “more prominence than others”, shaping perception (what we consider worthy of attention or interest), affectivity (what can be an object of suffering, shame, or pride) or thought (what is likely to be the object of reflexive reconsideration: to be represented or spoken about). (Provost 2023: 41; my transl.)

Provost defines the awareness we may have of oppression as perceptive attention, a way of working on what we experience. Perceptive attention to the experience of oppression is thus a mode of perception that works on the consequences of oppression, helping to undo the logics that perpetuate it. It reveals the “systematic” character of oppression as, somehow, tied in with language (cf. Provost 2023: 15). Thus, a language-based approach to oppression is not about re-articulating on a discursive (theoretical and legal) level the experiences of oppression that victims experience in a confused or limited way. More radically, it “aims to elicit new experiences—experiences lived as unjust, oppressive—that could not previously be lived as such due to an absence (discursive, descriptive, categorical) of available interpretations” (Provost 2023: 44 – 45; my transl.). This failure to interpret an experience that happens because of an absence (discursive, descriptive, categorical) of available interpretations refers to a second form of conceptual injustice, namely: a gap in the collective hermeneutical resource leading to hermeneutical marginalization in society.

Therefore, the point of being attentive to language is to reflect on new experiences, new affects, by establishing interpretive grids against which violence can be named, perceived, and felt in its profoundly unjust character. Using Fricker’s analysis of sexual harassment, Provost argues that ‘finding the right expression’ was not a label providing a translation of a reality which, until the legal institutionalization of the category and its circulation in language uses, would have been experienced as such but would not have had words to express itself. The legal category informs new experiences and provokes effects in what is and will be experienced. It enables what is experienced to be interpreted in the light of new feminist paradigms in which “sexual seduction” is redefined as violence, disempowerment, and humiliation (Provost 2023: 44; my transl.). In the field of feminist theory, this joint understanding of language and oppression has shed new light on what lies at the heart of the (epistemological) feminist method, namely: the importance of starting from an inadequately understood or muddled experience of oppression, describing it in order to better understand and clarify it.

As every experience is situated within a discursive and an ideological system that is not impermeable to dominant norms, how experience is expressed, therefore, depends on the discursive categories and norms that organize it and give it coherence. To articulate an experience of oppression requires the use of a repertoire of words and images that may themselves be structured by dominant, patriarchal discourses. The experience put into words in an oral or written narrative is therefore not the literal translation of a past experience but the recomposition by and in the language of an experience itself shaped by means of interpretation. It is fundamentally connected to discursive conditions capable of making it legible, of articulating by drawing on existing meanings that it may eventually contest or reinvent (cf. Provost 2023: 48). This implies that we need to be aware, at the very heart of conceptual engineering, of the forms of domination and marginalization that can constantly reoccur. And that requires constant attention to language, a never-ending mode of awareness of the social injustices that can be woven into language.

3.2 Hermeneutical Injustice: a Form of Conceptual Injustice

“Feminism has long been concerned with the way in which relations of power can constrain women’s ability to understand their own experience”, writes Fricker in Epistemic Injustice to introduce her notion of hermeneutical injustice (2007: 147). She argues that, in a society in which social power has an unfair impact on collective forms of social understandings, the powerful always tend “to have appropriate understandings of their experiences ready to draw on as they make sense of their social experiences, whereas the powerless are more likely to find themselves having some social experiences through a glass darkly, with at best ill-fitting meanings to draw on in the effort to render them intelligible” (2007: 148). In order to grasp why I define hermeneutical injustice as a form of conceptual injustice, let us consider Carmita Wood’s sexual harassment story—especially the political conceptual ‘victory’ of it, when Wood and the group of other women colleagues, victims of sexual harassment, search through their linguistic repertoire to discern which concepts would best describe their shared experience of oppression, i. e. a man’s harassment towards them, as it is narrated by Susan Brownmiller:

[…] ‘Eight of us were sitting in an office of Human Affairs’, Sauvigne remembers, ‘brainstorming about what we were going to write on the posters for our speak-out. We were referring to it as ‘sexual intimidation’, ‘sexual coercion’, ‘sexual exploitation on the job’. None of those names seemed quite right. We wanted something that embraced a whole range of subtle and unsubtle persistent behaviors. Somebody came up with ‘harassment’. Sexual harassment! Instantly we agreed. That’s what it was. (Quoted in Fricker 2007: 150.)

As Fricker comments, “here is a story about how extant collective hermeneutical resources can have a lacuna where the name of a distinctive social experience should be” (2007: 151), and, indeed, Wood’s sexual harassment story illustrates what it is like to suffer an acute cognitive disadvantage from a gap in the collective hermeneutical resources which is how hermeneutical injustice precisely works. By reinscribing this story within hermeneutical marginalization, where members of the disadvantaged group are suffering cognitive disablements preventing them from rendering intelligible and communicating their experiences when it would yet be in their interest to do so, Fricker highlights the fact that because there is a structural identity prejudice, which is discriminatory, therefore, there is an epistemic wrong not due to the epistemic subject and therefore an injustice.

The conceptual injustice lies in the fact that if the harasser’s cognitive disablement (he lacked the concept of sexual harassment too) is not a significant disadvantage to him, by contrast, the victim’s cognitive disablement is seriously disadvantageous to her. The cognitive disablement prevents Wood from understanding a significant aspect of her own experience—for without that understanding she is in trouble, isolated, and vulnerable to continued harassment. As a form of conceptual injustice, this unequal hermeneutical participation of Wood is the reason why her cognitive disablement constitutes an injustice, and an example such as how the concept of sexual harassment was missing in the 1960s is symptomatic of the unequal relations of power in society that often prevent dominated or marginalized groups of people from participating on equal terms with dominant groups of people in those practices by which collective social meanings are generated. I shall point out that when Fricker writes,

the primary harm of hermeneutical injustice consists in a situated hermeneutical inequality; the concrete situation is such that the subject is rendered unable to make communicatively intelligible something which it is particularly in his or her interests to be able to render intelligible. […] The first prejudicial exclusion [in testimonial injustice; C.B.] is made in relation to the speaker, the second in relation to what they are trying to say and/or how they are saying it. (Fricker 2007: 162)

she actually faces a form of conceptual injustice, as embodied in this impossibility for someone to speak, “in relation to what they are trying to say and/or how they are saying it”. That is part of the reason why I take the level of an objective conceptuality, and I do not remain in the subjective one, as I understand the gap in collective hermeneutical resources to be an objective difficulty and not a subjective failing.

If hermeneutical injustice accounts for the embedded political strength of language, what it implicitly captures is conceptual injustice. As Sabina Lovibond remarks towards Fricker’s story: the discovery of “sexual harassment”, a term so widely and fluently used by so many women around the world today, is indicative of a linguistic deficiency that was part of their vulnerability and submissiveness to men, something that “[met] an implicitly felt conceptual need”.

For us, the main lesson to be learned from the sexual harassment story (or rather from the faculty of purposive creativity which it represents) is that in achieving awareness of some socially conditioned shortcoming in our ‘hermeneutical resources’ at a given moment, we are not locked into a deterministic acceptance of the status quo, but on the contrary may be emboldened to intervene in the relevant social process. (The purpose of introducing the term ‘sexual harassment’ was to provide women, in particular, with the means of identifying and resisting a certain pattern of exploitative behaviour; its rapid and widespread acceptance can be regarded as evidence that the new term did in fact meet an implicitly felt conceptual need.) (Lovibond 2011: 41)

This is precisely where I would like to bring in the notion of conceptual engineering by presenting a successful scenario in which moral imagination—or what Fricker refers to as “imaginative moves in which existing resources are used in an innovative way that stands as a progressive move in moral consciousness” (2007: 104)—is paired with conceptual innovation and conceptual implementation, i. e. the case of Wood, whose discovery of the term “sexual harassment” was a “life-changing flash of enlightenment”. As Lovibond puts it,

because the lacuna in this case (namely, the lack of a name for ‘sexual harassment’) is not a mere accident with regrettable consequences for a certain class of speakers, but results from the fact that ‘the whole engine of collective social meaning is effectively geared to keeping these obscured experiences out of sight’ (that is, keeping women in a state of mystification which makes them easier prey for sexually exploitative bosses, or the like), the discovery or invention of something with which to fill it constitutes a political success. (Lovibond 2011: 40)

A parallel can be drawn with Sally Haslanger’s remark on the commonality problem which “questions whether there is anything social that all females can plausibly be said to have in common. […] On my analysis women are those who occupy a particular kind of social position, viz., one of sexually-marked subordinate” (Haslanger 2000: 45). Therefore, if we acknowledge Moi’s picture of the ‘woman’ as a combination of a socio-historical and conceptual situation, in other words, a kind of result of sexist cultural practices and traditions (cf. Moi 2001: 6 – 8), we understand how difficult it is, even in conceptual engineering, to avoid a normativity that may still encompass, as Nancy Fraser puts it, “sexist misrecognition” and “status subordination” (Fraser 2007: 28, 31). Haslanger writes: “[…] this leads us directly from the commonality problem to the normativity problem. […] One worry is that bias inevitably occurs in deciding which experiences or social roles are definitive […].” (Haslanger 2000: 46)

3.3 The Case of Women

The concept ‘woman’ seems illustrative of conceptual injustice: as feminist theory has shown it since many years now, the prototypical cases of womanhood are fundamentally political, i. e. they result from social relations relayed through institutionalized patterns of cultural value in society (cf. Díaz-León 2015, 2016, 2022). Trying to escape the sex/gender distinction, Moi puts back at the core of feminist epistemology Simone de Beauvoir’s phenomenological approach in What is a Woman?. Moi contends that a ‘woman’ is, first, a body which is a situation in the world and has no absolute meaning (in the sense that bodies have no absolute meanings). Doing so, she tries to overlap the assumption “that political exclusion is coded into the very concepts we use to make sense of the world”, arguing that “it is this idea that made some post-structuralists assume that the word ‘woman’ can never be used in non-ideological ways, that ‘woman’ must mean ‘heterosexual, feminine and female’. In this view, all concepts become bundle concepts: mention one word and hosts of others are taken to be implied.” (Moi 2001: 43) If Moi recognizes that language in general and concepts in particular often carry ideological implications, she advocates for the Wittgenstein method of looking at language-games: if the meaning of a word is its use, thus, used in different situations by different speakers, the word ‘woman’ might take on very different implications:

If we want to combat sexism and heterosexism, we should examine what work words are made to do in different speech acts, not leap to the conclusion that the same word must mean the same oppressive thing every time it occurs, or that words oppress us simply by having determinate meanings, regardless of what those meanings are. (Moi 2001: 43 – 45)

Looking at the word ‘woman’, and seeing what it concretely makes in society, of individuals, as much as how it structures both individual and collective relationships, is an example of putting attention to language into practice. Once Moi has made a sort of feminist genealogy of the various criticisms leveled at the concept of ‘woman’, from biological determinism to sexist ideologies and the sex/gender distinction—“the pervasive picture of sex” (Moi 2001: 2)—which she believes is insufficient to make this concept effective in telling us, if we define ourselves as ‘women’, she suggests repotentializing the concept using the phenomenological and existentialist Beauvoirian approach. We have here an interesting case of conceptual engineering, since the concept is not rejected as such nor is it supplemented by a new one better able to tell us—but it is reprocessed, Moi offering a new Beauvoirian use of it (cf. Moi 2001: 62). In particular, Beauvoir’s understanding of individual subjectivity is vastly different from sex and gender theories, e. g. as “it never occurs to her that an individual can be divided into a natural and a cultural part, in the way suggested by the sex/gender distinction” (Moi 2001: 70).^[8]

The phenomenology of Simone de Beauvoir shows that “the relationship between one’s body and one’s subjectivity is neither necessary nor arbitrary, but contingent” (Moi 2001: 82). Indeed, Beauvoir invites us to study the varieties of women’s lived experience:

One aspect of that lived experience is the way in which the individual woman encounters, internalizes, or rejects dominant gender norms. But this encounter is always inflected by the woman’s situation, and that means by her personal and idiosyncratic history as this is interwoven with other historical situations such as her age, race, class, and nationality, and the particular political conflicts in which she may be involved. (Moi 2001: 82; my emphasis).

4.1 Linguistic Interventions: Disagreement, Transgression and Transformation

To grasp conceptual ethics and the ethics of attention to language it is needed to understand how we can disagree in language—and why disagreeing in language can be ethical. For Moi, forms of life have a strong social dimension; they are arbitrary in the sense that they probably could have evolved differently, but they are meaningful and tell us something about a community’s shared way of life. The concept of form of life should remind us that many cultural practices are so fundamental to our ways of being in the world—in a way, they are natural to us—, that we cannot change them at will. Moi gives a bodily aspect to forms of life—something close to the Gestalt theory (cf. Moi 2017: 57). Forms of life invite us to rethink the role conventions play in our lives and to consider what we share with others, but Moi takes another direction than that of “social construction”: “we should see forms of life as absorbing mutually the natural and the social” (Moi 2017: 59). Thus, Wittgenstein’s agreement at stake in PI 1958: 241^[9] is something like the conditions of possibility for us to be able to actually have disagreements—to play the language-game of disagreement, opposition, objection—in the first place. To share a language is to make the same judgments concerning grammar, criteria, and language-games: we can only mark differences and disagree against a background of a shared understanding. Consequently, the willingness to look at language-games is the hallmark of the realistic spirit: by considering language as use, as a practice, Wittgenstein places the burden of responsibility on us: “you are responsible for your words, I for mine” (Moi 2017: 75).

Sometimes, we might feel ‘lost’ in language. In such moments, Moi argues that a “grammatical investigation investigates whatever it is that makes us puzzled and confused” (Moi 2017: 53), which involves putting words back into their everyday uses.^[10] The aim is to reach a “clear view of the use of our words”, a “perspicuous representation”, i. e. a concept which is of fundamental importance for Wittgenstein: “The concept of perspicuous representation is of fundamental significance for us. It earmarks the form of account we give, the way we look at things.” (PI 1958: 122) When combined with Murdoch’s thinking, this is where an ethics of attention to language can emerge. Indeed, Moi is arguing for an ordinary language philosophy enabling us to connect the ethics of attention developed by Simone Weil, Cora Diamond and Murdoch to questions of language and writing.^[11]

An ethics of attention to language can therefore be based on linguistic and cognitive strategies. Rachel Katharine Sterken defines these strategies as, sometimes, linguistic interventions (i. e. communicative activities on the part of a speaker that [intentionally and strategically] attempt to change the word-meaning pairs in circulation—see Sterken’s [2020: 420] account of Haslanger’s ameliorative analysis of the concept ‘woman’). Such linguistic interventions can involve disruption and transgression: in such cases they are defined as linguistic transgressions as they elicit “transformative communicative disruptions” (Sterken 2020: 423).

In such cases, the interventionist’s transgression is justified (it is outweighed by the potential benefits to be achieved—either representational or worldly), and she is engaging in activity whereby her interlocutor can reflect on the meaning of the given word, acquire the new meaning, and recognize the new meaning as an improvement. Sometimes being a good member of a linguistic community will involve disrupting and transgressing. (Sterken 2020: 423)

In this sense, inclusive writing (l’écriture inclusive) appears as a case of transformative communicative disruptions whose later ethical benefits (making the language less steeped in sexist norms and schemes) are more important than the discomfort of seeing the writing of words change.^[12]

4.2 The Ethical Aspect of Conceptual Engineering

Sometimes having certain word-meaning pairs in circulation in a population of speakers at a particular time, in a particular social-historical milieu, can be bad. Such word-meaning pairs might cause injustice or disadvantage, stifle discourse, deliberation and inquiry, or stall social progress. It is not hard to think of examples—take any slur. The population would be better off without such word-meaning pairs. Likewise, sometimes not having certain word-meaning pairs in circulation in a population of speakers at a particular time, in a particular social-historical milieu, can be bad. Not having these word-meaning pairs can cause injustice or disadvantage, stifle discourse, deliberation and inquiry, or stall social progress. (Sterken 2020: 417)

By recognizing that (whether or not) to have “word-meaning pairs in circulation in a population of speakers at a particular time, in a particular social-historical milieu, can be bad” Sterken actually faces conceptual injustice, saying no more and no less that “these word-meaning pairs can cause injustice or disadvantage” which create cognitive blockages and forms of cognitive disruption as it “stifle[s] discourse, deliberation and inquiry, or stall[s] social progress” (Sterken 2020: 417). But she also opens new avenues for reflection. Taking the stance of conceptual ethics, I join her in saying that “it matters what word-meaning pairs are in circulation for a given linguistic population; that which language we speak can have a significant impact on whether or not the world is as it should be,” and hence “that normative claims about word-meaning pairs are important to reflect on” (Sterken 2020: 418; my emphasis). When we pay attention to language, we begin to be more mindful to conceptual activity and issues of refinement. In this sense, we can grasp with Sterken why linguistic interventions can lead to epistemically transformative and personally transformative experiences, because, as Burgess and Plunkett write:

Arguably, our conceptual repertoire determines not only what beliefs we can have but also what hypotheses we can entertain, what desires we can form, what plans we can make on the basis of such mental states, and accordingly constrains what we can hope to accomplish in the world. (Burgess & Plunkett 2013: 1096 – 1097)

Therefore, how can linguistic imagination and conceptual change be connected not only for the sake of social justice but also for our own personal and epistemological experiences? This is not to say that imagination is in itself sufficient for conceptual change. But imagination not only “serves the critical function of questioning the correctness, reasonableness or validity of our own concept formations”, it also “stimulates us to reconsider something supposedly certain”, and “questions what is natural or self-evident to us, and the things that have escaped our remark because of their familiarity” (Trächtler 2023b: 8 – 9). As Provost writes in L’expérience de l’oppression: often, what we take to be ‘natural’ is merely what is ‘familiar’ to us—much closer to a fact of culture than a fact of nature (cf., e. g., Provost 2023: 55). Imagination encourages us to see things otherwise, to explore scenarios, and to counter practices of institutionalizing and naturalizing things and situations that are in no way strictly natural. It is not to defend a radical socioconstructivist vision of life and language but to reinstate our responsibilities when it comes to language. It is an attempt to address how things are actually done and undone in language—performed in language.

Trächtler considers imagination as “a cognitive faculty helping us to see certain conceptual prejudices and injustices as such in the first place—whether for philosophical purposes or political ones, to change the concepts—and with them, the things, the conditions in living together with others” (2023b: 9). From a Wittgensteinian standpoint, talking about conceptual change means certain language-games losing some of their importance while others grow in importance. Imagination therefore appears as a cognitive faculty and process that ultimately values knowledge as a means of acknowledging uncertainty, and values epistemological attitudes of skepticism for the sake of greater social justice.

What we can learn about conceptual change […] is first that an alteration in the use of a language’s vocabulary is gradual, beginning with the change of language-games leading to a change in concepts and with that to a change in the meaning of words. […] [S]uch a change may often be accompanied by uncertainty, a vacillation between old and (various) new ways of expression, which can currently be observed in many languages, for example, in the introduction of gender-inclusive pronouns and concepts. (Trächtler 2023b: 10)

Relying on Burgess and Plunkett, conceptual engineering concerns “the assessment and improvement of concepts, or of other devices we use in thought and talk (e. g. words). This often involves attempts to modify our existing concepts (or other representational devices), and/or our practices of using them” (Burgess & Plunkett 2020: 281). Drawing on their assessment, conceptual engineering follows the activities of conceptual ethics, conceptual innovation, and conceptual implementation; conceptual ethics is thus defined as (i) encompassing normative and evaluative questions about concepts, words, and other broadly “representational” and/or “inferential” devices we use in thought and talk; and (ii) a concern of which concepts we should use and what words should mean, and why. I argue for conceptual engineering to be a successful way of countering conceptual injustice, (i) as it recognizes the need for conceptual normative and evaluative inquiry, (ii) as it is an effective and creative approach, closely related to what Murdoch argues as improving our concepts from a moral perfectionist view (i. e. bettering oneself through language): conceptual innovation involving the practice of introducing new concepts and/or amending existing ones and conceptual implementation involving a practice of getting people to employ the concepts one previously developed doing conceptual innovation.

At any rate, one might doubt whether conceptual engineering fully answers the question of what to do when a concept is flawed and unfair in regard, for example, to the concept of ‘woman’: should we no longer use the concept of ‘woman’ at all or should we keep it but expand it by including more and more prototypes within the ‘woman’ category? Both options present risks. On the one hand, it is not sufficient to want to ‘put an end’ to a use for it actually to stop, and on the other, there is the risk of inventing a new concept that might in turn be unable to take all subjectivities and experiences of womanhood into account, that might leave part of reality outside the sphere of language. Conceptual ethics as part of conceptual engineering must be, then, a theoretical approach embodied in the sphere of daily uses and practices of giving meaning. Murdoch says as much when she writes: “We learn language in contexts where our vocabulary is increased and (ideally) refined in the everyday processes of living and learning. We learn moral concepts. […] This is the very texture of being and consciousness woven and working from moment to moment in language.” (Murdoch 1992: 385)

Imagination in language is what Fricker proposes to counter the cognitive handicaps that result from hermeneutical injustice: we should not see a culture’s moral discourse as a “finite monolith”, but rather consider conceptual resources as resources that indefinitely generate many new meanings, whether new applications of old concepts or new concepts. These are generative and dynamic resources for meaning, never entirely absorbed by the set of meanings actually put into practice at any given moment in history. Therefore, with this generative conception of moral meaning, we are able to differentiate between routine discursive movements in moral discourse and more imaginative, one-off movements, in which existing resources are used in an innovative way that constitutes a progressive move in moral consciousness.

An example of such an imaginative move might be the extension of a notion of respect to all humanity as made by someone ethically formed in a society that routinely reserved respect for the ruling classes; or the new application of the concept ‘cruel’ to violent punishments of children routinely meted out in the name of ordinary discipline. Most of us most of the time make routine moves and exhibit routine moral thinking, and we are lucky if we live in a culture where this means that our ethical thinking is on the whole decent; but sometimes people can rise to a challenge and succeed in something more imaginative. (Fricker 2007: 104)

Through these cases of “imaginative moves” in conceptual engineering towards a greater goal of social justice, we may better understand how the ethics of attention to language might counteract the cognitive handicaps that arise from situations of conceptual injustice. But these “imaginative moves” cannot be ‘leaps’ into extraordinary intelligibilities, and are rather part of a daily refinement in language, something that can be done in literature, poetry, or art in general. However, both Murdoch and Wittgenstein would argue that, ultimately, for these movements to have any real social significance, for them to speak to us, and have meaningful substance we would have to agree in language and uses: this is the collective nature of language which both Murdoch and Wittgenstein acknowledge—again, this is not to say that there is no intimacy in language but there is nothing private. Nor does this mean that conceptual engineering alone is enough to achieve greater social justice, for the latter also requires agreement in society’s institutions and laws.^[13]

Conceptual ethics and a virtuous practice of linguistic imagination has, as Sterken argues, diachronic aspects. She talks about the (e. g. feminist) “ameliorator” interested in the communicative “long game”, i. e. searching for not just what is going on in her own communicative context or with the community of speakers that speak their shared language at the time of her utterance, but intending for her speech to eventually be understood as she wants it to be. With its ameliorated meaning, the ameliorator has indeed “diachronic communicative intentions”—communicative intentions are not relevant to her context but project into future communicative contexts and future linguistic communities (cf. Sterken 2020: 432).

4.3 The Ethics of Attention to Language as an Antidote to Conceptual Injustice

In Murdoch’s apprehension of moral life, “goodness consists in a constantly perfectible apprehension of a perpetually receding reality” (Panizza 2015: 12). Murdoch’s question—how can we be morally better?—is not merely a question for her philosophy but also for a philosophy of language: how can we better ourselves within language? It implies that the process by which we can reconnect with conceptual resources requires transformation and moral effort: “We are all the time building up our value world and exercising, or failing to exercise, our sense of truth in the daily hourly minutely business of apprehending, or failing to apprehend, what is real and distinguishing it from illusion.” (Murdoch 1992: 304). By its very nature, attention pertains to something external to oneself: attention does not produce its results instantaneously but rather is part of a gradual, “piecemeal” enterprise of moral growth. As Silvia Panizza puts it in The Ethics of Attention, the moral concept of attention involves “particular epistemic attitudes and faculties that are meant to enable the subject to apprehend moral reality and thus achieve correct moral understanding and moral response” (Panizza 2015: 2).

Applied to language, attention is thus articulated as a cognitive penetration of perception. It involves understanding what impedes a correct perception of reality and understanding the flaws in our cognitive resources. From a moral standpoint, it invites to understand how a perceiving subject, if he does not turn away from self-concern, is likely to get caught up in fantasies and illusions. Panizza shows for example how fantasy, an epistemic-moral concept, is a distortion of reality that we are responsible for but not necessarily, and not often, conscious of. In this respect, an ethics of attention to language is precisely about understanding our responsibility in language, in those places where we are not conscious of the injustices that are being played out. Murdoch writes: “We should feel socially responsible about what in our society people always or never see.” (Murdoch 1992: 329) An ethics of attention to language is, thus, a redirection of consciousness within language and to what is actually happening within it.

Some might say, provocatively, that when Murdoch redefines a concept such as the one of ‘freedom’ as “a small piecemeal business which goes on all the time and not a grandiose leaping about unimpeded at important moments” (Murdoch 1970: 36), she merely manipulates language: how is that ethical? Where does trust lie? How can conceptual engineering in its ethical dimension, i. e. getting as close as possible to the truth, prevent injustices that can always return? From fake news to wokism, the least that one can say is that contemporary political field is far from lacking in examples showing the complex aspect of conceptual engineering. Murdoch would answer her critics exploring the fears she feels about the authoritarian nature of following rules in language, which does not necessarily follow the criteria of what is right or good, in this way:

How does meaning connect with truth? One language can be more potentially truth-bearing, more precise, more beautiful, richer in concepts than another. Tyrants destroy language, diminish vocabulary. A language is enlarged, improved (value judgement), by truthful utterance. People suffer and are damaged if prevented from uttering the truth. Assent, general agreement, has a background which must be scrutinized. Is there a reason why a despotic state could not be Lebensform? Any Lebensform may be subject to moral judgement.

[…] Why is Wittgenstein so anxious to set up this machinery which so pointedly excludes the individual peculiarity of speaking humans? And if it appears that we cannot accept this picture are we rejecting one of his fundamental tenets? Of course language depends very generally upon areas of ‘agreement’, but is also continuously lived by persons. Fine shades of behaviour, imponderable evidence, looks, glances, gestures, tones, whistling. (Murdoch 1992: 281)

This passage can be understood in two different ways: on the one hand, as an attempt to restore the place of the “person” within moral philosophy, to restore the importance of particularities and differences of perception that constitute the richness of the world, and on the other, as an endorsement of Wittgenstein’s theory that concepts are above all a matter of agreement in language. How exactly do we carry out a linguistic investigation, then? To echo John L. Austin: we look at what we should say when (cf. Austin 1979: 182). James Conant defines a grammatical investigation as “a convening of our criteria for the employment of a particular concept” (Conant 1998: 249). To use criteria is to exercise judgement—the judgment that this is the word to use, the sentence to utter in these specific circumstances. The underlying idea is one of discriminating or separating cases, of identifying by means of differences (cf. Cavell 1999: 17).

In The Sovereignty of Good, Murdoch introduces the well-known example of M&D (mother and daughter-in-law), often borrowed to grasp her phenomenology of attention or how inner life is a site of moral reconfiguration and moral perception (cf. Jamieson 2020).

A mother, whom I shall call M, feels hostility to her daughter-in law, whom I shall call D. M finds D quite a good-hearted girl, but while not exactly common yet certainly unpolished and lacking in dignity and refinement. D is inclined to be pert and familiar, insufficiently ceremonious, brusque, sometimes positively rude, always tiresomely juvenile.

[…] Time passes, and it could be that M settles down with a hardened sense of grievance and a fixed picture of D, imprisoned […] by the cliché: my poor son has married a silly vulgar girl. However, the M of the example is an intelligent and well-intentioned person, capable of self-criticism, capable of giving careful and just attention to an object which confronts her. M tells herself: ‘I am old-fashioned and conventional. I may be prejudiced and narrow-minded. I may be snobbish. I am certainly jealous. Let me look again’. Here I assume that M observes D or at least reflects deliberately about D, until gradually her vision of D alters. If we take D to be not vulgar but refreshingly simple, not undignified but spontaneous, not noisy but gay, not tiresomely juvenile but delightfully youthful, and so on. And as I say, ex hypothesi, M’s outward behaviour, beautiful from the start, in no way alters. (Murdoch 1970: 16 – 17)

For Murdoch, what M is trying to do is to see D justly and lovingly. It suggests a new picture of freedom no longer as the sudden leap of the isolated will in and out of an impersonal logical complex but a characteristic of the progressive attempt to see a particular object clearly. To sum it up very roughly: M’s mental activity is necessarily progressive, and continuously perfectible. The example introduces the idea of perfection and moral vision. But it seems to me that analyses too often neglect the linguistic dimension in this example. That is, the moral vision, the attentiveness, do not unfold outside a choice to change in the vocabulary: it is a case of attention to language. Murdoch’s conceptual engineering is first and foremost a form of moral perception and cognition driven by attention. It means feeling that something is obscure in experience, and needs to be groped out patiently. It is therefore also potentially subject to doubt, insofar as skepticism is endowed with epistemic virtues:

There are, however, moments when situations are unclear and what is needed is not a renewed attempt to specify the facts, but a fresh vision which may be derived from a ‘story’ or from some sustaining concept which is able to deal with what is obstinately obscure, and represents a ‘mode of understanding’ of an alternative type. Such concepts are, of course, not necessarily recondite or sophisticated; ‘hope’ and ‘love’ are the names of two of them. (Murdoch 1956: 51)

For Murdoch, concepts determine moral configurations of the world. This implies adherence to the view that facts and values are inextricably intertwined, i. e. that value follows two movements: on the one hand, a movement on the part of the moral agent (i. e. the moral gaze that attaches value to the facts of the world), and on the other, a reciprocal movement whereby value is also in the things and facts that call my gaze. Thus, “[f]or purposes of analysis moral philosophy should remain at the level of the differences, taking the moral forms of life as given, and not try to get behind them to a single form. I suggested above that ethics had in the past, in one of its aspects, been continuous with the efforts of ordinary moral agents to conceptualise their situations.” (Murdoch 1956: 58) Seeing aspects, differences, details, i. e. a whole philosophy of attention that redefines conceptual activity as attention to what counts, has counted up to now, and should count in the future, not only for the sake of social justice but also for continuous moral betterment. To follow up on the M&D story, conceptual change is not just a difference of vision, but it is a choice within the sets of concepts that one possesses. Attention to language thus implies conceptual learning (cf. Murdoch 1970: 30 – 31). Such is the path that Murdoch envisages, i. e. the ability for a concept not to reduce vision but to extend it. If we consider language and conceptual activity, we learn about the differences in human life through our attention to contexts, our vocabulary develops through fine-tuned attention to objects and people; and regarding these “imaginative [linguistic] moves” (Fricker 2007: 104), that is where art and literature are of great help, forming themselves as ethical experiences.

The ethical dimension of attention to language emerges within the willingness to see things in a way that comes closest to being truthful. Conceptual engineering therefore appears in this philosophy as a faculty of discernment (cf. Murdoch 1970: 88) put to work by attention but also as a kind of cognitive penetrability of perceptual experience (cf. Panizza 2022: 71 – 73). Concepts are evaluative instruments, but it does not mean that reality simply equates with each individual’s grasp of it or that each person has a set of purely private concepts: rather, as in Wittgenstein’s conceptual thinking, the application of concepts to specific situations is subject to public rules, and concepts themselves are not made up by the individuals, but learnt publicly with other people, within public contexts of use, in relation to specific objects or situations (cf. Panizza 2022: 73 – 74). As Murdoch says, “words vanish or alter their meaning ‘naturally’ in the course of language-use and social change” (Murdoch 1992: 327). It is all about learning, and perceiving justly. Like Wittgenstein, Murdoch argues for the very texture of being and consciousness as woven and worked from moment to moment in language. We are constantly engaged in understanding meaning through speaking, but paradoxically it is through speaking that we give meaning to what we say, and therefore see.