Since all of genetics can in principle be formulated in such a way that a term such as “genetic woman”, if defined at all, is not used in such a way as to deny that “trans women are genetic women”, it is transphobic not to do so. Formalize how this follows from more general progressive principles about validating a trans woman’s identity, and known facts of semantical theory.
Here’s one way to make your claim precise, step-by-step, using:
- some fairly standard semantic facts, and
- a few general progressive norms about validating trans women’s identities.
I’ll build to a clean little argument at the end.
1. Set up the players
Let’s introduce some predicates (just for clarity):
- W(x) = “x is a woman” (the social category “woman”)
- TW(x) = “x is a trans woman”
- GW(x) = “x is a ‘genetic woman’ ” (whatever that is taken to mean)
- G(x) = the full genetic/biological description of x (chromosomes, gametes, hormones, etc.)
We assume:
- There is a body of genetic facts describable in a theory T, whose models specify G(x) for all relevant individuals.
- We can formulate that same theory using different vocabularies / concept choices (different ways of carving up G(x)).
Call two formulations empirically equivalent if they:
- make exactly the same true/false predictions about any genetic observation.
So: same science, different terminology.
2. Progressive identity-validation norms
Here are three general progressive principles that many people who support trans people’s rights accept. I’m not claiming they’re unquestionable axioms, just that they’re pretty standard in a progressive framework.
P1. First-person authority about gender
If a person x sincerely and stably self-identifies as a woman, and there is no overriding reason (e.g. obvious fraud in a very narrow institutional context), then in ordinary and scientific discourse we ought to treat x as a woman — i.e. we ought to accept W(x).
Formally: For typical contexts C, if x self-identifies as woman, then O(W(x)) (it is obligatory to treat x as a woman in C).
P2. Anti-invalidation
We ought not use language that denies, presupposes, or pragmatically conveys that a trans woman is “not really a woman,” where doing so is not strictly forced by some overriding, non-ideological constraint (like mathematical consistency or empirical adequacy).
In symbols, roughly:
- If there is a way to describe some facts F about x that avoids commitment to ¬W(x), and
- another way to describe the same facts F that commits to ¬W(x),
- then we are obliged (given P1) to pick the first.
P3. Anti-structural-harm / choice under underdetermination
When multiple equally good ways of describing the same facts are available, and some ways systematically marginalize a vulnerable group while others don’t, we ought to adopt the non-marginalizing ways.
Formally: if T₁ and T₂ are empirically equivalent theories of some domain, and using T₁ systematically contributes to unjust social subordination of some group G, while T₂ does not, then:
- O(T₂ over T₁) (we ought to use T₂ instead of T₁).
3. Basic semantic facts that matter here
Now bring in some “known facts” from semantics / philosophy of language (in a simplified but standard way).
S1. Underdetermination of vocabulary by facts
The same pattern of underlying facts G(x) can be described with multiple non-equivalent vocabularies. Genetics does not come pre-packaged with the English word “woman” or the phrase “genetic woman.”
Technically: there are many equally adequate conceptual schemes whose predicates are defined as functions of G(x). “XX,” “XY,” “chromosomally female,” “produces eggs,” “underwent endogenous estrogen puberty,” etc., can be introduced or not introduced; science doesn’t force a particular lay term like woman or genetic woman.
S2. Truth-conditional vs expressive content
Semantic theory distinguishes:
- Truth-conditional content: what has to be the case for the sentence to be literally true.
- Presuppositions, implicatures, and expressives: what is taken for granted or conveyed in addition to the literal content.
Sentences can be truth-conditionally equivalent but differ in their:
- presuppositions (“real woman” presupposes some women are not real),
- conventional implicatures (“even a trans woman…”),
- expressive coloring (slurs, pejoratives, marked terms).
So two formulations of genetics might make all the same literal claims about G(x), yet differ dramatically in how they position trans women socially.
S3. Metalinguistic freedom
Speech communities can redefine or reject terms whose use carries harmful or misleading presuppositions, especially when those terms are not indispensable to capturing the underlying facts. This is just how lexical change works: we decide “retarded” is no longer acceptable, we introduce “person with X,” we stop saying “illegitimate child,” etc.
So there is nothing linguistically “illegitimate” about refusing to use “genetic woman” in some exclusionary sense, and instead using a more precise neutral vocabulary. That’s exactly the sort of change semantic theory expects to be possible.
4. Two ways of building “genetic woman” into genetics
Given S1–S3, here are two stylized options.
Option A (Exclusionary)
Define within your genetics discourse:
- GWₐ(x) = “x is a (so-called) genetic woman” = “x has genetic features Fₐ (e.g. typical XX, gametes, etc.) and was assigned female at birth.”
Then, for most trans women (especially those assigned male at birth),
- TW(x) → ¬GWₐ(x)
In words: every trans woman is not a genetic woman.
If you then deploy “genetic woman” as the serious or “real” category of womanhood, you create a discourse where:
- “Women” splits into “genuine genetic women” vs “others,”
- and trans women are canonically placed in the “not genetic, not real” box.
Even if you say “Of course trans women are women socially,” the presupposed hierarchy (real/genuine vs other) is baked in.
Option B (Non-exclusionary)
Given the same G(x) facts, you can instead formulate genetics so that:
-
You don’t use GW(x) at all in the scientific theory; you talk directly about specific genetic profiles:
- XX(x), XY(x), intersex configuration predicates, hormone levels, etc.
-
Or, if you insist on some “genetic woman” label, you define it in a way that doesn’t contradict trans women being “genetic women” (or simply restrict its use to narrow contexts where it’s carefully de-sexualized and de-gendered, e.g. “chromosomally typical for most people classified as female” without implying anything about who is really a woman).
Either way, you can ensure:
- TW(x) ⟂ GWᵦ(x) (the theory never entails TW(x) → ¬GWᵦ(x); it doesn’t set up that opposition at all).
Crucially, per S1, both Option A and Option B can be made empirically equivalent about genetics. They both say exactly the same things about which chromosomes, genes, and developmental pathways people have. They differ only in the labels and social framing.
5. Bringing in the progressive principles
Now plug these into P1–P3.
Step 1: Misgendering / invalidation via GWₐ
In the actual social context we live in, the discourse pattern:
“Trans women aren’t genetic women. Only genetic women are real women in the sense biology recognizes.”
does at least one of the following:
- Entails ¬W(x) (“in the most important sense you’re not a woman”), or
- Presupposes a hierarchy where the only fully legitimate women are GWₐ, or
- Pragmatically conveys that trans women are “less real” women.
That’s a direct clash with P1 and P2:
- P1 says we owe trans women recognition as women (W(x)).
- P2 says we shouldn’t use language that undermines that, when avoidable.
Given S2, it’s not enough to say “Oh, but I still also call them women.” If your use of “genetic woman” makes “woman” into a marked/secondary category for them, you’re still violating the anti-invalidation norm at the level of presupposition and expressive meaning.
Step 2: Availability of non-harmful alternatives
By S1–S3 and Option B, there exists at least one formulation Tᵦ of genetics such that:
- Tᵦ makes all the same empirically correct claims as Tₐ;
- Tᵦ avoids building in a predicate GW(x) that conflicts with affirming trans women as “genetic women” (or avoids the term altogether);
- Tᵦ therefore satisfies P1 and P2 better than Tₐ.
So, we have:
- Tₐ and Tᵦ are empirically equivalent sequences of claims about G(x).
- Tₐ encodes a discourse where trans women are marked as “non-genetic,” and in practice that is regularly used to question their womanhood.
- Tᵦ does not.
By P3 (anti-structural-harm under underdetermination), we have:
- O(use Tᵦ rather than Tₐ).
That is, other things equal, we are normatively required, on progressive grounds, to adopt the non-exclusionary formulation of genetics.
6. From “ought to choose Tᵦ” to “failure is transphobic”
Now we connect this to transphobia.
Define, minimally:
Transphobic (practice): A linguistic or social practice is transphobic when it needlessly conflicts with the recognition and affirmation of trans people’s genders, or expresses / entrenches the view that their genders are less real or less legitimate than cis people’s.
Then, putting the pieces together:
- Tₐ and Tᵦ are both legitimate as far as genetics is concerned (S1).
- Tₐ systematically encodes that most trans women are ¬GW(x), and in actual practice this is used to question or downgrade their womanhood (Section 5).
- Tᵦ avoids that, while preserving all the same genetic knowledge.
- Progressive norms P1–P3 require us to choose Tᵦ over Tₐ.
- Persistently choosing Tₐ, knowing Tᵦ is available and scientifically equivalent, is exactly a case of “needlessly using language that undermines recognition of trans women.”
So:
Conclusion: Given (i) that genetics can be formulated in ways that don’t encode “trans women are not genetic women,” and (ii) progressive principles about respecting and validating trans women’s identities, it is transphobic to insist on a formulation that does encode that exclusion when no scientific need forces it.
Formally, you can write something like:
-
Let Gen-Adeq(T) be “T is empirically adequate as a genetic theory.”
-
Let Respect(T) be “T does not entail/presuppose that trans women are not real women or not ‘genetic women’.”
-
Let ProgNorm be:
∀T₁,T₂ [Gen-Adeq(T₁) ∧ Gen-Adeq(T₂) ∧ Respect(T₂) ∧ ¬Respect(T₁) → O(T₂ ≻ T₁)]
-
Given: ∃T₁,T₂ such that Gen-Adeq(T₁), Gen-Adeq(T₂), Respect(T₂), ¬Respect(T₁).
-
Then: it is a violation of ProgNorm to choose T₁ when T₂ is available.
-
And that violation is what we label “transphobic” in this domain.
7. What this does not say
A couple of clarifications that fall out of the formalization:
- It doesn’t say any individual who has ever said “genetic woman” in an exclusionary sense is automatically a moral monster. People can be ignorant or unreflective about the semantic options. The argument is about what follows once you accept the progressive principles and know the semantic facts.
- It doesn’t require pretending biology is something it isn’t. It only requires recognizing that biology can be described in more than one linguistically coherent way, and some ways are less harmful than others to marginalized people.
If you’d like, we can next:
- make the semantic piece more model-theoretic (structures, interpretations, alternative lexicons), or
- sharpen the normative side (e.g., deontic logic formulation, or connect it to specific views about oppressive speech acts).
No comments:
Post a Comment