Despite being a lover of the ancient world, I tend to be suspicious when contemporary commentators make a show of bolstering their convictions with a nod to Plato or some other big name of the distant past. It is one thing to trace a careful line of intellectual descent (and dissent) from then to now. This is something I’m all in favor of. It is quite another to cherry-pick an idea from one of the great books and present it, largely out of context, as evidence for a course of action today.
Still, people do this all the time—I’m sure I’ve been guilty myself—and debates over both interpretation and applicability can certainly be gripping. Once in a while, perhaps especially when sexual mores are involved, they even make major news.
Probably the most famous American example in my lifetime took place thirty years ago. In November 1992, nine people, one of them Richard G. Evans, a gay man who worked for the mayor of Denver, filed suit to enjoin Colorado from enforcing an amendment to the state constitution that voters had passed earlier that month. This amendment prohibited the adoption of “any law or policy which provides that homosexual, lesbian, or bisexual orientation, conduct, or relationships constitutes or entitles a person to claim any minority or protected status, quota preferences, or discrimination.” In October 1993, when the case was still in state court, Martha Nussbaum (then at Brown but soon to move to Chicago) served as an expert witness for Evans et al., while John Finnis (Oxford) and my friend and former colleague Robert P. George (Princeton) played the same role for the Centennial State, led by Governor Roy Romer (who was himself personally opposed to the amendment but defended it in his gubernatorial capacity).
Things quickly became nasty.
Things quickly became nasty, with Finnis and George accusing Nussbaum of engaging in the shameless act of furthering her argument by deliberately misconstruing Plato. In particular, they alleged that Nussbaum was intellectually dishonest in her interpretation of what the philosopher meant when he put the word tolmēma, which might be translated as “shameless act,” into the mouth of his character the Athenian Stranger in the Laws (636c, in the standard Stephanus pagination) in connection with sex para phusin (contrary to nature). Nasty, as I say, but not without wry humor as well: as Daniel Mendelsohn put it in the September/October 1996 issue of the much-missed academic rag Lingua Franca, observers “found something comic in the sight of academic superstars earnestly debating Plato’s views on anal intercourse . . . a good 2,300 years after Plato himself presumably rejoined the realm of pure Ideas.”
As it happens, neither the Colorado Supreme Court nor the U.S. Supreme Court, both of which sided with Evans (the former in 1994, the latter in 1996), paid any heed to the heated battle over Ancient Greek lexicography that had taken place in the Denver courtroom. Be that as it may, it is difficult to imagine how even a fervent supporter of the ultimate decision could muster sympathy for Nussbaum’s egregious anti-philological performance, well described in The New Criterion’s “Notes & Comments” from December 1994 and a flurry of further articles from the time, including by Finnis and George in Academic Questions (the former in the Winter 1994 issue, the latter in Spring 1996, plus an addendum in Summer 1997).
“We should not be so quick to conclude that Plato’s final reflection on the art of politics has nothing to teach us” today.
And yet, a close examination of the matter by Randall Baldwin Clark, published in the Yale Journal of Law & the Humanities in Winter 2000, finds cherry-picking on both sides. To some extent, such behavior arises almost necessarily from (in Clark’s words) “the partisan character of legal discourse,” which in this case encouraged everyone to find ways to sway a judge “untutored” in Plato’s Laws, the significantly less famous follow-up and corrective to the Republic. Clark, who calls himself one of “Plato’s scholarly cheerleaders,” is surely right that although the professorial comments on the Laws were “not appropriately dispositive of the case at bar, we should not be so quick to conclude that Plato’s final reflection on the art of politics has nothing to teach us” today (emphasis in original). So, too, I argue, does another underappreciated Platonic dialogue have something to teach us about what is becoming an increasingly consequential question in contemporary life and law: what is a woman?
In the so-called culture wars, the spotlight has moved in recent years from gay rights—a major legal and philosophical issue in the United States from before Bowers v. Hardwick (1986) up through Obergefell v. Hodges (2015)—to trans rights. Every day brings new stories, new controversies. One can spend hours on social media watching clips of activist teachers making pronouncements about both their students and themselves that suggest they are unfit for the classroom and of irate parents shouting at school officials for promoting transgender ideology, often to their youngest charges. Alternatively, one can spend some minutes reading my analysis of pronouns and the law in the March 2023 issue of The New Criterion. But there’s a great deal more to say, and Plato can, I think, help us to say it, even though—or, in fact, because—there are severe linguistic problems with his approach.
People in today’s polarized world are increasingly acting as though words were things possessing truth in their very form. From progressives, we have grown used to hearing the mantra “Words are literal violence” (so is silence, it seems, but that is another matter), which is often associated with the idea that misgendering someone or otherwise not using the right language to or of a trans person is literally denying his/her/their/xir (etc.) existence. And in many contexts, including the medical and legal spheres, people across the political spectrum who have no interest in playing language games are being told to avoid using so ordinary a word as “woman” (alternatives include “menstruator” and “bonus-hole-haver”), with the consequence that a significant percentage of the human race feels slighted.
Much of this is new, but the impulse to draw connections between language and reality is not.
Much of this is new, but the impulse to draw connections between language and reality is not. Consider the Cratylus, one of the least read but arguably most interesting of the Platonic dialogues, which is specifically about the relationship between language and reality. At issue in the discussion Socrates has with two other philosophers, Cratylus and Hermogenes, is whether objects have names (onomata) by nature or are instead matters of local convention. Cratylus espouses the former position, known as naturalism, whereas Hermogenes takes the latter, so-called conventionalism.
The dialogue is enigmatic in many ways. The thrust of the following exchange between Socrates and Cratylus (430a–c), however, is clear. I quote here and below from the excellent translation of C. D. C. Reeve (very lightly revised in places):
Socrates: You agree, don’t you, that it’s one thing to be a name and another to be the thing it names?
Cratylus: Yes, I do.
Socrates: And you also agree that a name is an imitation of a thing?
Cratylus: Absolutely.
Socrates: And that a painting is a different sort of imitation of a thing?
Cratylus: Yes.
Socrates: Well, perhaps what you’re saying is correct and I’m misunderstanding you, but can both of these imitations—both paintings and names—be assigned and applied to the things of which they are imitations, or not?
Cratylus: They can.
Socrates: Then consider this. Can we assign a likeness of a man to a man and that of a woman to a woman, and so on?
Cratylus: Certainly.
Let’s pause and take stock. While contemporary linguists and philosophers have sharp disagreements about the role social convention plays in language—in brief, the late David Lewis thought it was fundamental to meaning, whereas Noam Chomsky does not—all scholars agree that language is conventional in some weak sense. And not just scholars. Leaving aside marginal cases like “Achoo!,” most readers of this essay will be wholly sympathetic to the idea that onomata are not the things themselves. Put another way, a rose by any other name (gül in Turkish, say, or waridi in Swahili) is still a rose.
That said, readers—especially monolingual ones, like most Americans though not the majority of humans—don’t generally think about the relationship between roses and “roses.” Instead, they act most of the time as though onomata were in fact pretty good reflections of what they represent, namely imitations (mimēmata) or likenesses (eikones, whence our word “icons”) of reality. And it is appropriate, therefore, to call men “men” (or, in Greek, andres; compare “android”) and to call women “women” (gunaikes; compare “gynecologist”).
So far, so good. Let us keep going (430c):
Socrates: What about the opposite? Can we assign the likeness of a man to a woman and that of a woman to a man?
What is the answer? Can we assign eikones to the wrong sex? Many will say that of course we can’t. But this is wrong:
Cratylus: Yes, we can.
Socrates: And are both these assignments correct, or only the first?
Cratylus: Only the first.
Socrates: That is to say, the one that assigns to each thing the painting or name that is appropriate to it or like it?
Cratylus: That’s my view, at least.
In other words, we could choose to call a rose a “spade,” but since spades are not roses, to do so would be an instance of incorrect naming—just as it would be incorrect to paint a picture of a spade and say that it represented a rose.
What we have here, and also in what immediately follows as Socrates and Cratylus continue in this vein, is a stark comment on the difference between reality and the language we use—necessarily but imperfectly—to express reality: men and women are creatures in the real world; “men” and “women” are words to express worldly reality; these words can be misused; but, nonetheless, calling men “women” and women “men” does not change reality. The quotation marks are doing important metalinguistic work.
Back to the text of the dialogue. Socrates responds to Cratylus’s opinion about the appropriateness of paintings and names as follows (430d, emphasis in original):
Since you and I are friends, we don’t want to engage in a battle of words, so here’s what I think. I call the first kind of assignment [i.e., of an imitation to the thing it actually resembles] correct, whether it’s an assignment of a painting or a name, but if it’s an assignment of a name, I call it both correct and true. And I call the other kind of assignment, the one that assigns and applies unlike imitations, incorrect, and, in the case of names, false as well.
Calling a rose a “spade” would, it turns out, be more than not correct (ouk orthēn; compare the “orth-” of the orthodontist who corrects your teeth). It would also be false (pseudē, whence our “pseud”).
What does that mean? For Socrates, names capture essence, and essence is revealed by etymological explorations, to which a large part of the Cratylus is devoted. Experts on Plato disagree about just how serious the etymologies are, but it is undeniable that—to put it in overly simplistic form—many premodern scholars of language operated under the assumption that words that looked or sounded alike were, or could without much argument be said to be, connected. Conversely, those that did not were unconnected: not only are roses and spades different entities, and not only do the Ancient Greek words rhodon (rose) and sminuē (spade) have as little in common formally as their English counterparts, but there is presumably something particular about the linguistic structures of these two words that makes each of them true—or, as Goldilocks might say, just right.
The Cratylus is silent on rhodon and sminuē, but there are numerous examples of other words to illustrate the point. Here is a representative passage (398c):
Hermogenes: But what about the name “hero” (hērōs)? What is it?
Socrates: That one isn’t hard to understand because the name has been little altered. It expresses the fact that heroes were born out of love (erōs).
Hermogenes: How do you mean?
Socrates: Don’t you know that the heroes are demigods?
Hermogenes: So what?
Socrates replies at length (398d–e):
So all of them sprang from the love of a god for a mortal woman or of a mortal man for a goddess. And if, as before, you investigate the matter by relying on old Attic, you will get a better understanding, since it will show you that the name “hero” (hērōs) is only a slightly altered form of the word “love” (erōs)—the very thing from which the heroes sprang. And either this is the reason they were called “heroes” or else because they were sophists, clever speech-makers (rhētores) and dialecticians, skilled questioners (erōtan)—for eirein is the same as legein (to speak). And therefore, as we were saying just now, in the Attic dialect, the heroes turn out to be speech-makers and questioners. Hence the noble breed of heroes turns out to be a race of speech-makers and sophists. That isn’t hard to understand.
In fact, from the modern linguistic perspective, this is hard to understand: it can hardly be the case that the Ancient Greek word for “hero” is a lightly modified form of the noun for “love” and also related to verbs that mean “to question” and “to speak.” (The actual etymology of “hero” is unclear, and I admit that the existence of at least three suggestions since the turn of the millennium by major Indo-Europeanists—“someone characterized by booty,” “someone in his flourishing time,” and “someone with durable (bone-)strength”—might make some doubt the solidity of the historical/comparative linguistic enterprise.) In any case, the point is that this and other passages strongly suggest that if Ancient Greek had had the words “man,” “woman,” “manacle,” “Manichaeism,” “monk,” “adamantine,” and “lo mein,” Socrates would not have found it difficult to tie them all together.
The process—an extremely common one—by which people make such linguistic connections is known as folk etymology. The results are often amusing, revelatory, or both. Some connections stick: pretty much everyone imagines that woodchucks chuck wood (whatever exactly “chucking” wood might mean), though the coinage “woodchuck” was rather an attempt by early settlers to make some sense of the name of the animal in an Algonquian language of New England (compare Narragansett ockqutchaun). Others are idiosyncratic: not understanding what “tenterhooks” are makes some people imagine that these hooks, though sharp, are somehow “tender.” And who could forget John Donne’s line “Get with child a mandrake root,” where the root’s well-known anthropomorphic nature is enhanced by the chance resemblance of the word “mandrake” (borrowed from Latin mandragora, which got it from Greek, which got it from who knows where) to “man,” “woman,” and “human” (on all of which see below)?
It so happens that, before his discussion with Cratylus about names as “imitations,” Socrates has already commented on the etymologies of the words for “man” and “woman.” His explanations take off from the singulars of the already-mentioned forms andres and gunaikes, respectively anēr and gunē (414a):
“Man” (anēr) indicates upward flow (anō rhoē). It seems to me that gunē (woman) wants to be gonē (womb).
Transported to twenty-first-century America, Socrates and Plato might not be too surprised, then, to discover such products as “Sacred Womban,” described online as a “very potent, special women [sic] Abyhanga Oil.”
How exactly the Greek philosophers conceived of the relationship between men and women is a matter that goes far beyond the scope of this essay and depends to a great extent on the understanding of Book 5 of Plato’s Republic, in which Socrates suggests that the two sexes have much in common: “men and women are by nature the same with respect to guarding the city, except to the extent that one is weaker and the other stronger” (456a; trans. G. M. A. Grube, rev. Reeve). A proper assessment would have to balance such claims (rejected a generation later by Aristotle) against the idea in Aristophanes’ speech in Plato’s Symposium that “long ago . . . [t]here were three kinds of human beings . . . not two as there are now, male and female” (189d; trans. Alexander Nehamas and Paul Woodruff); the sociopolitical environment of Classical Athens, in which men and women occupied largely different spaces; and much else. What is clear is that, from the point of view of the Greek language alone, men and women have little to nothing in common.
But what if Socrates were etymologizing in English?
Viewed cross-linguistically, the fact that anēr and gunē are entirely different is hardly unusual: witness Mann and Frau in German, hombre and mujer in Spanish, erkek and kadın in Turkish, and so on. When Socrates, using the plurals, says that it is incorrect and false to call andres “gunaikes” or gunaikes “andres,” there is, then, no linguistic reason to doubt it.
Not all languages are like this, though. In Swahili, for example, “man” is mwanamume and “woman” is mwanamke: mwana means “child,” and the designation of sex comes from the addition of either mume (husband) or mke (wife).
And then there is English, where there is a “man” in “woman.” Literally so. (In pronunciation, the vowel in the unstressed second syllable of “woman” is different from the vowel of “man,” but in our literate society, spelling rules.) What would Socrates make of this? Culturally knowledgeable, he would be aware of Genesis 2:22, perhaps in the King James Version: “And the rib, which the Lord God had taken from man, made he a woman, and brought her unto the man.” Since he would not know Hebrew (any more than most Anglophone readers of the Bible do today) and since in at least one place in the Cratylus (410a) he suggests that foreign words may be unetymologizable, I pass over what is going on in the original here and in the following verse (2:23, “And Adam said, This is now bone of my bones, and flesh of my flesh: she shall be called Woman, because she was taken out of Man”). Instead, I concentrate on the effect of the relationship between “man” and “woman” in present-day English—an effect, by the way, that the shared uniqueness of the plural highlights as well since there are no other examples of a plural in the language made by simply changing an a to an e. (Yes, again the second vowel of “women” is subtly different from the vowel of “men,” but this does not affect my argument.)
Was it always the case that a woman was in some sense a man plus “wo-”? Yes and no.
The linguistic relationship has some well-known orthographic consequences: such spellings as “womyn” (usually plural), “womxn” (singular or plural), “wimmin” (plural), and “womban” (singular) are meant to provide linguistic distance from the idea that a woman is dependent on a man, with the last calling attention to her womb rather than his manhood—or rib. The first two, attested from 1975 and 1991, respectively, now have entries of their own in the Oxford English Dictionary; the first three, as Christopher F. Rufo and others have pointed out, are described as “empowering” in a University of Texas at Austin “glossary” from 2023.
Was it always the case that a woman was in some sense a man plus “wo-”? Yes and no. In Old English, the word mann could occasionally be used for a specifically male human being, but its primary meaning was “any member of the species,” as in “mankind,” to use a disfavored term that is now generally replaced by “humankind” (on which more to come). A standard way a thousand years ago to distinguish a male mann from a female mann was through compounding: the former was a wǽpenmann (weapon-man), the latter a wífmann (wife-man). (The weapon was the penis. As for the wife, it is possible that the first element of the other compound has a cognate in the word for “shame” or “modesty” in Tocharian, a long-dead Indo-European language of Chinese Turkestan: kwīpe in one dialect, kip in the other; compare German Scham, which means both “shame” and “private parts, vulva.” But this etymology is controversial.)
Over time, people stopped adding wǽpen to the word for the male and started pronouncing the word for the female in a way that obscured the compound: “woman.” They also started spelling the plural with the same first vowel as the singular, thereby making “women” the only word in English where an o is pronounced as a short i. The only word, that is, aside from “ghoti,” pronounced “fish,” a bit of humor attributed (incorrectly) to George Bernard Shaw that hinges on the non-obvious spellings “enough,” “women,” and “action.”
Was it always the case that a female was in some sense a man plus “fe-”? Here the answer is more no than yes.
It is worth pointing out as well the somewhat different story of “male” and “female.” Was it always the case that a female was in some sense a man plus “fe-”? Here the answer is more no than yes. The word “male” entered our language in the late fourteenth century from the Anglo-Norman development of the Latin word masculus (the diminutive of mās, “male”), which we borrowed directly as “masculine” (and which turned in Spanish into macho, which we have also borrowed). The word “female,” attested in English from the middle of the fourteenth century, is also of Latin and Romance origin, but a standard spelling early on was femele (as in Latin fēmella, “girl,” the diminutive of fēmina, which we borrowed directly as “feminine”). The later spelling “female,” with an a, reflects the influence of “male.” In short, thanks to folk etymology, a “male” is now in “female” that didn’t used to be there.
The Oxford English Dictionary observes, obviously accurately, that “The genderless uses of man to mean ‘human being’ or ‘person’ are now often objected to on the grounds that they depreciate women, and are frequently replaced by human, human being, or person.” But is it not curious that there is also a “man” in “human,” not to mention a “son” in “person”? To be sure, the “-man” in “human” and “humankind” (borrowed from Latin hūmānus, itself a derivative of humus, “earth, soil”) has no deep connection to the nouns “man” and “mankind,” but then the connection between the “-male” in “female” and the noun “male” isn’t terribly deep either.
I stress that there is no reason why someone who is not a linguist should know any of this. Ordinary people do not have access to the histories of the words in the language or languages they speak: they use the words “son” and “person” without making a mental note that the one is a core part of English vocabulary (Old English sunu) while the other is a borrowing from Latin persōna (mask; character), which is itself likely borrowed from an extended form of Etruscan phersu (mask). But everyone—without regard to extent of education—makes folk etymological connections: a son is, after all, a kind of person, just as a man is a kind of human. And so is a woman.
What are we to learn from such meditations? Everyone has to admit that there is in a real sense a “man” in “woman.” Beyond this, however, there is no unanimity. I assume that some readers will agree with one or more of the following four statements while others will vehemently disagree:
(i) There is in a real sense a man in a woman (support might come from Genesis or from the presence of an X chromosome in each).
(ii) Spellings like “womyn” and “womxn” are good defenses against the patriarchy.
(iii) The similarity between “man” and “woman” (and between “men” and “women”) is a sign that men and women are largely the same.
(iv) The fact that “man” and “woman” (and “men” and “women”) are different words, even if they do have something in common, shows that the two sexes, for all their similarities, are not the same.
All but the first of these statements are about English and blur the line between word and world. As we have seen in the Cratylus, that line is regularly blurred in Ancient Greek as well. But because there is no anēr in gunē, the issues that I have brought up in connection with the translation of the dialogue do not arise.
You cannot turn a man into a woman by calling him a woman and you cannot turn a woman into a man by calling her a man.
Certainly you can turn a “man” into a “woman” by adding “wo-” or a “woman” into a “man” by subtracting the same. But in my view—and in Socrates’—you cannot turn a man into a woman by calling him a woman and you cannot turn a woman into a man by calling her a man. It is, of course, possible that Socrates and I are wrong about this. What I insist on, though, is the importance of understanding that language is not always an accurate reflection of truth and reality, that words all too often fail to reflect the world precisely. In a small way, the difference in how the Ancient Greek and English languages speak of men and women suffices to show this. After all, surely no one wants to claim that you can turn an American (or perhaps Swahili) man into a woman but not a Greek (or Spanish) one.
To be human—man or woman—is to play with language: you and I do it, children do it, Socrates and Plato did it. We notice words with similar sounds and spellings and fiddle with them, sometimes consciously, other times not. I take for granted, therefore, that the similarity in English between the words “man” and “woman” does have some effect—small for some, larger for others—on how speakers view the relationship between the sexes. And we may legitimately wonder, I believe, whether Socrates would have said something different about men and women if his words for the two sexes had been more alike. Even if they had been more alike, however, and even if Socrates had said the opposite, this would not change reality.
The reason to believe that men cannot be women and women cannot be men is biology, not linguistics or language games that tack on or remove “wo-.” You may hold the firm view that the sexes are not interchangeable. You may believe to the core that a man does not become a woman by saying that he is one or by undergoing surgery for augmented breasts, and that a woman does not become a man by saying that she is one or by having her breasts removed. But you would be wrong—it would be an instance of lazy thinking—to give as your reason either that an Ancient Greek philosopher said so (as though even a wise man couldn’t be mistaken) or that words in one or another language could not be manipulated to suggest otherwise.
