Philosophers have hitherto only interpreted the world in certain ways; the point is to change it.
Karl Marx, Theses on Feuerbach (1845), eleventh thesis1. Introduction
One of the core issues brought about by the “social turn” in present-day analytic philosophy concerns the concept woman.Footnote 1 This issue has important consequences for feminist philosophy regarding the problem of identity and whether the concept woman may be applied in such a way that one's gender may not be coincident with the sex one was assigned at birth; whether it may be applied to trans women contrary to what trans-exclusionary accounts of gender maintain; and how it relates to non-binary accounts of gender, which are not only adding new categories but also reshaping already existing ones and the dichotomy woman/man. The paper presents a Wittgensteinian, family resemblance account of the concept woman (§3). This approach is deemed superior to the Carnapian account, championed by Sally Haslanger (§2), on several fronts. Furthermore, it allows to locate the source of sexism not in the concept woman itself, but in the beliefs sexist people have about women. The compatibility of this account with semantic externalism is then explored, as well as its relationship with stereotypes and paradigmatic examples of women (§4). It is further shown how present-day attempts to extend the application of woman beyond those who were assigned female sex at birth may be made sense of in terms of a change of Wittgensteinian “hinges.” Hinges have the form of empirical propositions yet play a normative role (at least in context). Still, they may change according to context and over time. It is argued that hinges, understood as rules of evidential significance, which a community of speakers brings to bear on the determination of whether a person is a woman, are currently being revised under the pressure of social activism in favor of queer and transgender identities (§5). Indeed, such a revision is of a piece with a battle against forms of “hermeneutical injustice” which, in the case at hand and contrary to Miranda Fricker's (Reference Fricker2007) original characterization, are brought about by the presence in our shared conceptual repertoire of faulty concepts, rather than by a lack of conceptual resources. Finally, it is argued that, despite Wittgenstein's metaphilosophical quietism, such an account is politically significant. Hence, Wittgenstein's metaphilosophical pronouncements, far from being inimical to political and societal change, may in fact be conducive to it (§6).
2. The Carnapian, top-down approach
What are we talking about when we say that a person is a woman? According to ameliorative projects, to answer that question we should arbitrarily pick a definition, which doesn't have to match our ordinary use of the term “woman” (whatever that might be), or what we normally take to be the concept woman.Footnote 2 Moreover, this definition is typically politically charged to better expose the kind of social injustice perpetrated against women, with the hope that in the future there won't be any subjects falling under that concept. Notoriously, this is the approach taken by Sally Haslanger (Reference Haslanger2012, 230) who defines woman thus:
S is a woman iffdf S is systematically subordinated along some dimension (economic, political, legal, social, etc.), and S is “marked” as a target for this treatment by observed or imagined bodily features presumed to be evidence of a female's biological role in reproduction.Footnote 3
Haslanger's approach is distinctively Carnapian. Accordingly, we may conventionally propose new definitions of the terms/concepts under consideration, if they serve our emancipatory purposes better than what one might think is the way in which they are standardly used or deployed. The Carnapian approach is typically top-down, and was envisaged for scientific concepts.Footnote 4 By explicating a given concept/term, which may be already present in ordinary practice/language, we would then fasten onto some more specific concept/meaning that would make for a theory that is explicative and fruitful.
When it comes to the domain presently under investigation, however, things seem much more complicated for the Carnapian approach. For we are not trying to do social science, thus explicating concepts in the service of building theories that should be empirically testable, like sociological ones. Rather, as philosophers, we are trafficking in meanings and concepts available to us qua members of a given community, with the hope of bringing about certain societal changes. Thus one might wonder what the point of this top-down approach could ever be.Footnote 5
If the concept woman people have in our society is different from the one Haslanger proposes, it would be difficult to claim that by getting rid of women, in this second sense, we are improving the lives of women in the first sense. For one could easily be accused of equivocation or of wishful thinking. Of course, one might say that at least some women fulfill the definition proposed by Haslanger, and if the day comes when there won't be any women in Haslanger's sense, it will be an improvement.
Yet, counterintuitively, non-subordinated women (if there are any) would not be women, according to Haslanger, since the only relevant concept woman is the ameliorated one.Footnote 6 Moreover, there is no reason to go through such a re-engineered definition of “woman” to improve the lives of subordinated women. For it will be enough, first, to note that among women, ordinarily understood, many are de facto so subordinated. Second, we could try to improve the lives of these women, with the various means that feminists have normally advocated for to that end.Footnote 7
Alternatively, one will have to say that the definitions, meanings, or concepts proposed by the theorist are implicitly operative in our linguistic community.Footnote 8 This might give Carnapians some hope that their project could engage with reality. Yet, many—including hard-core feminists—would baulk at the idea that it is analytically true that women are individuals who ought to occupy subordinate positions in a society (see the revised definition of “woman” in Haslanger Reference Haslanger2012, 234).Footnote 9 It may be true, unfortunately, that many women do occupy such subordinate positions, or that sexists think they ought to. But this does not seem to be necessarily part of the concept woman. After all, it isn't a contradiction in terms to think that women may not and certainly should not be so subordinate. At most, thinking they should be is part of the sexist beliefs and practices that surround women and shape their lives.
A terminological remark is apposite here. As is standard in philosophy of mind, beliefs are propositional attitudes of acceptance of certain contents—or propositions—the grasp of which requires concepts. Only semantic holism holds that all beliefs containing a given concept C are constitutive of it. Semantic holism, however, dissolves languages and conceptual schemes into idiolects and has serious difficulties in making sense of substantive disagreement. More importantly, for present purposes, it is incompatible with the Carnapian approach, according to which concepts are individuated by means of a finite set of necessary and sufficient conditions. Thus, while it may be a true belief about women that many are subordinate, or even that sexists think they ought to be, this by itself is no sign that our concept woman—or even the best concept possible given feminist purposes—would have to include in the definiens reference to the subordinated role within the society that many women have or are expected to have by sexists.
Alternatively, we might interpret the Carnapian project as follows: (sexist) people tend to think that there is something belonging to a woman's essence that makes her naturally socially subordinate due to her biological role in reproduction. Let the metaphysical constructivist be interested in debunking this idea. One may then say that that is the concept woman sexists have. In a Carnapian spirit, then, one might think that sexists are projecting their faulty concept onto reality. Yet the idea that there is something like a woman's essence and that it is just as described, would be due to mistaking a semantic fact for an ontological one. Once we are clear about such a conflation, we should be able to debunk the idea that there is something like a woman's essence.
Yet one could raise the following criticism: let us admit that the sexist concept woman is as described above. Still, instead of taking for granted that it is vacuous, we should subject it to serious empirical scrutiny, just as we did, in the past, in the case of phlogiston, say. We should not commit the moralistic fallacy of thinking that simply by finding such a concept morally reproachable we will have ascertained that it isn't instantiated. Since many women, alas, may be taken to fall under that concept, it will be easy for sexists to point at these cases and claim that that is what “real” women are like.
Furthermore, simply by exposing the sexist concept as leading to a pernicious reification of sexist beliefs one would not have improved the lives of women—even in Haslanger's favorite sense of the term. For, clearly, much more would have to be done to improve women's lives. Thus, the political significance of such a move would be extremely debatable.
True, it would help raise awareness but this would not be sufficient for espousing the account. For awareness may be—and indeed has been—raised by feminists by exposing the fact that many women are so subordinated. This awareness would consist in acquiring a true belief about the social condition of many women and would not itself depend on making that social condition definitional of being a woman. Moreover, making it definitional of woman would hardly be helpful even from a political point of view. At the very least it may induce the impression that just by exposing the sexist concept something would thereby have been done to improve the lives of women.Footnote 10 Thus, I doubt this is the best way of going about the project of bettering society.
Nevertheless, I am keen on improving society. And my question is, what can philosophers interested in issues such as meanings, concepts, and epistemic practices do, qua philosophers, to foster the improvement of our society? Here I would like to explore an alternative approach to the Carnapian top-down approach embodied in Haslanger's work, inspired by Wittgenstein's notion of family resemblance.
3. The Wittgensteinian bottom-up approach
3.1. woman as a family resemblance concept
The Wittgensteinian, bottom-up approach takes a more descriptivist line than the Carnapian one. For it does not propose that philosophers should revise existing concepts, either to pass them on to society, as it were, or to better expose forms of subordination or discrimination present in society against some of its members. Rather, it urges us to pay attention to the ordinary, to the complex texture of our forms of life and our lived and embodied experiences and eschews abstract generalizations which inevitably cannot do justice to the complex vagaries of life. “Don't think, look!” (1953, 66) is Wittgenstein's famous injunction.Footnote 11
Yet there are also elements in common with the Carnapian approach; namely, the idea that meanings are arbitrary, and that issues that seem to take the form of metaphysical questions into the nature of reality—such as “What is a woman?” “What is race?”—are better approached by looking at the workings of the corresponding words/concepts.Footnote 12 For these meanings and concepts are not set in stone. They are the result of certain social conventions, and they may change over time. Hence, the Carnapian and the Wittgensteinian approach are both anti-essentialist projects. Still, these projects work quite differently.
First, according to the Wittgensteinian, bottom-up approach concepts or meanings are not simply determined by explicit definitions. Sometimes, of course, they are—like in the mathematical, the scientific, or in the legal context. Yet when we are interested in the terms and concepts that people in our societies use in their everyday interactions, then that approach is often unavailable. The way in which people typically use these terms is rather through family resemblance. Hence, the proposal I want to examine is to consider woman as a family-resemblance concept.
Accordingly, woman wouldn't be individuated through a set of necessary and jointly sufficient conditions. Rather, there would be certain individuals that are already considered instances of that concept and then other individuals would be taken to fall under it, if they share one or the other of the features of these cases, but not necessarily either all or a (weighted) majority of them. Family resemblance, therefore, is neither a prototype, nor a cluster theory of concepts.Footnote 13
As Friedrich Waismann explained the notion of family resemblance, borrowing Wittgenstein's example of game: “it may be merely that every two games are connected by intermediate links” (Reference Waismann1997, 82). That is, like tennis is a game which cannot be played alone and where there is winning and losing, so is solitaire, where there is winning and losing but must be played alone, and so is playing with dolls, which can be played alone and where there is no winning or losing. Solitaire is therefore the intermediate link that connects tennis and playing with dolls, even though these latter games have nothing in common.
This explains the relevance of intermediate links for Wittgenstein (Reference Wittgenstein1953, 122). Like the intermediate links in a chain connect separate links that do not touch each other in any point, intermediate instances of a concept C serve to unite instances that may have nothing in common between them, and they do so by means of having (at least) one feature in common with each of these separate instances, though not the same one(s).
This is important for several reasons. First, it avoids the idea that one could count as a woman only if certain (clusters of) traits are shared, and connectedly that there would be individuals who would be more of a woman than others if they shared more of these traits or more of the “heavy weight” ones.Footnote 14. Think again of the links in a chain: they are all on a par and none of them is more “central” than any other one, in and of itself.
Secondly, family resemblance clarifies in what sense concepts are open-ended (Waismann Reference Waismann1945 would have said “open-textured”). Since inclusion within a concept does not depend on sharing a set of predetermined features, new cases that share with one old member of the class a feature that they do not share with any other member of that class can be included within it. Thus, one important aspect of family resemblance is that it allows for the extension of the concept to new cases, in potentially unpredictable ways.
Thirdly, family resemblance should not be conflated with vagueness. That is, the proposed account of woman does not depend on the fact that biologically there may be human beings that fall in between male and female, or have some biological features of both, such that the concept woman would not be a clear-cut one.Footnote 15 Nor does it depend on noticing that one's gender and one's sex can come apart, so that it may not be fully determinate whether one falls under that concept. The point, rather, is that if woman works by family resemblance, the criteria for inclusion within the class of women—however subject to vagueness they might be—may expand to become more inclusive.
Finally, the concept woman, on this proposal, is more like a space-time worm than like a tridimensional individual. Hence, while evolving in time, it remains one. That is, no replacement of meaning/concept is affected simply by adding new criteria for membership within the concept. Rather, the concept woman would be a potentially open-ended, revisable disjunction of the various features that have entered the concept through time. Our definitions, therefore, would only be snapshots of moments in the life of the concept—corresponding to (sometimes overlapping) temporal stages of one single concept. Thus, we should not mistake the concept in its entirety with any of its temporal stages. Furthermore, the open-endedness of the concept leaves room for its extension to new cases which are not presently considered simply because they may not be salient at this time in history but may become so later. Thus, it is not inconceivable that, due to the development of AI and robotics, in the future the concept woman (or man) will have to be extended to non-biological creatures. The relevant point is that these new cases will be subsumed under the same concept (woman or man), rather than under a different, partially related one.
Now, conceptual continuity and change is a problem for the Carnapian approach because, as soon as we change the set of necessary and sufficient conditions, we change the concept. As we saw, if a woman is not subordinate, she is not a woman in Haslanger's sense. Nor is she one if she lacks all the physical features presumed to be evidence of females’ biological role in reproduction, given Haslanger's definition of woman. This might have the consequence, for instance, that some trans women would not be women.
For a Wittgensteinian the problem is easier to handle because, to repeat, there won't be a change of concept—that is, a replacement of the original one with a new one—by allowing some more individuals to fall under it, which were not initially considered part of its extension. Rather, that very concept will simply be extended to new cases, because the criteria Footnote 16 for inclusion within the class determined by the concept will be expanded. Similarly, the very same concept would be changed, just like an individual, thought of as a spatio-temporal worm, can change its properties while remaining one and the same.
Moreover, as we saw, the Carnapian approach can easily be accused of changing the topic by substituting certain definitions of contested terms with different ones.Footnote 17 And then the question would be how changing from a concept woman to a different one could ever be taken as ameliorating the initial concept. At most, it would count as a salutary replacement of the initial concept. On the Wittgensteinian approach, in contrast, we could still have the same concept woman and yet ameliorate it, in the sense that, for instance, we would be expanding it to new cases which were not initially considered to fall under it.Footnote 18
3.2. Amelioration in Wittgenstein and “serious politics”
According to the Wittgensteinian model, conceptual amelioration does not work by having a schoolmasterFootnote 19 dictate a new meaning for our old words, or by promoting the replacement of one concept woman with a different one, in hopes that one day there won't be any more women in this latter sense. Rather, the Wittgensteinian bottom-up approach enjoins fighting on the rough ground of “serious politics,” as Naomi Scheman (Reference Scheman2011, 17–18) calls it,Footnote 20 so as to have women who are not subordinate or to include transgender women, should they so wish.Footnote 21 If we want our concepts to be more inclusive, we need to “shift the ground”, as Scheman puts it (2011, 35), from which they arise.
Still, the extension of the relevant concepts to new cases is possible only thanks to changes in the social and political world, effected by serious political activism, where the agents are direct stakeholders and those of us who share the hopes for a better future for marginalized members of our society. This mode of amelioration is thus of a piece with a democratic, non-authoritarian, and non-paternalistic conception of how society works and improves itself.Footnote 22
Moreover, it is only (or mostly) by going bottom-up—that is, by acting on the rough ground of the material living conditions of women (including trans women), and improving those—that we can concretely bring about positive change. Conversely, just by exposing the fact the concept woman sexists have is as proposed by Haslanger we would not have done anything useful to help improve the lives of those individuals who fall under it.
Thus, the aim of contemporary analytic feminism should not be one day to do without the concept woman, after having it re-engineered the way Haslanger suggests. Nor is it really to substitute it with a new and hopefully better one, following a top-down approach. It is rather, one day, to have more and happier women—that is, more individuals subsumed under that concept, who are not subordinated to men and are therefore more satisfied.
4. Externalism and objectivity on a Wittgensteinian approach
As previously noted, there is a common element to the Carnapian and the Wittgensteinian approach, which is their anti-essentialism, if by “essence” we mean a natural essence. Thus, both the Wittgensteinian and the Carnapian approach do not take woman to pick out a natural—biological—essence that all and only individuals falling under that concept should allegedly share.
For a Wittgensteinian, moreover, the meaning of a term and the way in which a concept is identified depend on use. Since use “ain't in the head,” let alone in the head of individual speakers, and since, for Wittgenstein, the very idea of a private language is not just false, but altogether incoherent (1953, 244–70), the proposed account of woman would count as externalist. What would be alien to this kind of externalism is the “mysterian” aspect of some externalist positions whereby even the whole community may not know what woman refers to. If such a consequence of some externalist positions may be acceptable when we are dealing with natural kind terms and concepts, it is much less so where social concepts, which depend on our making, at least at some level, are concerned.
In an interesting passage Haslanger (Reference Haslanger2012, 374) writes:
Externalism is an option whenever there are relatively objective types, the notion of objective type needed is not too mysterious. A set of objects is more an objective type by virtue of the degree of unity among its members beyond a random or gerrymandered set. We might account for unity in various ways, but a familiar way, I'll assume for current purposes is in terms of degrees of similarity; the similarity in question need not be a measure of intrinsic similarity, that is, things can be similar by virtue of the relations (perhaps to us) they stand in.
Objective type externalism: terms or concepts pick out an objective type, whether or not we can state conditions for membership in this type, by virtue of the fact that their meaning is determined by ostension of paradigms (or other means of reference fixing) together with an implicit extension to things of the same type as the paradigms.
And then she adds: “we should not assume that objectivity is only found in the natural world. There are objective types in every realm: social, psychological, political, mathematical, artistic, and so on” (2012, 375).
Now, what is interesting in these passages is not only that a Wittgensteinian account of woman could count as externalist, but also that, even though family resemblance is not a prototype theory of concepts, there would be members falling into the category that will share more than one feature between them (or “heavier” ones). Those would be, I take it, what Haslanger calls the “paradigms.” Thus, we typically point to, say, a dog or a cat, when we want to exhibit a paradigmatic case of mammal, and not to a dolphin or a whale.
What is interesting, and it has been notoriously used in psychology by Eleanor Rosch and her lab, is the idea that while there may be paradigmatic cases of mammal, etc., our concepts extend to cases that may be further removed from the paradigmatic ones (Rosch and Mervis Reference Rosch and Mervis1975). When tested, subjects would show longer reaction times in categorizing these elements that are removed from paradigmatic instances of the relevant concepts. Thus, there is a testable empirical explanation of why we cognitively operate as we do—that is, for instance, by being more hesitant to categorize a trans woman as a woman. Such an explanation clearly shows the relevance to our cognitive performance of paradigmatic cases, with the attendant difficulty of recognizing new and somewhat different elements as falling under a given concept.
Yet paradigms are context sensitive and can change over time, precisely in virtue of the open-endedness of our concepts, as predicted by the family resemblance approach. Thus, political activism should work towards more and more inclusion to also foster a change in the prototype associated with woman.Footnote 23
From a theoretical point of view, this teaches us that those prototypes and paradigms, while cognitively relevant, aren't concepts. The latter are normative entities that determine when it is correct or incorrect to consider an individual as an element of a class, irrespective of whether it fits the prototype associated with it; whereas the former are mental representations that explain how de facto we operate, from a cognitive point of view, when engaged in certain specific tasks, like quickly recognizing elements belonging to a certain class, or when we must make fast and unreflective choices or inferences.
Finally, in keeping with constructivist approaches, we should deny that the objectivity of the prototype—that is, its being cognitively operative within multiple individuals’ minds—corresponds to an essence. For that would give the game away, if by “essence” we mean a natural or a social essence. There is no social essence, or role (or kind), I believe, corresponding to being a woman, let alone a natural one. Rather, the roles that women can play in a society are continuously renegotiated. Feminists have been vocal in advocating for these kinds of change. Yet, in a sense, this would be strictly speaking impossible, if women were taken to share a social essence.Footnote 24 For how could an entity change its own essence, if it really had one?
The family resemblance approach, as presented in §3.1, clearly dispenses with that as well as with the idea that individuals falling under the concept woman would have to share a cluster of properties. One might object, however, that the account would make it too easy to be included within the extension of the concept; or, alternatively, that the concept would not pick out any “real” or “objective” kind.
The former objection can be countered by noticing—indeed in keeping with the Carnapian, pragmatic conception of concepts—that our purposes and needs will de facto drive the process and it is extremely unlikely that they will bring about the inclusion into the class of women of any possible entity as long as it has something in common, albeit not the same feature, with some individual already included within that class. The latter objection, in contrast, is no objection, because the whole point of adopting the proposed account of the concept woman is to dismantle the idea that there is a real kind—be it biologically or socially functionally determined—that our “real”, or “true” concept woman should pick out.
5. Hinges and hermeneutical injustice
Let us now turn to another important aspect which hasn't been properly considered in the literature. Namely, the role of hinges in relation to epistemic injustice (Fricker Reference Fricker2007)—especially, hermeneutical injustice.Footnote 25
To that end, it is important to consider a problem in Fricker's understanding of hermeneutical injustice. According to Fricker, hermeneutical injustice arises when there is a lack in the conceptual resources belonging to a certain community, so that experiences had by some members of the community are precluded from collective and personal understanding. This kind of lack of conceptual resources within the community is kept in place by power structures that are afflicted by identity prejudice against those members.
The example she gives is sexual harassment. Actions that later would be considered instances of it were present well before that concept was available to society. However, prior to its introduction, members of the society who were victims of acts of sexual harassment could make sense neither to themselves, nor to others, of the harm they had undergone. Furthermore, there was no interest in having such a kind of harm recognized by the dominant members of the society, who would actually benefit from downplaying the significance of those actions and from harboring several prejudices against women—both with respect to there being a real harm inflicted to them by means of various actions that would later be classified as cases of sexual harassment, and with respect to the subordinated role of women in relation to men.
However, there are also cases in which hermeneutical injustice arises differently. I have argued for that elsewhere (Coliva Reference Coliva2024b), by looking at the persistence within our conceptual resources of the concept hysteria, which has been shown to be a flawed one, and consequently was withdrawn from DSM III in the 1980s. For it was realized that hysteria didn't pick out any real disease, let alone one only creatures with a uterus (given that hystera in Greek means “uterus”) could suffer from. The case of hysteria, therefore, shows that hermeneutical injustice arises also when there are, available in a society, faulty conceptual resources that are kept in place due to identity prejudice.
Mindful of such cases, we should then re-engineer the very concept hermeneutical injustice to include not only cases in which this injustice arises owing to a lack of conceptual resources in a society, due in its turn to power structures that discriminate against certain individuals, but also when it arises because of the presence of faulty conceptual resources that are kept in place by discriminatory power structures.Footnote 26
With this re-engineered understanding of hermeneutical injustice in mind, if we reflect on the claim “I am a woman” and on the concept woman from this point of view, we can see how the traditional, conservative criteria for inclusion under the concept woman, that block the extension of that concept to new cases, are not allowing some members of our society fully to make sense of their own experience to themselves and others. Consequently, sticking to conservative criteria for falling under woman perpetrates a form of hermeneutical injustice.
Notice that the Carnapian approach is often considered politically more salient than the Wittgensteinian one because it would allow choosing how to explicate a concept based on pragmatic factors, broadly construed to include also political ones. Ameliorative projects are in fact predicated on this feature of the Carnapian approach. Given what we have been saying so far, however, the Wittgensteinian approach would be equally open to pragmatic and political encroachment.
To see how hinge epistemology comes into the picture, it is important to keep in mind that hinges have the form of empirical propositions, yet play a normative function, at least in context. For they can be (1) norms of evidential significance, that is, what needs to stay put for certain evidence to be brought to bear onto the semantic assessment of ordinary empirical propositions; and/or (2) meaning-constitutive norms, often embodied in judgments which serve as paradigmatic instances of correct applications of a given term.
Now, we can interpret discussions and political battles concerning woman as cases in which there is a conflict between hinges that would then impinge on what would count as a paradigmatically correct application of that concept, as well as on what counts as evidence for being counted as a woman. Interestingly, Wittgenstein, in On certainty (1969, 79–83) considers the correlative case of “I am a man” as an example of a hinge. He writes:
That I am a man and not a woman can be verified, but if I were to say I was a woman, and then tried to explain the error by saying I hadn't checked the statement, the explanation would not be accepted.
The truth of my statements is the test of my understanding of these statements.
That is to say: if I make certain false statements, it becomes uncertain whether I understand them.
What counts as an adequate test of a statement belongs to logic. It belongs to the description of the language-game.
The truth of certain empirical propositions belongs to our frame of reference.
Hence, “I am a man,” for Wittgenstein, is a hinge. Even if it looks like an empirical proposition—a description of a contingent state of affairs, about a specific individual, which could be empirically verified—it actually plays a rule-like role (1969, 95), and in this specific sense is part of what he calls “logic”. Playing such a rule-like role, for him and his community, neither Wittgenstein nor his community would have considered anything as proof to the contrary.
What is relevant to our present purposes is the fact that Wittgenstein takes it for granted that it is mutually exclusive whether one is either a man or a woman. That is, he operates with a conception of gender as coincident with sex as assigned at birth based on one's observable physical characteristics. Yet it is part of his philosophical account that hinges can and do change over time. For instance, when he was writing, “Nobody has ever been on the Moon” was a hinge, while it is no longer one for us.
Hence, the interesting point of On certainty is not whether the myriad hinges listed there are the ones we would nowadays subscribe to. Rather, what is relevant is the conception of their nature and role as propositions of the form of empirical ones, which, while playing a normative role with respect to what is to count as evidence for what, and/or with respect to what is to count as a correct application of a given word, may change over time and according to context.
Wittgenstein's hinge—which is still quite operative in our society, alas—excluded other forms of evidence—especially, how one feels about one's own gender—from playing a role in answering the question “What gender am I?” (and vis-à-vis the corresponding third-personal questions). Non-cisgender people no doubt had certain characteristic experiences even back then, but those experiences did not count towards determining the answer to those questions, at least in the eyes of their community. Since, for Wittgenstein, “what counts as an adequate test of a statement belongs to logic” (1969, 82), precluding one's own experience from being an adequate test of whether one is a man or a woman, amounts to normatively excluding that evidence from playing an epistemic role in the determination of one's own gender.
Armed with this idea we can now look at current debates about the extension of woman to trans women, and/or about non-binary identities as debates about renegotiating the hinges—that is, the rules of evidential significance—that will have to be operative when we want to answer the previous questions. Connectedly, such a hinge disagreement will involve which judgments would count as paradigmatically correct applications of that concept.
That is, just as the (de jure) hinge “There is an external world” (or There are phyiscal objects) needs to stay put for sensory—and, therefore, mind-dependent—evidence to be taken to bear onto the semantic evaluation of a proposition or a belief about a mind-independent object such as “Here is a mug,” say, so the (de facto) hinge “Gender is solely determined by biological factors as manifested at birth” needs to stay put for the observation of genitalia, say, to count towards the determination of whether a person is a man or a woman, while excluding a different kind of evidence, like one's self-identification, as bearing onto such a determination.Footnote 27 And just as “This is my hand,” in a Moore-like scenario would count as a paradigmatically correct application of “hand,” so “I am a man,” in Wittgenstein's own example, would count as a paradigmatically correct application of “man,” while excluding judgments based on self-identification as counting as correct ones.
Thus, we see that on this ground there is a battle between, on the one hand, people holding on to the conservative hinge that gender is binary and that whatever gender one belongs to is determined by others at birth, based on certain observable bodily characteristics of the individual in question; and, on the other, people holding on to the progressive hinge “Gender is determined by the way in which subjects self-identify.”
Importantly, we are already operating differently from an epistemic point of view than in the past. At the time when Wittgenstein was writing, the conservative hinge was operative, and there was no practice of asking people which pronoun others should use to refer to them. Nowadays, this has become standard practice, at least in institutions of higher education like universities, particularly in North America. Thus, the way in which people self-identify is more relevant than whatever sex they may have been assigned at birth by other people based on certain biological characteristics, no matter whether their physical appearance conforms to our existing stereotype.
Thus, allowing for personal experiences to count as evidence in the determination of one's own gender can be likened to a case of hinge-change, which goes together with the extension of the concept woman to individuals previously not considered to fall under the concept. Such an extension, moreover, may allow for marginalized people to make better sense of their own identity to themselves and others. This, in turn, provides a tool to overcome the kind of hermeneutical injustice that conservative applications of the concept, with their attendant heteronormative and scientistic hinge, perpetrate.
6. Quietism, activism, and imagination
Finally, philosophers moved by the desire to ameliorate society, qua philosophers, find Wittgenstein's alleged quietism—that is, the view that for him we can only describe our communal use of words, and that philosophy should leave everything as is—inimical to their projects.Footnote 28 For it would not be for philosophers to engage in these battles, qua philosophers, and there wouldn't be anything specific they could bring to the cause of ameliorating society.
I doubt that this is the correct description of what Wittgenstein was after. His quietism, as I have argued elsewhere (Coliva Reference Coliva2021), is much more apparent than real. If he was against theory, it was only in a specific sense of “theory,” mostly to be found in metaphysics and in foundationalist projects in epistemology. Roughly, Wittgenstein was against the kind of theory, as applied to philosophy—not to science—that vertically tries to ground more complex truths on a limited set of more fundamental ones, and that aims to unveil the hidden essence of phenomena. He was not against a kind of theory, in philosophy (but also in anthropology or sociology) that, horizontally, tries to map and give us a perspicuous and synoptic presentation of the complex interrelations of our concepts and that teaches us differences which may go unnoticed under the superficial similarity of the words we use. Furthermore, Wittgenstein coined several philosophical notions, such as “family resemblance,” “language-game,” “form of life,” “hinge proposition,” “world-picture,” etc. which, albeit not explicitly defined by him, serve the theoretical purpose of understanding phenomena such as meaning, concepts, rules, justification, knowledge, etc., and, as I have been arguing in the previous sections, have significant political consequences.Footnote 29
I have already extensively expounded on why the notions of family resemblance and of hinge proposition can play a role at a theoretical level to help us make sense of conceptual change, especially with respect to contested concepts, like woman. If what has been said is roughly on the right track, these are better theoretical tools than the ones put forward by Carnapians and give us a clearer sense of what is possible for philosophers qua philosophers to do vis-à-vis these societal battles and changes. The take-home message would be to join Carnapians in their anti-essentialist attitude, but then afford descriptively more adequate tools to make sense of present-day activists’ battles, thus fostering a better theoretical understanding of what they are about. While doing that, considerations having to do with the pragmatic implications of abiding by more or less progressive hinges, such as fostering or impeding hermeneutical injustice, can be brought to bear onto any decision.
Furthermore, the later Wittgenstein continuously asks us to imagine alternative practices, epistemic or otherwise. Consider the famous case of the odd wood sellers in The remarks on the foundations of mathematics (Wittgenstein Reference Wittgenstein1956), with their alternative epistemic practice of measuring the quantity of wood and of paying for it; or the case of the builders in the Philosophical investigations, with their partially different language-games; or the different forms of life, he asks us to imagine—let them be biologically different ones (1953, 223e), or only culturally different ones (1969, 92, 132).
These exercises in imagination have the role of exposing the contingent nature of our epistemic practices, language-games, and form of life. They aim at debunking the idea that there are essences and metaphysical necessities at every corner where we are tempted to see them. Rather, once tested through imagination, alleged essences, or metaphysical necessities, often turn out to be projections of our way of operating with language and thought. To the extent that such an essentialist conception of reality is shared by lay members of the community, becoming able to see it for what it is—that is, a projection of our categories and practices onto reality—can be a liberating message for philosophers to share with the rest of their community.
One of Wittgenstein's most well-known exercises in imagination—albeit an aborted one in his opinion—starts with the famous words “If a lion could speak …” Now think of this as a template and start producing instances of it such as “If a woman could speak (in the public sphere) …,” “If a woman could work (outside the household) …,” “If a woman could vote …,” “If a human being could affect a transition …,” “If a human being could determine their gender …” Without such imagined scenarios, no change would have been possible, no new forms of life, or, more modestly, new forms of political life—or our being together in a polis—would have been possible. Imagination is thus key in world-building—in bringing about new ways of making it possible for human beings to realize their potential and thereby live (more) freely.Footnote 30
Thus, there is something specific and liberatory about being trained in this way of doing philosophy that places considerable value in imagining alternatives to what seems to us to be normal, natural, true as a matter of objective truth, and of mind-independent, objective reality. For, on the one hand, it shows us that what appears to be of the essence of, or as belonging to the nature of, being this and that is often a product, or a function of the way in which we operate as members of a linguistic and epistemic community and as members of a society. On the other hand, it makes us aware that all these elements are largely revisable, and thereby opens spaces for new forms of political activism.
Still, by insisting on the primacy of use and practice over theory, the Wittgensteinian perspective reminds us that, to bring about societal changes, philosophers will have to engage with society directly, because it is only by “shifting the ground” from which they arise that philosophers will have the power to ameliorate our concepts and hinges.
7. Conclusions
The aim of this paper has been to show how a family resemblance approach to woman fares better than a Carnapian one on several fronts. Not only is it more apt as a methodological tool to make sense of our existing linguistic practice, but it also allows for a more inclusive use of that concept. By tying this discussion to Wittgenstein's notion of hinges, we have seen how the rules of evidential significance we bring to bear onto the determination of the extension of woman are now changing, to make room for self-identification but also for further categories beside cisgender ones. The political consequence of these new hinges is to potentially end forms of hermeneutical injustice perpetrated by more conservative, heteronormative, and biologistic hinges. Finally, by considering the later Wittgenstein's recourse to the use of imagination, we have seen how it has liberatory effects against essentialism and how it is conducive to the kind of political activism that aims at world-building. Thus, contrary to the traditional picture of Wittgenstein's thought as having no political significance and no consequence in practice, the paper has demonstrated the extraordinary political potential inherent in several key notions and methodological perspectives characteristic of the later Wittgenstein's philosophy.
Acknowledgments
Many thanks to Delia Belleri, Anna Boncompagni, Daniele Moyal-Sharrock, Katherine Ritchie, Karl Schafer, two anonymous referees, and Hypatia's editors, especially Aidan McGlynn, for their comments on previous versions of this paper, as well as people in attendance at the “Identities and Epistemic Injustice” conference (University of California, Irvine, March 12–13, 2021), the “First International Congress of Logic, Epistemology and Methodology of the University of Costa Rica” (June 21–28, 2021), the European Epistemology Network Meeting (Glasgow, June 15–17, 2022), the “XXVIII Convegno della Società italiana di filosofia del linguaggio. Regole, prassi e forme di vita. A partire dalle Ricerche filosofiche di Wittgenstein settant'anni dopo” (Università della Calabria, September 21–23, 2023); and at seminars and colloquia at the Universities of Campinas, UC Santa Barbara, Macau, and East Anglia.
Annalisa Coliva is Full Professor in the Department of Philosophy at the University of California, Irvine. She is the author of Singular thoughts: Perceptual-demonstrative and I-thoughts (with E. Sacchi, 2001), Moore and Wittgenstein: Scepticism, certainty and common sense (2010), Extended rationality: A hinge epistemology (2015), The varieties of self-knowledge (2016), Wittgenstein rehinged: The significance of On Certainty for contemporary epistemology (2022), and of two opinionated introductions: Relativism (with Maria Baghramian, 2020) and Skepticism (with Duncan Pritchard, 2022). She is currently working on a monograph on Social and applied hinge epistemology, and on a Cambridge Element on Wittgenstein and social epistemology.
