Philosophia

[Link]

Why the Trans Inclusion Problem cannot be Solved

Tomas Bogardus1

© The Author(s), under exclusive licence to Springer Nature B.V. 2022

Abstract
What is a woman? The definition of this central concept of feminism has lately
become especially controversial and politically charged. “Ameliorative Inquirists”
have rolled up their sleeves to reengineer our ordinary concept of womanhood, with
a goal of including in the definition all and only those who identify as women, both
“cis” and “trans.” This has proven to be a formidable challenge. Every proposal so
far has failed to draw the boundaries of womanhood in a way acceptable to the Ame-
liorative Inquirists, since not all those who identify as women count as women on
these proposals, and some who count as women on these proposals don’t identify as
women. This is the Trans Inclusion Problem. Is there any solution? Can there be?
Recently, Katharine Jenkins, pointing to the work of Mari Mikkola, suggests that the
Trans Inclusion Problem can be “deflated” rather than solved. We will investigate
this proposal, and show that, unfortunately, Jenkins is mistaken: Mikkola’s project
will not help us answer the Trans Inclusion Problem. After that, we’ll look at Robin
Dembroff’s suggestion that we “imitate” the linguistic practices of trans inclusive
and queer communities, and we will evaluate whether this would help us solve the
Trans Inclusion Problem. Unfortunately, this strategy also fails to solve the problem.
By the end, we’ll have a better appreciation of the challenges faced by Ameliorative
Inquirists in their project of redefining “woman,” and clearer view of why the Trans
Inclusion Problem cannot, in fact, be solved. That’s primarily because, no matter
what it means to be a woman, it’s one thing to be a woman, and another thing to
identify as a woman.

Keywords Philosophy of Gender · Gender · Feminism · Inclusion Problem · Trans

Philosophy · Trans · Woman

* Tomas Bogardus
[Link]@[Link]
1
Philosophy Department, Pepperdine University, Malibu, California, USA

13
Vol.:(0123456789)
 Philosophia

1 Introduction: What Is a Woman?

Since at least the crest of feminism’s second wave, the definition of its principal
concept—woman—has become a matter of great controversy within feminism.
Prior to that, the question seems to have been much less controversial. If one reads
the founding documents of First-Wave Feminism—for example the Declaration of
Sentiments from the Seneca Falls Convention—there’s little question as to whose
rights were at issue: those of adult female humans.1 As Talia Mae Bettcher (2009,
105) put it, “On the face of it, the definition ‘female, adult, human being’ really
does seem right. Indeed, it seems as perfect a definition as one might have ever
wanted.” But things have changed. Simone de Beauvoir’s ([1949]1956) The Sec-
ond Sex sent many in search of a purely social definition of woman, in terms of the
social processes or conditions by which one, in de Beauvoir’s terms, “becomes a
woman.” A search, in other words, for “the social meaning of sex,” as Catharine
MacKinnon (1983, 635) called it.2
However, despair set in with the rise of intersectional analyses of gender and
oppression (cf. Crenshaw, 1989), and the growing realization that there is no the
social meaning of sex, no one social role that all women share across time and
space. As Elizabeth Spelman famously argued, due to the diversity and dissimilarity
of women’s cultural and social experiences, it’s simply not the case that “underneath
or beyond the differences among women there must be some shared identity—as if
commonality were a metaphysical given” ([1988] 1990, 13). At least, not any shared
social or psychological identity.
This is a problem for those who abandoned the biological definition of women,
since the concept of womanhood is central to feminism. That is, as Linda Mar-
tín Alcoff put it, this concept is “the necessary point of departure for any femi-
nist theory and feminist politics, predicated as these are on the transformation of
women’s lived experience in contemporary culture and the reevaluation of social
theory and practice from a woman’s point of view” (Alcoff, 2006, 133). If women
are to be defined in terms of a shared social role or psychology, and yet nothing

1
Even Simone de Beauvoir, in her famous and influential book The Second Sex ([de Beauvoir,
[1949]1956, 15) says, “Woman has ovaries; a uterus…” and later (ibid., 59) she says, “…there have
always been women. They are women in virtue of their anatomy and physiology.” As late as 1976, it
seems that de Beauvoir endorsed the traditional definition of “woman,” saying in an interview, “A posi-
tive definition of ‘woman’? Woman is a human being with a certain physiology…” (Brison, 2003, 192).
For an overview of the reasons and arguments given for the abandonment of these biological definitions
of “woman” and “man,” see Bogardus, 2020b.
2
For a few examples of such projects, see Friedan, 1963 (now much derided), Rubin, 1975, MacKin-
non, 1983, 1987, Haslanger, 2012, Ásta, 2011 (published under “Ásta Sveinsdóttir”), and Antony, 2020.
Antony (ibid., 540) argues that “gender” designates a social hierarchy “between those who exist ‘for
themselves’ and those who exist ‘for others’,” this hierarchy (ibid., 543) is “erected on…difference in
reproductive role,” and that “man” and “woman” name social roles, though she denies “that there is any
specific content to these social roles.” For a recent dissent, see Stock (ms).

13
Philosophia

seems to fit the bill due to intersectionality, then there would be no women. And
if there are no women, whither feminism?3
This problem became more pressing with the emergence of trans-inclusive femi-
nism, which seeks to create a political space for trans individuals within the tent
of feminism, and especially conceptual space for trans individuals who identify as
women to count as genuine women.4 That is, lately, some feminist philosophers
have sought to revise our concept of womanhood to include trans individuals who
identify as women (and to exclude trans individuals who identify as men, and those
who identify as non-binary), in order to advance the cause of social justice. Follow-
ing Sally Haslanger, I’ll call this deliberate conceptual engineering “Ameliorative
Inquiry.” The idea is to reflect on which concept we should express with the English
word “woman,” in order to best achieve the goals of feminism, perhaps e.g. bringing
attention to unjust discrimination and subordination of women, honoring the gender
identifications of trans individuals, and the like.5 Then, we’re meant to specify the
intension of that concept—its meaning—typically via the articulation of necessary
and sufficient conditions, and, finally, to appropriate our word “woman” to express
this revisionary concept. Coming up with a suitable revisionary concept has proven
to be no easy task, and, following Katharine Jenkins (2016, 2018), I’ll call the prob-
lem of conceptually engineering our concept of womanhood so as to respect every-
one’s self-identification, “The Trans Inclusion Problem.”
In this vein, Jenkins (2016, 2018) proposes—as an explicit bit of Ameliorative
Inquiry, of conceptual engineering—that being a woman is a purely psychologi-
cal matter of having an internal ‘map’ that is formed to guide someone classed as a
woman through social and material reality. In other words, to be a woman is to take
enough of the right sort of gender norms to be for oneself, to be about oneself.6 As
Jenkins puts it (Jenkins, 2016, 411), “what is important is that one takes [norms of

3
Theodore Bach (2012, 234) calls this the ‘Representation Problem’: “if there is no real group ‘women’,
then it is incoherent to make moral claims and advance political policies on behalf of women.”
4
An individual who identifies as a woman is said to be “trans” when this individual was “assigned male
at birth.” If the assignment was accurate and if sex in humans cannot, at the moment at least, feasibly be
changed, this individual remains biologically male. Mutatis mutandis for trans individuals who identify
as men, or individuals who identify as “non-binary,” i.e. as neither men nor women (nor boys nor girls).
As for biological sex, a male is a member of a sub-type of a species whose body is organized toward—
i.e. has the function of—producing spermatozoa. (Though not all males actually produce spermatozoa.
For any number of reasons, the body may not do what it’s meant to do, what it evolved to do. It may mal-
function, just as a heart or a kidney may malfunction, or it may not yet fulfill its function, due to immatu-
rity.) Similarly with females, though with bodies organized toward the production of ova.
The definitions of “male” and “female” are not mutually exclusive, nor are they exhaustive. Some
flowering plants are both male and female simultaneously, as are common garden snails, and many fish
“change sex” during the course of a lifetime, and, so, enjoy bodies designed to produce both spermatozoa
and ova, though at different times. Other organisms don’t reproduce sexually at all, and so are neither
male nor female. It’s broadly logically possible for a species to reproduce via several sexes, though for
convenience evolution seems to have settled on no more than two.
5
For a view of the goals of feminism that contrasts with that of Haslanger, Jenkins, et al., see Holly
Lawford-Smith (forthcoming).
6
True, Jenkins proposes two distinct senses of “gender”: gender-as-class and gender-by-identity. But
she’s quite clear (Jenkins, 2016, 417) that she supports “using the term [‘woman’] to refer to people with
a female gender identity and not, in general, using it to refer to people classed as women.”

13
 Philosophia

femininity] to be relevant to oneself; whether one feels at all moved to actually com-
ply with the relevant norms is a distinct question.” Jenkins proposes this as the con-
cept we should express with the word “woman,” in order to solve the Trans Inclusion
Problem, a problem that besets social-role theories like Haslanger’s: namely that such
theories exclude a significant number of trans individuals who identify as women.7
However, Jenkins’ revisionary definition faces inclusion (and exclusion) problems
of its own, and therefore offers no trans-inclusive solution to feminism’s demarcation
problem.8 First, the proposal seems to exclude many women. For example, a woman,
raised perhaps by radical feminists, who has not internalized enough of the right sort
of gender norms to meet Jenkins’ definition; a woman who, perhaps due to cogni-
tive impairment, is unable to internalize enough of the right sort of norms to meet
Jenkins’ definition; a woman who, due to difficult relocation or immigration, has not
internalized enough of the right sort of norms to count as a woman in her new culture.
Jenkins’ definition, then, seems to marginalize (or flat-out exclude) women with non-
normative backgrounds, cognitive impairments, or difficult immigration experiences.9
Also, Jenkins’ revisionary definition will exclude many trans individuals who identify as
women. For Jenkins’ definiens features Haslanger’s social-role concept of womanhood: some-
one “classed” as a woman, someone subordinated on the basis of observed or imagined female
biological sex. To be a woman, on Jenkins’ definition, is to take enough of the norms about
women-in-Haslanger’s-sense to be relevant to oneself. But, then, trans individuals who are
excluded from Haslanger’s definition can’t sensibly take norms about women-in-Haslanger’s-
sense to be norms about them, and so they can’t sensibly be women on Jenkins’ view.10

7
Haslanger’s (2000, 39) definition is this: S is a woman if and only if she is systematically subordi-
nated along some dimension (economic, political, legal, social, etc.), and she is “marked” as a target for
this treatment by being, regularly and for the most part, observed or imagined to have bodily features
presumed to be evidence of a female’s biological role in reproduction. Jenkins (2016, 399-400) gives
examples of trans individuals identifying as women who would be excluded by this definition. For exam-
ple, trans individuals who do not publicly present as women may well not meet Haslanger’s definition:
they won’t “function as women,” in Haslanger’s sense, since they won’t be observed or imagined to have
female bodily features. Additionally, Jenkins says (ibid., 400), some trans individuals who do publicly
present as women but are nevertheless treated as men wouldn’t meet Haslanger’s definition. And, finally,
there may be trans individuals whose gender identifications are respected unconditionally, and not on the
basis of observed or imagined bodily features associated with a female’s biological role in reproduction.
Though such a person may well function as a woman, Jenkins says (ibid., 400–1), this person would not
count as a woman on Haslanger’s definition, since this person would not be treated as a woman on the
basis that Haslanger says is essential for womanhood.
8
Jenkins is concerned primarily with revising the concept of womanhood so that it is acceptably inclu-
sive, i.e. includes trans individuals who identify as women. But, of course, one also wants the concept
to be acceptably exclusive. Presumably, for example, Jenkins would like the concept of womanhood to
exclude trans individuals who identify as men, individuals who identify as non-binary, etc.
9
See also Elizabeth Barnes (2020, 710-11, and forthcoming) for similar points.
10
If they do take these norms to be relevant to them, to be about them, this will be a mistake, a result
of irrationality, ignorance, delusion, or some such shortcoming. And this would count as marginaliza-
tion, by Jenkins’ own lights. See also a recent objection from Matthew Salett Andler (2017). Andler
argues that Jenkins’ account will not be fully trans-inclusive, since “gender maps characteristic of people
socially positioned as either women or men only allow a body to be experienced as unified if its features
‘match’ along the axis of a single sex,” and some trans people experience bodily unification without such
“matching.” This is possible, Andler says, “[o]nly by adopting a radical gender map—distinct from the
gender maps characteristic of people socially positioned as either women or men…” And therefore these
trans individuals wouldn’t, on Jenkins’ view, count as women or men.

13
Philosophia

Jenkins herself (Jenkins, 2018, 733) accepts a criticism from Talia Mae Bettcher
(2017), involving the case of a trans individual S who has only just begun to present
outwardly as a woman, and so who still has a gender ‘map’ that is organized around
the norms that are applied to men in S’s social context, because these are the norms
that have been applied to S by others up to the present. Such an individual would
count as a man on Jenkins’ view, contrary to this individual’s own self-identifica-
tion. (By the same token, a trans individual who identifies as a man, or an individ-
ual who identifies as non-binary, may, due to upbringing, still have a gender ‘map’
organized around norms applied to women. This individual will count as a woman
on Jenkins’ proposal, contrary to the person’s self-identification.) This is a problem
for Jenkins, because she asserts (Jenkins, 2016, 396) as a “foundational premise,”
that “trans gender identities are entirely valid—that trans women are women and
trans men are men.” Further, she says, “[f]ailure to respect the gender identifications
of trans people is a serious harm and is conceptually linked to forms of transphobic
oppression and even violence.” So, by failing to respect the self-identification of a
significant number of those who identify as women (both cis and trans), and the self-
identification of many trans individuals who identify as men or as non-binary, Jen-
kins’ proposed ameliorative definition of “woman” falls short by her own lights.11
This brings us nearly up to date. What I’d like to focus on in this paper is the next
step in the dialectic. Jenkins admits (Jenkins, 2018, 733) that her norm-relevancy
account “does not entail that everyone is always right about their own gender iden-
tity.” (In fact, things are worse: her account entails the people can be wrong about
their own gender identity. As we’ve seen, someone may identify as a woman without
really being a woman, on the norm-relevancy account.) Jenkins takes this to be a
serious problem. But, in a footnote to that admission, Jenkins says this:
Establishing that the norm-relevancy account entailed that everyone has the
gender identity that they think they have was important for the purposes of that
paper because I was seeking to solve the ‘inclusion problem’, the challenge of
finding an appropriately inclusive concept of ‘woman’ for use in feminist prac-
tice. Since that time, Mari Mikkola has argued, in my view convincingly, that
the inclusion problem rests on a confusion and can therefore be deflated rather
than solved.
This is an intriguing prospect, “deflating” rather than solving the inclusion prob-
lem. Perhaps, as Jenkins thinks, recent work by Mari Mikkola (Jenkins cites
Mikkola, 2016, chapter 5) provides a satisfying answer to what Jenkins calls the
inclusion problem—but which we’ll call the Trans Inclusion Problem—and even
to feminism’s definition problem more generally. Also, more recently, Dembroff
(2018) has proposed an “Imitation Approach” for revising our gender concepts in a

11
Jenkins (2018, 739) concedes this is a cost, saying, “[c]ertainly, an account of gender identity that
entailed that everyone has the gender identity that they think they have whilst also meeting all of the
desiderata… would be preferable to the norm-relevancy account.” But she’s pessimistic that this is pos-
sible, and so in that paper she continues to advocate for her norm-relevancy account as “good.”

13
 Philosophia

trans-inclusive way, which one might think offers the prospect of solving the Trans
Inclusion Problem.
So, in what follows, I will consider Mikkola’s deflationary strategy, along with
objections, and then we’ll turn to Dembroff’s Imitation Approach, along with objec-
tions. We will see, in the end, that neither proposal offers a satisfying answer to
the Trans Inclusion Problem, or to feminism’s definition problem more generally.
Mikkola’s strategy was simply not intended to solve the Trans Inclusion Problem,
so Jenkins seems to be putting it to a task beyond its design. And, as a solution to
the Trans Inclusion Problem, Dembroff’s “Imitation Approach” suffers from formi-
dable difficulties. Indeed, I will argue that any attempt to solve the Trans Inclusion
Problem on behalf of Ameliorative Inquirists faces insurmountable problems. It is
impossible, in principle.

2 Mari Mikkola’s Deflationary Strategy

Mari Mikkola (2016) sets out to solve two puzzles. The Semantic Puzzle (ibid.,
3–4) is this: “Given that ordinary language users tend not to distinguish sex and
gender (treating ‘woman’ largely as a sex term, or a mixture of social and bio-
logical features), what precisely are feminists talking about when they talk about
women?” And the Ontological Puzzle is this: “How should we understand the cat-
egory of women that is meant to undergird feminist political solidarity, if there are
no necessary and sufficient conceptual conditions underlying our gender talk? Do
women make up a genuine kind, or simply a gerrymandered and random collec-
tion of individuals?”
Mikkola’s main contention is that, as she puts it (Mikkola, 2016, 105), “the
semantic and ontological issues are not as pressing as feminists make them out to
be. We can talk about women without providing a thick articulation of woman, and
we can understand the phenomena usually discussed under the rubric of the sex/
gender distinction in a way that avoids certain unintuitive and undesirable ontologi-
cal implications. My aim is to show that we can deflate the ontological and semantic
worries, and so there is no need to seek solutions to them.” Mikkola understands
the semantic and ontological issues as challenges to define “woman” in a satisfac-
tory way, on the assumption that, to do so, one must articulate necessary and suf-
ficient conditions. Instead of answering the challenge head-on, by supplying such
conditions, Mikkola says she will “deflate” the worries, by defining “woman” while
rejecting the assumption that, to do so, one must articulate necessary and sufficient
conditions. The project of feminism can proceed, Mikkola thinks, even without
articulating a definition of “woman.”
How is this “deflation” to be accomplished? In a word, ostension. The idea is
that we can sidestep concerns about articulating satisfactory necessary and sufficient
conditions for our concept woman by, instead, pointing at the relevant category, or,
more precisely, by deferring to the “reference-fixing extensional intuitions” of ordi-
nary language users. Mikkola says (ibid., 106), “…we need not precisely specify
what it takes for someone to count as a woman (or satisfy the intension of woman)

13
Philosophia

in order to make good our reference fixing. Thus, my proposal is that in order to
retain gender talk for politically relevant social explanations, we can merely rely on
the reference-fixing extensional intuitions.”
What exactly are these intuitions? She explains (ibid., 109), “our (perceptualist) exten-
sional intuitions about ‘woman’ are simply about which individuals we think the term
applies to. They do not, then, inform us about the content of any concept, manifest or
operative. …Extensional intuitions understood as predoxastic perceptualist intuitions give
feminists what they need: a way to retain gender talk in a deflated manner.” Finally, she
concludes (ibid., 110), “Quite simply: focusing on ordinary language users’ willingness to
apply ‘woman’ is enough to pick out women’s type, and this is sufficient to answer the
representation problem (how to fix feminism’s subject matter). In order to avoid getting
bogged down by conceptual problems, this is what feminist philosophers should settle for.”
Here’s how I understand the dialectic at this point. The project of Ameliorative
Inquiry—of intentionally revising the concept of womanhood in order to advance the
cause of feminism—is meant to proceed in two stages: a Preparation Stage, and then a
Revisionary Stage. In the Preparation Stage, we identify what Sally Haslanger (2012,
375) calls “the dominant manifest meaning” of the term “woman,” or what Toril Moi
(1999, 8) calls “the ordinary understanding.” This is the subject of our engineering
project, the concept which is to be repaired or enhanced. And then, in the Revision-
ary Stage, the goal is to modify this concept via conceptual engineering, and thereby
arrive at an acceptable “target” concept of womanhood, the definition of which draws
just the right boundaries. Crucially, though Mikkola was not engaged in Ameliorative
Inquiry in the passages I’ve quoted, it seems to me that her answer to the Semantic
Puzzle and the Ontological Puzzle can help the Ameliorative Inquirist only with the
Preparation Stage. Mikkola argues we can identify the ordinary, manifest meaning of
“woman” without articulating necessary and sufficient conditions. We instead adopt
an attitude of deference to ordinary users of “woman,” and resolve to use the term in
whatever way they’re using it. We thereby join the conversation, and “fix feminism’s
subject matter,” as she puts it. Crucially, for the purposes of conceptual engineering,
by using the word “woman” with this intention to use it as the folk do, we thereby
come to use that very concept. We get the concept on the operating table, as it were, by
following ordinary users’ extensional intuitions. And we do all this, to repeat, without
articulating necessary and sufficient conditions for the application of “woman.”
As for the second stage of Ameliorative Inquiry, the Revisionary Stage, Mikkola
makes no attempt, but seems to think it would proceed in the way commonly envi-
sioned by Ameliorative Inquirists. She says (ibid., 113), “in appealing to extension
[sic] intuitions, which I do take to be highly uniform, I am not proposing to cement
the kinds’ boundaries. Furthermore, disagreements generated by extensional intu-
itions that involve trans* people and intersexes are not primarily to be settled by
semantics— they are a political issue.”12 And, later, she says (ibid., 116), “Despite

12
Citing a Tumblr blog (http:// youkn​owyou​retra​ns.​tumblr.​com), Mikkola (ibid., 23) explains her use of
“trans*”: “In many activist circles, the denotation ‘trans*’ has recently been used to refer to ‘transgender,
transsexual, genderqueer, non- binary, genderfluid, genderfuck, intersex, third gender, transvestite, cross-
dresser, bi-gender, trans man, trans women, agender’… ‘Trans*’ is also considered to be more inclusive
than ‘trans’, which is taken to refer to medically or hormonally altered transsexual men and women.”

13
 Philosophia

a high degree of uniformity, extensional intuitions do not deliver rigid kind bounda-
ries since there are times when ‘our’ intuitions come apart. But, as noted, in these
cases the debate is not to be had on the level of semantics, but on the level of poli-
tics.” In these quotations, Mikkola seems to be alluding to the Revisionary Stage
of Ameliorative Inquiry: after “fixing feminism’s subject matter” in the Preparation
Stage, one may well go on to make conceptual changes in the Revisionary Stage.
She doesn’t mean to “cement” the boundaries of womanhood; those kind boundaries
are not “rigid,” and may be changed during Ameliorative Inquiry in light of political
(normative) considerations. In these respects, she seems to be on exactly the same
page as those who have actually attempted the Revisionary Stage, like Haslanger
and Jenkins.
So, this is the deflationary proposal from Mikkola that Jenkins points to as an
answer to the Trans Inclusion Problem. However, I think there are serious concerns
about the prospects of conscripting Mikkola’s project as a satisfying answer to the
Trans Inclusion Problem. Indeed, Mikkola did not intend it to do so. Jenkins is
mistaken to prescribe Mikkola’s deflationary strategy for uses beyond its officially
approved indications.

3 Objection: The Preparation Stage isn’t where Ameliorative

Inquiry’s Problems Lie

Recall that Jenkins believes Mikkola has argued convincingly that the Trans Inclu-
sion Problem rests on a confusion, and can therefore be deflated rather than solved.
But the problem is this: Mikkola’s project might be repurposed to help, at most, with
the Preparation Stage of Ameliorative Inquiry, but the Preparation Stage of Ame-
liorative Inquiry is not where Jenkins’ problems lie. It’s fairly commonly granted
that the manifest/ordinary concept of womanhood is analyzable as adult female
human. Though she herself goes on to reject this definition, Jennifer Saul (2012,
196) admits, “According to most ordinary speakers and dictionaries, ‘woman’ is
a sex term—a term that picks out those who have certain biological traits.” Mik-
kola herself agrees. Of the extensional intuitions she points to in order to specify the
subject matter of feminism, Mikkola says (2016, 110), “Ordinary language users…
apply ‘woman’ with a high degree of uniformity. They are seldom confused about its
extension…” What is this extension? Mikkola says (ibid., 21), “Speakers ordinarily
seem to think that ‘gender’ and ‘sex’ are coextensive: women and men are human
females and males, respectively, and the former is just the politically correct way to
talk about the latter.” And, elsewhere, speaking of Haslanger’s proposed target con-
cept (which apparently excludes the Queen, as she’s not subordinated on sex-marked
grounds), Mikkola says (ibid., 82), “[Haslanger’s] proposal is strikingly counterin-
tuitive. It seems that most people would be willing to call the queen a ‘woman’,
and their willingness to do so might suggest that Haslanger is simply wrong about
the content of gender concepts.” So, it’s not difficult to “fix feminism’s subject mat-
ter,” as Mikkola puts it; it’s not hard to locate the manifest/ordinary concept of

13
Philosophia

womanhood. That concept is straightforwardly analyzable as adult female human.

That’s the concept that Ameliorative Inquirists mean to place on the operating table
during the Preparation Stage of Ameliorative Inquiry, in order to surgically modify
it during the Revisionary Stage.
The Revisionary Stage of Ameliorative Inquiry is the tricky part; that’s where the
difficulties are. Recall, the Revisionary Stage is where the conceptual engineering is
meant to happen. In this stage, we’re meant to modify, enhance, or perhaps simply
replace the ordinary concept we identified in the Preparation Stage, and come out
the other side with our new and improved target concept to express with the word
“woman,” a concept which draws satisfactory boundaries. There are three challenges
associated with this stage of Ameliorative Inquiry. First and foremost, we must pro-
vide an acceptably inclusive target concept. This is challenging if, from the concep-
tual engineer’s perspective, it’s difficult to see what all and only women have in com-
mon. If your target extension of “woman” includes individuals from all across each
dimension of biological sex, social role, ‘lived experience’, performativity, sexual ori-
entation, internalized gender ‘maps’, etc., then it will be very difficult to specify a tar-
get concept that draws the boundaries in a satisfying way.13 Also, we cannot success-
fully draw these boundaries if we attempt to specify a target concept in a viciously
circular way, as seems to be the case with “sincere self-identification” approaches,
e.g. Bettcher’s (2017, 396) proposal that being a woman is entirely a matter of sin-
cerely identifying as a woman.14 (No concept is successfully specified by a viciously
circular proposal.) As detailed above, both Haslanger’s and Jenkins’ proposed target
concepts founder on this first challenge, since they exclude some trans individuals
who identify as women (and, for similar reasons, plausibly include some trans indi-
viduals who identify as men, and some individuals who identify as non-binary).
Turning now to Mikkola’s deflationary strategy, what’s important to note for
present purposes is that Mikkola makes no attempt to provide such a defini-
tion. No part of her deflationary strategy speaks to the Revisionary Stage of
conceptual engineering, producing a target concept with the right boundaries.
Her deflationary strategy of pointing to the extensional intuitions of ordinary
speakers yields only the traditional, biologically-based definition of “woman.”

13
Or, if you prefer to think in terms of intensions—functions from contexts of utterance to referents—it
will be very difficult to specify what this function is, i.e. how it works, what characteristics determine
the application of the term. Note: not what sort of heuristics or clues we happen to use, psychologically,
to apply (or perhaps misapply) the concept in thought. But, rather, what characteristics determine the
proper application of the concept, i.e. what characteristics are necessary and sufficient for genuine kind
membership.
14
Though see Bettcher, 2009, 109ff) recognition and response to the problem of circularity, in terms of
“existential self-identification.” The idea seems to be that “I am a woman” may be intended to express a
desire to be treated in a certain way (ibid., 108), or an answer to questions like “What am I about? What
moves me? What do I stand for? What do I care about the most?” (ibid., 110). On this account, in either
case, the sentence “She is a woman” may be intended to express a proposition that is true of a trans indi-
vidual who identifies as a woman, but in neither case does that sentence express the proposition that she
is a woman.

13
 Philosophia

One might use Mikkola’s insights as an acceptable approach to the Prepara-

tion Stage of Ameliorative Inquiry, but as a satisfactorily trans-inclusive target
concept of the Revisionary Stage, all parties can agree it is surely lacking. So
Mikkola’s deflationary strategy cannot help Jenkins avoid the Trans Inclusion
Problem. It offers no answer at all to this challenge of Ameliorative Inquiry. It
was not meant to.
Worse, there seems to be an insurmountable general problem here for Ame-
liorative Inquiry of gender terms. It is widely agreed by Ameliorative Inquirists
that, for an output of the Revisionary Stage to be a satisfactory target concept, it
must align with the criterion of self-identification. As Jenkins (2016, 396) puts
it, “an important desideratum of a feminist analysis of gender concepts is that
it respect these identifications by including trans people within the gender cat-
egories with which they identify and not including them within any categories
with which they do not identify.”15 And Mikkola (2016, 113) says, “[T]heory of
gender that point-blank excludes trans* women from women’s social kind is sim-
ply unacceptable.”16 But the self-identification criterion itself cannot be used to
specify a new target concept, since that would be viciously circular.17 So, for the
Ameliorative project of trans-inclusive feminism to be successful, the engineered
target concept must somehow respect self-identification, without using self-iden-
tification itself to draw the boundaries.
The problem is, it sure looks like any other criteria will not be necessarily co-exten-
sive with people’s self-identifications: no matter which criteria we choose, whether
someone meets those criteria seems to have no necessary connection with her identi-
fying as a woman, in the sense of “woman” cashed out by those criteria. In that case,
there will always be a failure to respect self-identification, and so the Ameliorative
project of trans-inclusive feminism is bound to fail. This would explain why propos-
als such as Haslanger’s and Jenkins’ have both failed (by their own lights) to correctly
draw the boundaries: because they used, as criteria, features that do not necessarily co-
vary with self-identification, and this is verboten for the trans-inclusive Ameliorative

15
Later (Jenkins, 2018, 723), Jenkins gives this as her second desideratum for an account of gender
identity: “D2: The definition should be compatible with a norm of FPA.” Of this desideratum, Jenkins
says (ibid.), “This demand can be understood as a call for what Bettcher (2009) has described as a norm
of ‘First Person Authority’ (FPA), a principle stating that a person should be treated as the final and deci-
sive authority on their own gender identity.” Jenkins proposes her norm-relevancy account as an account
of gender identity, the same concept which she proposed we should express with the word “woman,” in
her earlier work (Jenkins, 2016, 417). So I take it that, in her 2018 paper, Jenkins remains committed to
the importance of including trans people within the gender categories with which they identify and not
including them within any categories with which they do not identify.
16
In passing, let me mention a second general problem in the neighborhood, suggested to me by Alex
Byrne (though any infelicities in its expression are my own). If at least some trans individuals are self-
identifying using Preparation-Stage concepts, then no novel ameliorative concepts emerging from the
Revisionary Stage will succeed in respecting this self-identification. And insofar as Ameliorative Inquiry
necessarily originates novel concepts (see below), there looks to be an insurmountable problem for Ame-
liorative Inquiry here—at least if it really is the case that at least some trans individuals self-identify
using our ordinary Preparation-Stage, pre-ameliorated concepts.
17
Or, if not viciously circular, then trans-exclusive. For details, see Bogardus (2020a).

13
Philosophia

Inquirist.18 In any case, nothing about Mikkola’s deflationary strategy will help us sur-
mount this difficulty. And so, again, contrary to Jenkins’ claim that Mikkola answers
the Trans Inclusion Problem—by ‘deflating’ it rather than solving it—it seems that
Mikkola’s project offers no answer at all to the Trans Inclusion Problem, because that
problem arises during the Revisionary Stage of Ameliorative Inquiry, and Mikkola
was concerned only to pin down the ordinary meaning of “woman,” which could help
at most with the Preparation Stage of Ameliorative Inquiry.19
The second challenge facing the Revisionary Stage of Ameliorative Inquiry is
this: don’t change the subject. That is, those engaged in Ameliorative Inquiry intend
to continue a long-standing conversation about gender, men, women, etc. They don’t
mean to simply borrow the words “gender,” “man,” and “woman,” and use them
to express novel concepts, and to speak, rather misleadingly, of new topics. This is
challenging because it’s quite plausible that, like words, concepts are individuated by
their origins, and originating uses of concepts—that is, mental acts that instantiate
new concepts, rather than existing concepts—are marked by an intentional lack of
deference to established practice. As Mark Sainsbury and Michael Tye put it (Sains-
bury & Tye, 2012, 42), deference occurs when “we aim to conform in our usage to
our previous usage, and to the usage of those in our conceptual community… This is
typical of non-originating uses. The deference takes the form of intending to use the
concept as it has been used by oneself or others on previous occasions… One char-
acteristic of originating uses is the absence of any such conformist requirement.”20
Since it is a self-aware project of conceptual engineering, Ameliorative Inquiry
evidently lacks any intention to conform to the usage of those in our conceptual
community. That is, those engaged in Ameliorative Inquiry lack, as Sainsbury and

18
As another example, consider Bach’s, 2012 proposal, on which gender is a natural kind with a historical
essence. Bach is quite explicit that this proposal won’t respect everyone’s self-identification, saying (Bach,
2012, 269), “Because one’s status as a man or woman is conditional upon historical relata, individuals
cannot simply stipulate their gender status as ‘woman’ or ‘man’. Indeed, individuals do not have privileged
access to historical processes, and it follows that they can be mistaken about their own gender status…”
Though he goes on to say that it’s less likely that they are mistaken about their gender identity, their inter-
nal sense of what gender they are, Bach’s view seems to be that one’s gender identity can be mistaken.
The view also entails that there are women in the United States if and only if there are no women in Japan.
Though he uses both “American women” and “Japanese women,” the sad fact is that, on Bach’s view
(ibid., 262), these groups “do not descend from the same ancestral population of women and are therefore
not members of the same historical gender kind.” Though the groups are members of the same teleofunc-
tional kind, this will not help rescue the view, since, for Bach, “woman” is not (merely) a teleofunctional
kind. It’s a historical kind, it has a historical essence, and these groups do not have the right histories to
be members of the same historical kind. So, again, on Bach’s view, there are women in the United States if
and only if there are no women in Japan. This is an implausible implication, to say the least.
19
To forestall possible misunderstanding here, allow me to reiterate that I am not saying that Mari
Mikkola was engaged in Ameliorative Inquiry in her 2016 work. She wasn’t. Rather, she was aiming to
“deflate” what she calls the Semantic Puzzle and the Ontological Puzzle, roughly, the challenge of defin-
ing “woman” by articulating necessary and sufficient conditions. She “deflates” the challenge by instead
defining the ordinary use of “woman” via ostension. This move could be used by one so inclined during
the Preparation Stage of Ameliorative Inquiry. But it will be of no use at all during the Revisionary Stage.
20
As to why this view is plausible, confer Sainsbury and Tye (2012) to see how it offers elegant solu-
tions to seven traditional puzzles of thought and language, including the puzzle of Hesperus and Phos-
phorus, the puzzle of Paderewski cases, and the puzzle of empty thoughts.

13
 Philosophia

Tye say (ibid., 70), “the recognition that others already use a concept, together with
the desire to use the very concept they use, with the very reference it has in their
uses.” Ameliorative Inquirists certainly recognize how others already use the rele-
vant concept—in the case of our concept of womanhood, we’ve seen the widespread
concession that the manifest/ordinary concept refers to the property of being an
adult female human—but they have no desire to use the concept with the very refer-
ence it has in common uses. So, the worry is that the Revisionary Stage of Amelio-
rative Inquiry necessarily originates a new concept, appropriates an existing word to
express this new concept, and therefore equivocates, changing the subject.21
Mikkola herself seems to agree (Mikkola, 2016, 110): “The prospects of an adequate
revisionary analysis of woman also look unpromising since this tends to make gender
terminology unfamiliar to ordinary speakers. Common gender terms become feminists’
theoretical terms, which … is politically unhelpful.” And the problem is that Mikkola’s
deflationary strategy offers no solution to this problem of changing the subject. (As I
said before, it was not meant to solve this problem.) While deferring to ordinary users’
extensional intuitions may help fix the subject of feminism during the Preparation Stage
of Ameliorative Inquiry, such deference ceases in the Revisionary Stage, when we self-
consciously attempt to alter, modify, or engineer that concept. And, so, it continues to
look as though the output of Ameliorative Inquiry is a new concept, not a modified old
concept. In which case, we’ve changed the subject, and introduced ambiguity.22

21
This is a well-known concern about conceptual engineering in general. See, for example, Part III of
Cappelen, 2018.
22
Herman Cappelen (2018) takes seriously the concern that conceptual engineering necessarily changes
the subject, and proposes that there are cases in which two subjects can be truly said to be saying the
same thing, even though they use a context-sensitive expression that has a different extension in each
of the contexts (and therefore, he says, different intensions as well). His example (ibid., 110) is light on
details, but the idea focuses on the context-sensitive term “smart,” and seems to be that subject A might
truly utter, in one context, “Serena is really smart,” and subject B might utter, in a different context, “Ser-
ena is really smart,” and we can truly say, in some contexts, “A and B said the same thing about Serena.”
And then Cappelen concludes, “These descriptions of them as samesayers are true even if it turns out
that A’s and B’s respective contexts fix somewhat different extensions and intensions for ‘smart’.” So
two people can say the same thing—and therefore speak to the same topic—even if the relevant term
differs in extension and intension, Cappelen reasons. By way of reply, we might first point out that, even
if A’s and B’s uses of “smart” have different extensions in these contexts, this is not enough to prove
that they have different intensions, since intensions are functions from contexts of utterance to exten-
sions, and A and B are in different contexts. And the claim that A and B are samesayers is plausible only
when we assume sameness of intension. But if they are in fact using the words (to express concepts)
with different intensions, it’s not at all clear that they really are saying the same thing, pace Cappelen.
Secondly, as Mark Schroeder (2020) put it, “In general, appeals to semantic context-dependence are illu-
minating when they appeal to a common core meaning.” Think of the relevant common core meanings
revealed by Kaplan’s treatment of the “character” of indexicals, Kratzer’s treatment of modal verbs like
“must,” DeRose’s theory of “knows,” etc. In a similar way, any illuminating treatment of context-sensi-
tive expressions like “smart” will reveal a commonality between A’s utterance and B’s utterance. Cap-
pelen prefers not to speak of concepts (he’s non-committal on their existence), but this seems like a grand
opportunity to appeal to them: A and B are expressing the same concept with “smart.” If they do succeed
in saying the same thing, sameness of concept will help explain why. But, in that case, we won’t have an
example here that could help us see how Ameliorative Inquirists might originate a new concept (with a
new intension and extension), and yet manage to stay on the same topic, because, in Cappelen’s example,
A and B are using the same concept. Neither A nor B originates a new concept, while nevertheless stay-
ing on topic.

13
Philosophia

The third and final challenge for the Revisionary Stage of Ameliorative Inquiry
is that of avoiding a circular methodology. During the Revisionary Stage, we’re
meant to engineer a target concept with an eye toward advancing the cause of
feminism. The problem is that, in order to understand the cause of feminism, it
seems that we must use some concept of womanhood, since feminism is the cause
of women. But which concept of womanhood shall we use when evaluating how
well these proposed, ameliorative target concepts advance the cause of women?
We can’t use the novel target concept itself at this stage of Ameliorative Inquiry,
because we’re still not sure whether it’s a good concept to use; it’s the concept
we’re evaluating, not yet employing. But if we use the ordinary, biological con-
cept of womanhood in this evaluation process, the challenge is to explain how
the revisionary target concept advances the cause of adult female humans. To my
knowledge, nobody engaged in Ameliorative Inquiry has tried to spell out exactly
how this would go. For it seems somewhat obvious that, if the goal is to advance
the cause of women, understood as adult female humans, then probably the best
way forward is to reserve the word “woman” to express the concept of an adult
female human. If things are not how they seem here, one should like to know
why.
Mikkola demurs, saying (Mikkola, 2016, 116), “if one takes feminism to be
organized around women, and if extensional intuitions provide a way to pick out this
kind, not being able to articulate a solution to the semantic puzzle by explicating the
applicability conditions for woman does not present a serious political concern.” On
the contrary, I think it does, at least if one intends to engage in Ameliorative Inquiry,
revising the concept in light of political or normative concerns, e.g. respecting
self-identification. For, if Mikkola is right about extensional intuitions, one would
have to explain how modifying the definition of “woman” for political reasons ben-
efits the cause of women, understood in the sense expressed by ordinary speakers’
extensional intuitions, which is, according to Mikkola, adult female humans. Mik-
kola makes no attempt to do this. So here, too, her deflationary strategy provides no
answer to the challenges besetting the Revisionary Stage of Ameliorative Inquiry.
Mikkola did not intend it to.
After I presented an ancestor of this paper last year (February 2021) at a sym-
posium together with Katharine Jenkins, Jenkins updated the paper she presented
there with a substantial section on this issue of deflation. In this recently published
work, Jenkins (2022) seems to indicate that this point from Mikkola is what she
(Jenkins) had in mind in her 2018 footnote, when speaking of Mikkola’s deflation-
ary strategy. In that more recent work, without providing any direct quotations from
Mikkola, Jenkins says (Jenkins, 2022, 250), “Mikkola argues that rather than start-
ing from the idea that feminist action must be justified by some sort of significant
commonality between women (that we urgently need to locate), we should think
of feminist action as justified by the recognition that people are wronged by sex-
ism.” For reasons given above and below, I don’t think this is an accurate character-
ization of Mikkola’s project. But notice Jenkins’ talk of justification here. Jenkins
here seems to change her conception of the Trans Inclusion Problem from what she

13
 Philosophia

said in 2016 and 2018.23 In this more recent work, Jenkins (2022, 238) character-
izes the problem as the thought that “our account of gender kinds should capture
a commonality between women that justifies the formation of a social movement.”
Notice, again, Jenkins’ shift to a characterization of the inclusion problem as one of
justifying feminist action, instead of, as she put it in 2018, “the challenge of finding
an appropriately inclusive concept of ‘woman’ for use in feminist practice.”
Jenkins (2022, 250) then argues—contra no one, as far as I can tell—that
we need not rely on a definition of “woman” to justify a feminist social move-
ment. That’s true enough. “For that,” she continues, “pointing out the many
wrongs of sexist societies—wrongs that affect people who are not women as
well as many who are women—will do just fine.” Well, there I’m not so sure. It
seems to me that would depend on what sort of feminist movement one would
like to justify. Who is included in this movement, and who is excluded? Identi-
fying the goals of feminism—for example, the goal of eliminating sexism—will
require some understanding of feminism and of sexism. Both of those concepts
bear a priori entailment relations to our concepts of women, of men, and others.
Depending on what concepts we use—for example, if we use the traditional,
biological concepts of womanhood and manhood—the sort of feminism justi-
fied by the wrongs of sexism may not be as inclusive as philosophers like Jen-
kins would prefer, or at least explicitly did prefer in her 2016 and 2018 papers.
Jenkins remains sensitive to this concern, and she is still engaged in Ameliorative
Inquiry in her 2022 paper; indeed, she offers a new, pluralist account of gender.24

23
In the abstract of her 2016 paper, Jenkins characterized the problem this way: “Feminist analyses of
gender concepts must avoid the inclusion problem, the fault of marginalizing or excluding some prima
facie women.” Later (ibid., 396), she argues against Haslanger’s proposal for target gender concepts on
the grounds that it “does not in fact solve the inclusion problem because it does not include trans peo-
ple within their identified genders.” In footnote 24 her 2018 paper, Jenkins characterizes the inclusion
problem as “the challenge of finding an appropriately inclusive concept of ‘woman’ for use in feminist
practice.” Notice that there’s no talk here of justifying feminist action, there’s talk only to the effect that
“an important desideratum of a feminist analysis of gender concepts is that it respect these identifications
by including trans people within the gender categories with which they identify and not including them
within any categories with which they do not identify” (Jenkins, 2016, 396).
24
Jenkins’ new pluralist account is, in general form, this: “to be a member of a gender kind is to be constrained
and/or enabled in the ways characteristic of that kind” (Jenkins, 2022, 247). These constraints and enablements
can, according to Jenkins, vary by context, by type, and by granularity. So, she says (ibid., 247): “A given gender
kind is to be understood as an explanatory kind where what the members have in common is that they fall under
a particular bundle of constraints and enablements that has a certain scope, breadth, and granularity.” Which par-
ticular bundle, you ask? The bundle characteristic of that kind. So, for example, to be a woman is to be con-
strained or enabled in ways characteristic… of women. Taken at face value, then, this pluralist account of gender
looks to be viciously circular: the concept to be defined shows up in the definition. But, for reasons given above,
even if “is a woman” as Jenkins proposes to use it expresses some genuine feature or set of features, it will not
necessarily co-vary with self-identification as a woman, and therefore will not “avoid the inclusion problem, the
fault of marginalizing or excluding some prima facie women,” as she put it in the abstract of her 2016 paper.
Because, again, it will be one thing for a person to constrained or enabled in a particular way, and another thing
for a person to identify as being constrained or enabled in that particular way. Whatever exactly Jenkins’ new
pluralist proposal amounts to, then, it will suffer from the same fault that Haslanger’s proposal does, according to
Jenkins (2016, 396), namely that “it does not include trans people within their identified genders.” And this is a
serious problem according to Jenkins (ibid.), because “[f]ailure to respect the gender identifications of trans peo-
ple is a serious harm and is conceptually linked to forms of transphobic oppression and even violence.”

13
Philosophia

So, which concept of womanhood does she propose to use when evaluating whether
new, revisionary definitions of “woman,” like her new pluralist account? That is,
we’d like to check whether Jenkins’ new pluralist account of gender helps advance
the cause of feminism, or helps eliminate sexism. But, to think about feminism or
sexism, we must exercise some concept of womanhood. What will that concept be?
Our ordinary, biological concept of womanhood, i.e. the one exercised in ordinary
language users’ extensional intuitions, as Mikkola says? If so, again, it’s hard to see
how a revisionary concept of womanhood will better advance the cause of femi-
nism than the traditional, biological concept. But we can’t use Jenkins’ novel plural-
ist concept itself at this stage of Ameliorative Inquiry, because we’re still not sure
whether it’s a good concept to use; it’s the concept we’re evaluating, not yet employ-
ing. That methodology would be circular, and impossible to complete.
In conclusion, then, Mikkola’s proposal offers no answer to any of the problems
of the Revisionary Stage of Ameliorative Inquiry. But those are all the problems
distinctive of Ameliorative Inquiry. So, although Jenkins points toward Mikkola’s
deflationary strategy as the next step in the dialectic related to the Trans Inclusion
Problem, Mikkola’s proposal does not advance that dialectic one step. Jenkins’s
hope that Mikkola’s project could be applied to the problems besetting Ameliorative
Inquiry is, I believe, misguided.
The reason is not far to seek. Mikkola is quite clear that she was addressing
what she calls (2016, 3) “gender skeptic” views, which say both that, “woman is
ultimately meaningless and necessarily open-ended, which renders any attempt to
define it futile and even politically counterproductive,” and also, secondly, that,
“the social kind of women does not exist in any meaningful sense: either its exist-
ence is illusory (Butler, 1999) or the kind is so hopelessly fragmented that any
talk of women in plural is meaningless (Spelman, [1988]1990 [original 1988]).”
Mikkola aims to prove that (ibid., 6), “this dispute is theoretically bankrupt and
intractable,” but that, “we can provide a theory of injustice from a feminist per-
spective without having to elucidate a thick social conception of woman.” This
project could help the Ameliorative Inquirist with, at most, the Preparation Stage
of Ameliorative Inquiry: identifying the ordinary, biological concept of woman-
hood exercised by ordinary speakers, and corresponding concepts of feminism and
of sexism. But this biological concept of womanhood that Mikkola points to won’t
help solve or deflate the Trans Inclusion Problem. It’s the source of the Trans
Inclusion Problem. And that problem could be solved only via successful comple-
tion of the Revisionary Stage of Ameliorative Inquiry. But Mikkola offers no guid-
ance on that stage.
So, it is no surprise that Mikkola’s work will not avail us during the Revi-
sionary Stage of Ameliorative Inquiry; it wasn’t intended to. But since the Revi-
sionary Stage is where Jenkins’ problems arise, Jenkins is mistaken to think that
Mikkola’s deflationary strategy will provide an answer to the Trans Inclusion
Problem. So, Jenkins and other Ameliorative Inquirists must look elsewhere for
a satisfying answer to the Trans Inclusion Problem, and to feminism’s definition
problem more generally.

13
 Philosophia

4 Robin Dembroff’s “Imitation Approach”

Mikkola’s appeal to ordinary language users’ extensional intuitions may help us set-
tle the subject matter of feminism prior to engaging in conceptual engineering, but
it cannot guide us during that engineering process toward an appropriately inclu-
sive target concept. However, Dembroff (2018) has recently proposed an alternative
strategy. In this section, I would like to explain that strategy. In the next, I will share
some concerns about the strategy.
Dembroff’s idea is that we might improve dominant gender kinds via imitation of
trans inclusive and queer communities. Perhaps—one might think—we could thereby
solve the Trans Inclusion Problem. In brief, the suggestion is that the conceptual engi-
neering process has already been accomplished by these communities, in a way that
draws the boundaries of gender terms in acceptable ways, and so all that’s left for us
to do is start using the concepts developed and in use by these communities. As Dem-
broff, 2018, 37) puts it, “why reinvent the wheel when it is spinning and fabulous?”
Why try to manufacture what we can easily import? Dembroff acknowledges concerns
that such imitation might destroy trans and non-binary gender kinds, but Dembroff
(ibid.) says that at least trying to import such kinds is preferable to the status quo.
Here’s how Dembroff (ibid., 36) describes the approach:
On this different approach–the ‘imitation’ approach–classification practices can be
revised based upon those that already exist within other communities. Given that
operative gender kinds vary across contexts, looking at other communities will reveal
alternative gender kinds and corresponding classification practices. From here, those
in one context can attempt to revise their operative gender kinds by mirroring or oth-
erwise imitating the structures and practices that already exist in other contexts. To
improve dominant gender kinds, trans inclusive and queer communities are obvious
places to begin. The operative gender kinds in these contexts are markedly different
than those in dominant contexts. For one, within these communities, nonbinary gen-
der identities have intelligible social meaning because of systems that accommodate
these identities… Similarly, within these communities, membership criteria for men
and women differ from the corresponding criteria in dominant contexts. Genitalia—
much less natal genitalia—does not determine gender classification… Gender clas-
sification practices in such communities defer to self-identification, and do not take
anatomical information or gender presentation to determine gender.25

25
In a much earlier work, Bettcher (2009, 108) also directs our attention to how gender terms are used
in “subaltern contexts.” In these contexts, Bettecher says, “gender presentation is not understood to com-
municate genital status but, instead, indicates how persons want to be treated. Individuals’ self-identi-
fications are generally accepted at face value.” This much sounds in accord with Dembroff’s proposal.
Bettcher, however, seems less sanguine about importing any of these concepts “as is,” since their use
even in these subaltern contexts are not “politically ideal.” In particular, Bettcher says, “even in these
subaltern contexts, some interactions continue to reflect mainstream transphobic attitudes. I mention two
examples. First, despite the fact that gender presentation isn’t taken to communicate genital status, the
importance of genital status has not been abolished. Sometimes transpeople inquire about somebody
else’s status inappropriately and freely circulate information without consideration. Second, it is a sad,
peculiar fact that some MTFs (who have no difficulty referring to each other with appropriate pronouns)
when first learning to interact with FTMs can’t or won’t transfer the practice.”

13
Philosophia

So, to repeat, the idea is this: these trans inclusive and queer communities have
already performed the labor of conceptual engineering, and the target concepts
they’ve produced are appropriately inclusive (and exclusive). They draw the
boundaries in satisfactory ways, in ways that “defer to self-identification.” So,
Dembroff recommends that we at least try to improve dominant gender kinds by
“imitating”—or, more accurately, by incorporating, taking up, employing—the
concepts and use-patterns that we observe in these other communities.26 One
can understand it in this way: instead of using ostension to specify a concept
prior to Ameliorative Inquiry, as Mari Mikkola proposes, Dembroff suggests we
instead use ostension to specify concepts that have already completed the pro-
cess of Ameliorative Inquiry. And, just as we defer to medical experts regard-
ing what arthritis is, we might defer to trans-inclusive and queer communities
regarding what women are, and thereby use these concepts that these communi-
ties have developed. The question now is, could this proposal offer a satisfying
solution to the Trans Inclusion Problem, and to the problems facing Ameliora-
tive Inquiry in general?

5 Problems for Dembroff

I believe that Dembroff’s Imitation Approach cannot help us surmount any of the
three challenges of the Revisionary Stage of Ameliorative Inquiry. Recall, those
three challenges are: (i) provide an acceptably inclusive target concept, (ii) don’t
change the subject, and (iii) avoid circular methodology. Let’s look at each of these
three challenges in turn, to see whether Dembroff’s Imitation Approach could help
us meet the challenges.
Our first Revisionary Stage challenge is: Provide an acceptably inclusive target
concept. With regard to womanhood, what exactly is the acceptably inclusive target
concept that we are meant to imitate and take up from trans inclusive and queer
communities? What concept, exactly, do these communities express with the word
“woman,” or the predicate “is a woman”? A genuine predicate expresses a func-
tion from objects to TRUE or FALSE (in Frege’s terms), or IN or OUT with regard
to an extension. In these trans inclusive and queer communities, what function is
expressed by “is a woman”? How does it work?
Dembroff gestures toward a Bettcher-style sincere-self-identification definition,
saying that, in these communities, “[g]ender classification practices in such com-
munities defer to self-identification.”27 The problem is that, as mentioned above,

26
Dembroff (ibid., 37) responds to the question “whether trans and nonbinary gender kinds could be
incorporated within dominant structures… without being destroyed.” Dembroff (ibid., 38) concludes
that, “In the end, we may find that there is no true middle ground between destructive oppression and
destructive assimilation. But I think working to find one is preferable to the status quo…”
27
See also Dembroff and Wodak (2018, 374), where the authors say, “[Catherine] McGregor, for
instance, is a transgender woman because she identifies as a woman, but she was assigned the sex male
at birth… For one to be transgender (as opposed to cisgender) is for one’s gender identity to (sometimes)
differ from the sex they were assigned at birth.” Here the authors seem to endorse a Bettcher-style self-
identification definition.

13
 Philosophia

attempting to define a concept in terms of self-identification in this way is viciously

circular, and therefore unintelligible. To see this, suppose that, since I’m not sure
what concept “woman” is being used to express in these communities, I ask Dem-
broff for clarification. If Dembroff answers that, on this concept of womanhood,
someone is a woman if and only if she identifies as a woman, in the very sense
to be defined, we’ve gone in a circle. I still would not know what concept is being
expressed, since the term to be defined has shown up in the definiens. Insofar as this
was an attempt to convey the meaning of this new target concept, the attempt fails.28
Of course, we should be careful to avoid the Socratic fallacy: just because we
cannot articulate necessary and sufficient conditions for some term does not mean
the term fails to express any coherent concept. Just because the attempt to convey
the definition of this concept failed (due to circularity), maybe there is a definition
nonetheless. Perhaps there are necessary and sufficient conditions for the application
of “woman” in these trans inclusive and queer communities, even if these conditions
are difficult to articulate, difficult to communicate to someone who doesn’t already
grasp the concept. Just as Mikkola proposes using ostension to “fix feminism’s sub-
ject matter”—in the “settle” sense of “fix”—perhaps Dembroff’s idea here is to use
ostension in order to fix feminism’s subject matter, in the “repair” sense of “fix.”
Mikkola points to ordinary use of “woman” in the general population to sidestep
giving necessary and sufficient conditions for the ordinary concept, while Dembroff
points to alternative use in trans inclusive and queer communities to sidestep giving
necessary and sufficient conditions for the revised, engineered, ameliorated concept.
There remains a general worry. Let’s suppose that the predicate “is a woman”
expresses some function in these communities, and “woman” expresses a satisfac-
tory target concept in these communities. That is, suppose the meaning of the predi-
cate successfully picks out some feature (or set of features) W, that is necessary and
sufficient for the concept to apply.29 Let’s grant that it would be difficult to articulate
this feature (or set of features) W, as this would require a careful combination or
disjunction of, perhaps, age, species, biological sex, sexual orientation, social role,
mental state, etc. Perhaps we’re too conceptually impoverished to do so; fair enough.
But the suggestion is there is some such feature (or set of features) W, and that’s
what it is to be a woman, according to these communities, and W will necessarily
co-vary with identifying as a woman.

28
If “woman” in the definiens expresses the very concept that was being defined, and I do not yet grasp
that concept, then the definition is of no help to me in understanding or grasping the concept. If, on the
other hand, “woman” in the definiens expresses some other concept with which I’m already familiar—
perhaps the ordinary, adult female human concept, or Haslanger’s social role concept—then, while I can
understand the definition, it will not be trans inclusive (or appropriately exclusive), for reasons given
above and in Bogardus 2020a.
29
One sometimes hears allusions to “cluster concepts” at this point in the dialectic. (See, for example,
Hale, 1996 and McKitrick, 2015.) But, if “woman” names a category, there are necessary and sufficient
conditions for membership in this category, even if those conditions are something like sufficient resem-
blance to certain paradigm cases. So I don’t think that an affinity for cluster concepts will absolve one of
the commitment to necessary and sufficient conditions for category membership. There is also a serious
worry about circularity, when selecting the paradigm cases: S is a woman if and only if S sufficiently
resembles paradigm cases... of women.

13
Philosophia

To recapitulate the general worry expressed earlier, there are reasons to be

pessimistic about this idea, since it seems fairly obvious that any woman—or,
indeed, any person at all—could identify as having feature W, or not. Because
of the contingency of human mental states and our powers of attention and con-
ceptualization, there seems to be no necessary connection between having fea-
ture W—which we’ve supposed is necessary and sufficient for the application of
whatever concept is expressed by “woman” in these trans inclusive and queer
communities—and the criterion of self-identification. This is no stainless realm
that we inhabit; we occasionally make mistakes. One might have feature W and
fail to identify as having feature W; or one might identify as having feature W
and yet not really have feature W. This is why any coherent concept of woman-
hood, used in any community whatsoever, will fail to necessarily co-vary with
self-identification.
But, of course, it may be that, really, I merely fail to appreciate the impossibil-
ity of womanhood and self-identification as a woman coming apart, and I mistake
that for seeing the possibility of these two things coming apart. I suggest, though,
that the onus is on those who champion this allegedly satisfactory target concept
to explain what it means, and how it necessarily co-varies with self-identification,
rather than on the agnostic.
Here’s another way to put the worry. Suppose someone new enters one of these
trans inclusive or queer communities. There should be something that determines
whether or not this person is a woman. What would this be? The answer is the
meaning of the predicate “is a woman” in these communities, and the answer must
be a set of necessary and sufficient conditions, even if these conditions are ineffable
due to our conceptual poverty.
Suppose we’re given this answer: our newcomer is a woman if and only
if she identifies as a woman. The problem is that we seem to be getting two
answers to our question, which is one too many. This person would have to
identify as a woman, in order to be a woman. What would she be identify-
ing as? A woman, evidently. That’s one sense of “woman” that this answer
presupposes: the sense that gives the content of the intentional state of iden-
tifying as. Yet we’re also told that to be a woman just is to have that inten-
tional state: identifying in that way is what makes her a woman. This is a sec-
ond, distinct sense of “woman” that the answer proposes. These two senses
appear to be distinct, not necessarily equivalent, and this is the source of my
confusion.
Consider the definition in the form of this (implicitly necessitated) biconditional:
S is a woman if and only if S identifies as a woman.
If I’m told that the occurrence of “woman” in each bijunct of the proposed bicon-
ditional expresses the same concept, the same sense, then the biconditional looks
necessarily false. For how could it be, for any feature at all, that to have this feature
is to identify as having it? It seems that, in the case at hand at least, each bijunct

13
 Philosophia

could be true while the other is false.30 I conclude, then, that Dembroff’s Imitation
Approach won’t help us solve the first challenge of the Revisionary Stage of Ame-
liorative Inquiry, which is the provision of a coherent target concept.31 Indeed, it
looks as though it must fail to solve this challenge; that challenge looks to be broadly
logically insurmountable. And, if so, the Trans Inclusion Problem cannot be solved.
Let’s move on, then, to the second challenge of the Revisionary Stage of Amelio-
rative Inquiry. That challenge, recall, was: Don’t change the subject. I believe that,
if put to the use of the Ameliorative Inquirist, Dembroff’s Imitation Approach
will face a trilemma: either these trans inclusive and queer communities are origi-
nating a new concept (and, with a new referent) and therefore changing the sub-
ject, or they’re not originating a new concept, and therefore, either misapplying
the mainstream concept, or changing the subject via organic reference shift.
It’s more difficult, in this case, to determine which lemma will hold for these communities than
it is in the case of philosophers we’ve discussed. Typically, philosophers working in this area
are self-consciously engaged in the revisionary project of Ameliorative Inquiry. In their use of
“woman,” they knowingly and intentionally depart from the mainstream, traditional, biolog-
ically-based use. So, it looks rather obviously like they are originating new, ‘target’ concepts
rather than (mis)applying the original, mainstream concept. In the case of trans inclusive and
queer communities, though, it’s less clear that the use of “woman” is intentionally revisionary.
These communities chose to appropriate existing terms like “man” and “woman” for a reason,
and they often appear to disagree with ordinary users of those words about the extensions of
those words, etc. Yet insofar as their use is revisionary, it looks like a new concept has origi-
nated, the subject has changed, and we’ve introduced ambiguity and merely verbal disputes.
We’ve already explored above the implications of the first possibility, on which the pri-
mary intention of users in these communities is to depart from the common use. On this
possibility, there is origination of a new concept (one with a new referent), and therefore
subject change, equivocation, merely verbal disputes, and the like. So, let’s consider the
second possibility: members of these communities do not mean to depart from the main-
stream use of “woman”; their primary intention is to stay on topic, and engage with main-
stream and traditional conversations about women. If so, then—and bear with me now—
there are two further possibilities. On both possibilities, the deference shown by these
communities to common usage ensures that they are using the same concept as are main-
stream communities. But does this concept retain the same extension, the same referent?
One possibility is that it does. If so, it looks as though the use of “woman” in these trans
inclusive and queer communities involves a degree of deference, an intention to conform

30
There might be recherché features like that of identifying in some way, such that if one identifies as hav-
ing that feature, then necessarily one has that feature. But surely the other direction of the (implicitly neces-
sitated) biconditional will be false: it will be possible to have this feature without identifying as having this
feature. Also, notice that the concern is not that, as David Lewis (1986, 26) put it, the predicate picks out a prop-
erty, but the property makes for an “unduly miscellaneous class of things.” Rather, it’s that the alleged predicate
fails to pick out any property at all, due to vicious circularity. It specifies no extension. It’s not a genuine predi-
cate. In that case, the proposal has the unwelcome implication that there are no women. Or, if “is a woman”
does specify a predicate, the biconditional “S is a woman iff S identifies as a woman” will be necessarily false.
31
I’m grateful to Alex Byrne for a series of questions he raised to me that pressed me to think more
carefully about this point.

13
Philosophia

usage to mainstream usage. This is typical of uses of terms that do not originate a new con-
cept (cf. Sainsbury & Tye, 2012, 42ff). But, if there has been no shift in reference, then use of
“woman” by these communities would be like Arthritis Man’s (Burge, 1979) use of “arthri-
tis.” Arthritis Man defers to his doctor, accepts correction, and thereby manifests his the desire
to use the very concept his doctor uses, with the very reference it has in his doctor’s uses
(cf. Sainsbury and Tye, ibid. 70). It is on this basis that we judge Arthritis Man to be using
our common concept of arthritis, instead of a novel, idiosyncratic concept, tharthritis.32 But,
if this is right, then although these trans inclusive and queer communities have not changed
the subject, they stay on topic at the cost of saying many false things about women. Just as
Arthritis Man says, falsely, that he has arthritis in his thigh (instead of saying, truly, that he has
tharthritis in his thigh, using a novel, idiosyncratic concept), members of these communities
are frequently saying, falsely, that so-and-so is a woman, or that so-and-so is not a woman.33

32
Though, notice, that accepting correction from an expert is just one way of exhibiting an intention to
conform in one’s usage. Arthritis Man may instead have replied to his doctor in this way, “Yes, I know what
you think arthritis is, and that it can’t be in my thigh, but your corrupt modern medicine misunderstands
arthritis. Actually, arthritis can be had in the thigh.” I think that we’d still judge, in this case, that Arthritis
Man uses the common concept rather than an idiosyncratic concept. It’s his intention to engage and cor-
rect with the doctor that inclines us in this direction, intentions that one also observes among many users
of revisionary gender concepts. But, if that’s the case, then it looks like the Imitation Approach reaches a
bad result: just as Arthritis Man is misapplying the common concept of arthritis, and saying things that are
false—e.g. “there’s arthritis in my thigh”—these communities are misapplying the common concepts of
manhood and womanhood, and thereby saying things that are false, e.g. “trans women are women.”
33
I take it this was perhaps the crucial point of contention between Alex Byrne (2020 and 2021) and
Robin Dembroff (2021) in their recent exchange. Byrne (2021) thinks that many individuals—philoso-
phers and otherwise—in trans activist and queer communities often have “simple disagreements” about
women, with mainstream and gender-critical folks: “there is a single proposition expressed using the word
‘woman’, which one party believes and the other does not.” This would require that all parties to the debate
share a single concept (or, at the very least, concepts with the same referent). Dembroff (2021) disagrees,
and points to “the language use of feminist philosophers, as well as entire linguistic communities where
‘woman’ is explicitly not used to mean ‘adult human female’.” It seems that Dembroff’s interpretation of
“women” in such uses is motivated primarily by a principle of charity, or truth-maximization. That is, Dem-
broff seems to assume that because statements like “trans women are women” express true propositions in
these contexts, “woman” must not be expressing the ordinary concept. But, as we’ve seen, the cost of this
view is a change in subject, a shifting of concepts resulting in equivocation: those who say “trans women
are women,” using this new concept, are not contradicting those who say “trans women are not women,”
using the ordinary concept. If we instead want to maximize relevance, and make sense of what seem to be
genuine disagreements about women among different communities, we are pushed rather in the direction of
sameness-of-concept. But, the cost there is that, if there has been no reference shift, many members of the
communities Dembroff invites us to imitate are frequently saying false things about womanhood, e.g. “trans
women are women.” And, if there has been a reference shift, then there has been a change in subject, and
there is really no substantive disagreement after all, only ambiguity and equivocation.
Although, the claim that e.g. “Trans women are women,” or “I am a woman” (said by a trans individual
who identifies as a woman) would be false when using the ordinary, biological concept of womanhood
depends on a face-value interpretation of that sentence. But perhaps Bettcher (2009, 108) is right in saying
that, in these “subaltern contexts,” gender presentation generally, and self-descriptions in particular, are not
meant to indicate genital status, but instead to indicate “how persons want to be treated.” In that case, if
“I am a woman” does not express the proposition that I am an adult human female, but instead a proposi-
tion more like I wish to (or should) be treated like a woman (should be treated), then such utterances may
well express true propositions, propositions featuring the ordinary, biological sense of “woman.” Bettcher’s
suggestion here may better explain the loci of contemporary debates over gender identity, access to shared
spaces, etc., and, so, in my opinion it deserves more attention and development than it has so far received.

13
 Philosophia

A second possibility is that, due to their deference to use in mainstream communi-

ties, users of “woman” in trans inclusive and queer communities express the ordinary
concept. Yet, on this second possibility, there has been an organic reference shift, as
discussed by Saul Kripke (1980, 93) with respect to the name “Santa Claus,” and by
Gareth Evans (1973, 195-196) with respect to the name “Madagascar.” Prior to the
year 1800 at least—and even today in other cultures—cognates of the name “Santa
Claus” express a concept that refers to St. Nicholas, the historical bishop of Myra.
But, in contemporary mainstream North American cultures at least, that same con-
cept, expressed by “Santa Claus,” has shifted its reference, to a creature of fiction. If
something like this happened with terms like “woman,” as used by trans inclusive and
queer communities, then there is no risk of widespread misapplication of the concept
within these communities. But the specter of a change in subject has returned. Plausi-
bly, those who use “Santa Claus” to talk about St. Nicholas and those who use “Santa
Claus” to talk about a jolly Nordic elf are talking about different things, different sub-
jects. If one says, “Santa Claus is not alive” and the other says “Santa Claus is alive,”
there is no deep disagreement here, since they’re talking about different things. The
same would be true, mutatis mutandis, with terms like “woman” in these trans inclu-
sive and queer communities. In which case, this proposal entails a change in subject.
It seems, then, that no horn of this trilemma is an attractive one for the Ame-
liorative Inquirist who wishes to employ Dembroff’s Imitation Approach: the use
of “woman” in these communities originates a new concept, and therefore changes
the subject, or the use of “woman” in these communities is non-originating, and,
therefore, either these users are saying many false things concerning womanhood,
or they’re saying many true things, but about a new, changed subject. So, I conclude
that the Imitation Approach does not offer a satisfactory answer to this second chal-
lenge of the Revisionary Stage of Ameliorative Inquiry.
Finally, let’s consider the third challenge of the Revisionary Stage of Ameliorative
Inquiry: Avoid circular methodology. Recall that, in the Revisionary Stage, we are
meant to evaluate proposed target concepts that “woman” might come to express in
light of advancing the cause of feminism. And the concern was that we cannot evaluate
how well a proposed meaning of “woman” will advance the cause of feminism, unless
we already know who the women are. Because feminism just is the cause of women.34
So it seems that the process can’t get going, due to these recursive instructions.35
What does the Imitation Approach have to say about this? While introducing the
approach, Dembroff (ibid., 36) says, “looking at other communities will reveal alter-
native gender kinds and corresponding classification practices.” But the problem is:
how do we know these kinds and classification practices that Dembroff points to in
these other communities are gender kinds and gender classification practices? What

34
Even if you construe Ameliorative Inquiry as intending to advance the cause of social justice gener-
ally, and not feminism in particular, you will still run into this worry of circularity, assuming that your
conception of social justice includes women. And it does, doesn’t it?
35
Or, if we’re meant to evaluate proposed revisionary meanings of “woman” in light of how well they
will advance the cause of women in the traditional sense of “woman,” i.e. adult female humans, then it’s
rather doubtful that we will select a trans-inclusive target concept to be expressed by “woman.” For how
would letting “woman” express any concept other than adult female human advance the cause of adult
female humans better than the status quo? See Bogardus 2020a.

13
Philosophia

is gender? No doubt these communities use the words “man” and “woman,” but,
surely, merely using those words is not enough to ensure that their referents are gen-
der kinds, or that one has expressed gender concepts. As a quick and silly example to
help see why, suppose I decide to use the words “man” and “woman” in an unusual
way, classifying everything in the east half of my house “the men” and everything
in the west half “the women,” deferring not at all to normal uses of these terms, and
thereby originating unusual new concepts to be expressed by these words. Obviously,
I have not thereby created alternative gender kinds and classification practices. I have
not expressed gender concepts at all with this use of “man” and “woman.”
But a similar point holds with regard to the trans inclusive and queer communities
that Dembroff suggests we imitate. If the ordinary definitions of “man” and “woman”
are correct, and these are gender terms, then gender is a matter of only age, species, and
biological sex: men are adult male humans, and women are adult female humans.36 But,
in that case, there are no “alternative gender kinds” at all, and a fortiori there are none
in the communities that Dembroff points to, for there remain in those communities only
adult male humans and adult female humans.37 In other words, by claiming that there
are alternative gender kinds in these communities, Dembroff is already presupposing a
revisionary definition of gender, revisionary definitions of “man” and “woman.”
But that’s what the Revisionary of Ameliorative Inquiry is meant to give us as an output.
Here, Dembroff is using it as an input. At the start of the Revisionary Stage of Ameliorative
Inquiry, we are not yet in a position to say that these communities have alternative gender
kinds. We couldn’t sensibly do that until the Revisionary Stage is completed, and we have
a revisionary concept of gender that would deem the alternative kinds in these communi-
ties gender kinds. But, if we were to follow Dembroff’s Imitation Approach, completing the
Revisionary Stage would require that we already know that these communities feature alter-
native gender kinds: we would be meant to look to these communities, recognize that these
alternative kinds are indeed gender kinds, and then take up the corresponding concepts. But,
in that case, we would have to have already completed the Revisionary Stage of Ameliora-
tive Inquiry before we started it. And, thus, we have not escaped the worry about circular
methodology that plagues the Revisionary Stage of Ameliorative Inquiry. The proposed

36
We’re setting aside for the moment “girl” and “boy,” if those are gender terms. But what goes for
“man” and “woman” here also goes for those terms, if these are all gender terms. Also, if you believe
that at least some intersex individuals are neither men nor women (nor girls nor boys) and yet have gen-
ders, then I suppose you’ll think there are additional gender categories comprised of intersex individuals.
However these categories are defined—perhaps in biological terms (to reflect the biological characteris-
tics of their members)—so long as they also apply in these trans inclusive and queer communities, then
this argument against the Imitation Approach will go through. Also, even if you use “gender” in a more
traditional way, to mean something like “the social meaning of sex,” then gender is still defined in terms
of sex. And, since there are no additional sexes in these communities, there are no additional genders in
these communities. There will remain only norms about males (i.e. masculinity), norms about females
(i.e. femininity), etc., just as there are outside of these subaltern contexts, even if the norms vary by con-
text.
37
And, as mentioned in the previous footnote, you may wish to add additional categories for intersex
individuals. Still, these categories also apply in these trans inclusive and queer communities. These com-
munities may feature classificatory practices that differ from mainstream classificatory practices, but if
the mainstream use of gender terms is the correct use, then these communities will be systematically
misclassifying many people.

13
 Philosophia

Imitation Approach would be impossible to complete, due to circularity: we must complete

the Preparation Stage before completing the Revisionary Stage, but we also must complete
the Revisionary Stage before completing the Preparation Stage. This is impossible.
In other words, Amelorative Inquirists are trying to engineer a revisionary ‘tar-
get concept’ to be expressed by “woman,” which best advances the cause of feminism
(or social justice generally, which includes the cause of women). And, if they try to do
so by following Dembroff’s Imitation Approach, they face another dilemma. The first
horn is: If we use the traditional concept of womanhood to evaluate whether adopting
the “woman”-using practice of these other communities, we’ll need to see a reason why
doing so would advance the cause of adult human females. (But what could that reason
possibly be?) The second horn is: If we instead bracket off any particular understanding
of womanhood, how can we discern whether these communities are using a concept that
should interest us? They may well be using some other, irrelevant concept, expressed
by the word “woman.” But such a concept may be of no use or interest to women. For
we’re interested in concepts expressed by “woman” only if they mean the right thing, i.e.
only if they’re concepts of women. But if we genuinely set aside our concept of women
to evaluate the language-using practices of these communities, we’ll be unable to tell if
their practices are of any use or relevance to the cause of women.
So, Dembroff’s Imitation Approach does not provide satisfactory answers to any of the
three challenges of the Revisionary Stage of Ameliorative Inquiry. But that was the goal of
the Imitation Approach: to provide improved, revisionary target concepts for our gender terms
(in addition, perhaps, to additional gender kinds). So, like Mikkola’s proposal before it, Dem-
broff’s proposal would offer no way forward for the Ameliorative Inquirist. In that case, Ame-
liorative Inquiry’s Trans Inclusion Problem remains undeflated, unanswered, and unsolved.
We’ve also seen reason to believe that the Trans Inclusion Problem cannot be solved,
at least so long as one criterion of success for any conceptually engineered, appropriately
inclusive concept of womanhood is that it necessarily covaries with self-identification as
a woman. And that’s because, no matter what it means to be a woman, it’s one thing to
be a woman, and another thing to identify as a woman. This follows from a more general
principle: it’s one thing to be some way, and another thing to identify as being that way.
So the project of crafting a concept of womanhood that necessarily respects everyone’s
self-identification is doomed to fail. The Trans Inclusion Problem cannot be solved.
Though the project of this paper was largely negative, I hope that, by clearing away
proposals that cannot succeed, we have a better view of the challenges facing Ameliorative
Inquiry, and the options available. It looks as though the only way forward for the Amelio-
rative Inquirist involves abandoning the idea that gender self-identification is infallible.38

38
In a recent paper, Elizabeth Barnes (forthcoming) targets the claim that self-identification as a woman
is necessary for being a woman, since (ibid., 16) “cognitively disabled females are women,” even if they
are unable to identify as such. But Barnes endorses the sufficiency of self-identification for gender clas-
sification, saying (ibid., 3), “we need to say that self-identifying as an x is at least sufficient for being
categorized as an x.” Though it may look like this variable is bound by an implicit universal quantifier
ranging over an unrestricted domain, I believe the most charitable reading requires that we restrict the
domain to gender kinds, since it’s quite implausible that, in general, for any x, self-identifying as an x is
sufficient for being categorized as an x, and certainly not for actually being an x. One main purpose of
this present paper is to point out that self-identification as a woman is neither necessary nor sufficient for
being a woman, no matter what “woman” means.

13
Philosophia

Acknowledgements This paper benefited greatly from conversations at a workshop back in early 2020.
I’d especially like to thank Sophie Allen, Alex Byrne, Holly Lawford-Smith, Mary Leng, Jon Pike, and
Kathleen Stock. The paper was also presented as part of a symposium at the Central Division meeting of
the American Philosophical Association in February 2021. I’m grateful to feedback from the audience
there, and from the other symposium participants, Sophie-Grace Chappell and Katharine Jenkins.

References
Alcoff, L. (2006). Visible Identities. Oxford University Press.
Andler, M. S. (2017). Gender identity and exclusion: A reply to Jenkins. Ethics, 127(4), 883–895.
Antony, L. (2020). Feminism without metaphysics or a deflationary account of gender. Erkenntnis, 85,
529–549.
Ásta. (2011)(published under "Ásta Sveinsdóttir"). The metaphysics of sex and gender. In C. Witt (Ed.),
Feminist metaphysics: Explorations in the ontology of sex, gender and the self (pp. 47–65). Springer.
Bach, T. (2012). Gender is a natural kind with a historical essence. Ethics, 122, 231–272.
Barnes, E. (2020). Gender and Gender Terms. Noûs, 54(3), 704–730.
Barnes, E. (forthcoming). Gender without gender identity: The case of cognitive disability. Mind.
de Beauvoir, S. [1949] (1956). The Second Sex, translated by H.M. Parshley (New York: Vintage Books)
Bettcher, T. M. (2009). Trans identities and first-person authority. In L. Shrage (Ed.), You’ve changed:
Sex reassignment and personal identity (pp. 98–120). Oxford University Press.
Bettcher, T. M. (2017). Through the looking glass. In A. Garry, S. Khader, & A. Stone (Eds.), In Rout-
ledge companion to feminist philosophy. Routledge.
Bogardus, T. (2020a). Some internal problems with revisionary gender concepts. Philosophia, 48(1),
55–75.
Bogardus, T. (2020b). Evaluating arguments for the sex/gender distinction. Philosophia, 48(3), 873–892.
Brison, S. J. (2003). Beauvoir and feminism: Interview and reflections. In C. Card (Ed.), The Cambridge
companion to Simone de Beauvoir. Cambridge University Press.
Burge, T. (1979). Individualism and the mental. In P. French, T. Uehling, & H. Wettstein (Eds.), Midwest
studies in philosophy (4th ed., pp. 73–121). University of Minnesota Press.
Butler, J. (1999). Gender Trouble (2nd ed.). Routledge.
Byrne, A. (2020). Are Women Adult Human Females? Philosophical Studies, 177(12), 3783–3803.
Byrne, A. (2021). Gender muddle: A reply to Dembroff. Journal of Controversial Ideas, 1(1), 1–24.
Cappelen, H. (2018). Fixing language: An essay on conceptual engineering. Oxford University Press.
Crenshaw, K. (1989). Demarginalizing the intersection of race and sex: A black, feminist critique of anti-
discrimination doctrine, feminist theory and antiracist. Politics University of Chicago Legal Forum,
1989, 139–167.
Dembroff, R. (2018). "Real Talk on the Metaphysics of Gender," Philosophical Topics ,46(2), 21–50
Dembroff, R. (2021). Escaping the natural attitude about gender. Philosophical Studies, 178, 983–1003.
Dembroff, R., & Wodak, D. (2018). He/she/they/Ze. Ergo, 5(14), 371–406.
Evans, G. (1973). The causal theory of names. Proceedings of the Aristotelian Society, Supplementary,
47, 187–208.
Friedan, B. (1963). The feminine mystique. W. W. Norton.
Hale, J. (1996). Are lesbians woman? Hypatia, 11(2), 94–121.
Haslanger, S. (2000). Gender and race: (what) are they? (what) do we want them to be? Noûs, 34(1),
31–55.
Haslanger, S. (2012). Resisting reality. Oxford University Press.
Jenkins, K. (2016). Amelioration and inclusion: Gender identity and the concept of woman. Ethics, 126,
394–421.
Jenkins, K. (2018). Toward an account of gender identity. Ergo, 5(27), 713–744.
Jenkins, K. (2022). How to be a pluralist about gender categories. In R. Halwani, J. M. Held, N. McK-
eever, & A. Soble (Eds.), The philosophy of sex: Contemporary readings (8th ed., pp. 233–259).
Rowman & Littlefield.
Kripke, S. (1980). Naming and necessity. Harvard University Press.
Lawford-Smith, H. (forthcoming). ’Gender: What is it, and what do they want it to be? In In sex matters:
Essays in gender-critical philosophy. Oxford University Press.

13
 Philosophia

Lewis, D. (1986). Against structural universals. Australasian Journal of Philosophy, 64(1), 25–46.
MacKinnon, C. (1983). Feminism, Marxism, method, and the state: Toward feminist jurisprudence.
Signs, 8(4), 635–658.
MacKinnon, C. (1987). Feminism unmodified. Harvard University Press.
McKitrick, J. (2015). A dispositional account of gender. Philosophical Studies, 172(10), 2575–2589.
Mikkola, M. (2016). The wrong of injustice: The role of dehumanization in feminist philosophy. Oxford
University Press.
Rubin, G. (1975). The traffic in women: Notes on the ‘political economy’ of sex. In R. Reiter (Ed.), In
toward an anthropology of women. Monthly Review Press.
Sainsbury, R. M., & Tye, M. (2012). Seven puzzles of thought and how to solve them: An originalist
theory of concepts. Oxford University Press.
Saul, J. (2012). Politically significant terms and philosophy of language: Methodological issues. In A.
Superson & S. Crasnow (Eds.), Analytic feminist contributions to traditional philosophy. Oxford
University Press.
Schroeder, M. (2020). Review of semantics for reasons, eds. Weaver, B. R. and Scharp, K. Notre Dame
Philosophical Reviews, Accessed January 10th, 2021 [Link]
ns/.
Spelman, E. ([1988]1990). Inessential woman: Problems of exclusion in feminist thought. Beacon Press.
Stock, K. (ms) “Not the Social Kind.”
Toril, M. (1999). What is a Woman? (Oxford University Press)

Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published
maps and institutional affiliations.

13