Them Without Chests - Part II
It doesn't matter where words come from, because they change their meanings. But do they really? Yes, they do. (🤔)
In a heartwarming twist, it turns out the apparatus of modern media hasn’t yet eradicated every trace of serendipity and it’s still possible to have your creativity nourished by happy accidents. As I was sitting down to write this second part of the series, I opened up the New York Times app on my phone to distract myself (as you do) and discovered John McWhorter ruminating on language change (as he does).
McWhorter’s twenty-something-year-old popular-writing cannon is a great place to trace the long-term effectiveness of linguists’ campaign to convince the public that—like it or not—languages don’t sit still. Back in the day, he had to work a lot harder to lift that load. People really get it now, though. And that’s good, because the point is (to a great extent) demonstrably true. But, that said, I think there’s something a little rigid about one facet of the language-change creed: namely the tenet that word meanings shift like clouds in the sky. Even though that’s also sort of true, there’s something too dogmatic, not just about the way the public has picked up the ball and run with it, but even the expansive way linguists themselves espouse it. I’d like to push back a tad.
The more things change, . . .
The realization that word meanings are unstable is nothing new. In a famous passage from the Analects, one of Confucius’ disciples, Zhou You, also known by the endearing nickname Zi Lu [子路], tells Confucius that the ruler of the feudal state of Wei (one of many vying for complete control) wants him (Confucius) to govern the entire country. Zi Lu asks him what his first official act would be, and Confucius’s response has become known as the “rectification of names”:
Confucius said: “First it is necessary to rectify the names.” Zi Lu said: “Is that really what has to be done? You are being too pedantic, aren’t you now? How will you rectify these names?” Confucius said: “Zhong You, you are too unrefined. A gentleman, faced with a matter that he does not understand, takes a skeptical attitude. If names are not correct, one cannot speak smoothly and reasonably, and if one cannot speak smoothly and reasonably, affairs cannot be managed successfully. If affairs cannot be managed successfully, rites and music will not be conducted. If rites and music are not conducted, punishments will not be suitable. And if punishments are not suitable, the common people will not know what to do. So, when the gentleman uses names, it is necessary to be able to speak so that people understand. If one can say it, one can definitely do it. A gentleman should not be careless with words.”
As the Chinese philosopher Cao Feng (whose translation I’m using) notes, Confucius’ plans for “rectification” have ironically led to some serious confusion because the word he used for names (ming 名) could maybe be interpreted as referring to characters (zi 字) (as in the logograms of the language) or to the social positions—social status (mingfen 名分)—which, if they could be clearly defined, could reorder the then-chaotic political life of the country. Or maybe he was proposing a whole semantic theory! Or maybe a little bit of all of that. In any case, no matter which of those possibilities is really what Confucius intended, it was the shifting meanings of words that compelled him. So the discontents of language change were not unknown. 600 years later, the philologist Xu Shen traced the changing meanings of characters in his Shuowen Jiezi (說文解字), the earliest Chinese dictionary.
Meanwhile, about a quarter of the way around the world—and by the standards of 2500 years ago, it might as well have been on Mars, so distant was it from any possible cultural exchange—Plato wrote one of his dialogues on the topic of language change and the dodgy promises of etymology: the Cratylus. As always, it involves someone striking up a conversation with Socrates (whose workday apparently consisted of getting up, eating a fig, brushing his teeth with eggshell and bone dust, kissing his wife Xanthippe goodbye, and heading to the agora to quarrel with people). In this case he’s approached by two men having a squabble about whether the connection between words and things is natural or just an arbitrary convention.
The defender of the former viewpoint is Cratylus, who believes that the names of things are divinely provided. His sparring partner, Hermogenes, thinks that’s bull plop: names just come about through back-and-forth interaction; someone somewhere calls a dog a dog, and people follow along. In some other place, folks might use a different string of sounds to point to the same thing. He’d like to know what is “natural” about names and is frustrated that Cratylus isn’t forthcoming about where his so-called knowledge comes from. And he’s particularly pissed because Cratylus has the nerve to suggest that Hermogenes isn’t his natural name! So they appeal to Socrates to help settle their argument.
At first, the Gadfly of Athens seems to be thoroughly on team Cratylus. It might not be the case, he admits, that entire words are perfect correlatives, in every curve and line, of the things they refer to. But words work by describing things. Often they’re made by combining other words (what we now call compounds) or bits of sounds that already mean something (in our terminology, words with affixes). And those sounds are made of even more basic sounds. Those most basic sounds, Socrates asserts, have the power to describe by imitation.
He gives lots of examples of what he insists are “appropriate” sounds for imitating nature. Here’s one: ρ, which you can think of as a little like our letter r, but not really—our English sound is actually the post-alveolar approximate [ɹ]; in Attic Greek it was probably1 a voiceless alveolar trill [r̥] at the beginnings of words, maybe with a little breath beforehand, and voiced with the vocal chords in the middle of words [r]; think of your worst stereotype of a Scottish person rolling the r in “Yerrrr off yerrrr head!” You get it. Anyhow, Socrates says that the sound [r̥], which he identifies as the written letter rho (ρ), “seems to me to be an instrument expressing all motion[;] . . .”
a fine instrument expressive of motion to the name-giver who wished to imitate rapidity, and he often applies it to motion. In the first place, in the words ῥεῖν (flow) and ῥοή (current) he imitates their rapidity by this letter, then in τρόμος (trembling) and in τρέχειν (run), and also in such words as κρούειν (strike), θραύειν (break), ἐρείκειν (rend), θρύπτειν (crush),κερματίζειν (crumble), ῥυμβεῖν (whirl), he expresses the action of them all chiefly by means of the letter rho; for he observed, I suppose, that the tongue is least at rest and most agitated in pronouncing this letter, and that is probably the reason why he employed it for these words.
As we (excluding modern Greek speakers, maybe) don’t use the sound he refers to systematically, it’s hard to understand the correlation between sound and meaning he’s positing (which makes it hard to feel like his point is warranted). We do have certain intuitions like this; but we resist thinking of them as natural.2 In any case, Socrates goes off on a series of etymological derivations, at Hermogenes’ request, finally grounding the “naturalness” of words in nine such “appropriate” sounds.
But then Socrates deftly turns on his heal. The “name-giver” he refers to in the quotation above is a synecdoche for all of the ancients who preceded them and used—or sometimes failed to use—the “appropriate” sounds to describe things. Their choices of words may have had “fitting” phonetic qualities for the meanings they carried, but sometimes that very fitness betrayed cultural values that maybe weren’t so good! Socrates, as Plato’s voice box, believes there are eternal forms for all things, and inasmuch as the ancients associated goodness with sounds of motion and change and Heraclitian flux, and not with sounds of fixedness—fixed like the perfect forms, of which we have only an inkling—then that laid bare their own intellectual folly! Therefore, Socrates argues, knowing the origins of words won’t lead you to the “truth.”
Language change appeared as a theme throughout the Roman era. The transformations of mythological figures in Ovid’s Metamorphosis, for example, are mirrored by etymological wordplay, manifesting his theme that the gods will change, just as nature will change, and language can be no different.
But another figure made a contradictory case: namely that some simple and judicious rules of language—indeed, of all modes of expression—are universal and immutable. This figure styled himself (so to speak) as the lodestar of formal Latin oratory, and it’s hard to argue with his self-confidence. As the Classicist Stephen Harrison writes3, “Latin literature in the period 90–40 BC presents one feature that is unique in Classical, and perhaps even in the whole of Western, literature. Although it is a period from which a substantial amount of literature in a wide variety of genres survives, more than 75 per cent of that literature was written by a single man: Marcus Tullius Cicero."
That would be enough to apotheosize any writer, but it was Tully (as English scholars in the early modern period called him) who laid the groundwork for beliefs about discerning language for millennia. They broke the mold with Cicero. His writing was prodigious (nothing but his nemesis Mark Antony’s even more priapic ambition could cut it down). And with some assistance from his fourteenth-century-CE hype man, Petrarch—who rediscovered his letters to Atticus, Quintus, and Brutus in Verona in 1345—Cicero’s style inaugurated the Renaissance and served as an example of a refined and cultivated mind for figures of the Enlightenment: for empiricists like Locke and Hume, and rationalists like Christian Wolff; but also for prickly poets like Dryden and Pope.
Cicero certainly didn’t come up with his aesthetic principles whole cloth, but social and historical accidents have left him—and the Latin style he espoused—enthroned as the non plus ultra of verbal concision and clarity. He writes in De Oratore that Latin speakers should be careful in their choice of words and their wielding of verb tense, and noun number, person, gender, and case:
Atque, ut Latine loquamur, non solum videndum est, ut et verba efferamus ea, quae nemo iure reprehendat, et ea sic et casibus et temporibus et genere et numero conservemus, ut ne quid perturbatum ac discrepans aut praeposterum sit . . . .
(When we speak in Latin, we must not only be careful to emphasize those words that no one can justifiably criticize [i.e. as misplaced or misused], . . . [but we must also] arrange them in such a way that, through the use of the cases and tenses and gender and number, no confusion or preposterous discrepancies arise . . . .)
As for the way speakers pronounce words, and enunciate them with more or less volume, he declares his fealty to the goldilocks rule: He likes neither hearing people articulate each vowel and consonant with forced precision (Nolo exprimi litteras putidius,) nor obscured by slackness (nolo obscurari neglegentius); neither voiced with quiet, hasty breathlessness (nolo verba exiliter exanimata exire), nor with excessive loudness, as if they were “heavily panting” (nolo inflata et quasi anhelata gravius).
That’s a little too finicky for my tastes. Nevertheless, I’d allow it if he didn’t go and ruin things by saying it grinds his gears hearing someone “who is practically devoid of rhythm and out of tune” (aut quasi extra modum absona atque absurda) or—and here’s the kicker—who speaks too softly and effeminately (mollis vox aut muliebris; literally “like a woman”). I’m not surprised Cicero harbored sexist attitudes, but they don’t jibe with my values, and so it makes me question his critical faculties. And it gets worse. He rails at certain speakers (Catullus) who affect a countrified accent because they think it sounds old-fashioned (quod loquitur, priscum visum iri putat, si plane fuerit rusticanum)—and “old-fashioned” implies authentic, authoritative, and real.
If Cicero were simply critical of the whole pretense that older language is necessarily more authentic, that would be fine. But he’s not. On the contrary! He thinks the rustic vocal affect of speakers like Catullus is phony—precisely because it isn’t genuinely old fashioned.
This is where the wheels really fall off of Cicero’s logic. As an example of someone whose language has none of the supposed affectedness of Catullus, he cites his mother-in-law Laelia. Cicero chooses her because it’s easier for women—he claims (I’m just the messenger here)—to stick to “pure” old pronunciation, since they don’t talk to a lot of people and therefore retain the accents they heard growing up. When he listens to Laelia speak, he says, it feels like he’s listening to Plautus or Gnaeus Naevius: the sound of her voice is supposedly so unaffected and natural that she doesn’t let in even a trace of showiness. He figures listening to her is like listening to her father or her father’s ancient ancestors speak.
Meanwhile, Catullus’s (by Cicero’s lights) janky bucolic accent and his use of “vulgar”4 slang, and the way he uses an E sound [iː] in place of the “shorter” I [ɪ], are all signs that he’s not inspired by the “real” old-fashioned users of Latin, but rather by farmers and working-class folks.
This judgement is utterly wrong. There is no language that isn’t first established by farmers and working-class folks, or at least by people on the margins who have no interest in maintaining a power hegemony like the big wigs (e.g. Cicero) do. It is the meek who are the drivers of language change. And among those drivers, women are the main innovators! Far from unsullied, virginal, recluse vessels of “pure” language, women have traditionally had no easy access to privileges or rights, and so why should they give a frick about language rules? Maybe that wasn’t the case with Cicero’s patrician mother-in-law Laelia. But she was, let us assume, . . . an outlier.
In any case, it’s safe to say that “language change innovators” were not a group on Cicero’s radar (language change perpetrators, maybe). But they ought to have been! It was certainly the farmers—and in particular the farmer women—who for innumerable generations had picked and pulled at vowels here, battened down consonants there, and used words in new ways—and even grouped some of them together to give them further new meanings—until from out of the Proto-Italic language had emerged the Osco-Umbrian and Picene languages, as well as the Latinian language that had eventually spread from the region of Latium to push out its cousins and become the language of Empire. It wasn’t the elite who made Latin—they just made it elite.
And though Cicero might not have liked it, it was the laborers and the ladies in the provinces whose “vulgar”5 Latin would give rise to not only Italian, French, Spanish, Portuguese, and Romanian, but also Aragonese and Asturian, Galician and Jèrriais, Occitain and Piedmontese, and about 40 other Romance languages.
Given his iffy batting record, when Cicero posits the etymologies of the words superstition and religion in De Natura Deorum (On the Nature of the Gods), you might guess he gets it wrong. And you’d be right, sort of. Trying to clarify just what the distinction between those two ideas is (a worthy question, and one we still ask today), he appeals to the words’ origins to get at something important about them: “Non enim philosophi solum,” he writes, “verum etiam maiores nostri superstitionem a religione separaverunt” (“Not only philosophers, but also our ancestors separated superstition from religion”).
Regarding superstition, Cicero insists that it is in reference to the desire to make sure one’s children survive them. He paints a fabulous image of men spending days in prayer and sacrifice “so that their children might be their survivors” (“ut sibi sui liberi superstites essent”). These men were supposedly therefore labeled “superstitious.”
Now, the word superstitio definitely comes from superstare, which in a more literal context could mean “stand above/over,” but usually just meant “overstand” (i.e.“survive”). Cicero seems to think the -sti- affix in supersti- comes from the plural nominative of superstes (“survivor”)—namely superstites. That’s why he specifically claims that men perform superstitious rites so that their children might survive. He needed a third-person plural survivor to explain the morphology of the word. Superstes has multiple meanings, however. For one, it could mean “remains,” as in “what remains after death”; that is to say “the body,” but also one’s personal effects. Second—and this is the salient sense—it could mean “survival.”
Supersti- is probably a contraction of the subjunctive past participle stem superstetis-. The process by which that contraction took place was, as far as we know, simply unintentional phonological change: People dropped the -te- from superstetis- (which often happens when two similar sounds butt up against each other), and they found it handy to use the participle form as a noun, so you got superstit-; then they started adding the suffix -io, which usually means “the act of doing,” so you get superstitio; and boom!—Bob’s your uncle. But not Cicero’s uncle. In his mind, no proper word could have evolved by such haphazard means. For Cicero, unintentional language development wasn’t a thing. The ancestors actively meant to do it.
Meanwhile, regarding religion, he insists that religio comes from relegere. The word legere (“read”) comes from a root for “choose,” “gather” or “collect,” which has heritage thousands of years prior to the evolution of Latin. But that metaphorical sense was clearly apparent in the minds of Latin speakers in Cicero’s day. Otherwise he wouldn’t be able to coherently argue to his readers, as he does, that legere is the common denominator in diligens (“diligence,” from diligere, a combination of dis+legere, meaning “select apart”) and elegans (“elegance,” from eligere, meaning “carefully choose”),6 as well as religio. In other words, Cicero believes religio comes from “reread” or “recollect.” So in his conception, those performing religion (as opposed to superstition) go through the process of regathering all the important things (either in reading or in thinking).
Only, there is no evidence that Cicero is right about that etymology. A tradition carried on from the fourth century CE, starting with the Roman grammarian Servius, to the Christian philosopher Lactanius, on down to St. Augustine, bases the etymology of religio on a reading of a line by Lucretius: “religionum animum nodis exsolvere” (“to untie the knots in religious minds”). That tradition considers the root of -lig- in religio not legere but rather ligare (“bind” or “tie.”)7 Most scholars consider that etymology more parsimonious and therefore more plausible, and—barring any less circumstantial evidence—plausibility wins the day.
But you should note a tension in Cicero’s etymologies: In the case of superstition, he basically gets the constituent phonological parts right; though in his determination to understand where the -sti- comes from, he overdetermines. There’s no reason to think the human desire for survival is specifically about a person’s children just because you insist that the morphology of the word comes from a third person plural form. Survival in general is enough. After all, there is no theme more universal and ancient than the tension between, on the one hand, belief in the magical power to wrest control from death and, on the other hand, faith, which is still belief but embraces uncertainty about what dreams may come after death. What is the Epic of Gilgamesh about if not that? De Natura Deorum is itself an important contribution to that tradition.
At an even more basic level, this idea of standing as “resisting” seems to pervade our vocabulary: understand, withstand, I can hardly stand it. The act of opposing the force of gravity—possibly the most rudimentary conflict of human existence—so captures our imagination that we English speakers call upon it in order to make ourselves (as it were) understood in other domains. And the historical phonological record shows us that we share that same metaphor with the rest of our Proto-Indo-European language family.
Meanwhile, in the case of religio, Cicero probably gets the phonological history wrong. And yet his assertion that religio stems phylogenetically from relegere—mistaken as it is—enriches our semantic concept of religion. The notion of rereading, regathering, recollecting, and remembering calls to mind the purpose of creeds. And beyond that, you can’t help but think that the idea of “tying” or “binding” isn’t all that distinct from “gathering” or “collecting,” and you might be wondering whether maybe Cicero didn’t tap into something deeper connecting legere and ligare.8
In any case, the mixed bag that was Ciceronian scholarship—at one pole, perceptive insights into the human condition, as well as precise, elegant language; at the opposite pole, entrenched power, chauvinism, and essentialism about inherited language and tradition; and in between a consolidation of historical awareness, style, and values—held fast for generations. In the English language, it reached its zenith in the eighteenth century.
Samuel Johnson often professed a certain disdain for the literature of Classical antiquity. He thought Hellenic and especially Roman culture was too hung up on fame and prestige and didn’t sufficiently appreciate down-to-earth modesty. One of the most noteworthy instances of him making this point is specifically at Cicero’s expense. In Rambler No. 118, Johnson says that
Cicero has, with his usual elegance and magnificence of language, attempted, in his relation of the dream of Scipio,9 to depreciate those honours for which he himself appears to have panted with restless solicitude, by shewing within what narrow limits all that fame and celebrity which man can hope for from men is circumscribed.
Basically, Johnson thinks Cicero pays good (rapturously good!) lip service to the foredoomed way of all ambition, but that Cicero’s own biography proves he—and the whole culture around him—was nigh-on defined by colossal ambition, and so Cicero was sort of a hypocrite.
First of all, Johnson’s not wrong: Cicero was ambitious, even delusional in his conviction that he knew best how to define what Rome should be. And I have the same visceral reaction when I read the Classics, from the pre-Socratics on down: “Tone it back, yo!”
Maybe that’s unfair. I live in another time, and so perhaps Cicero’s way of understanding the world is completely incommensurate with my own. As Johnson himself writes of Homer, “Time and place will always enforce regard. . . . [C]onsideration must be had of the nature of our language, the form of our metre, and, above all, of the change which two thousand years have made in the modes of life and the habits of thought.” I can’t set myself back through all of the ethical paradigm shifts that have spanned from Cicero’s or Homer’s time to now. Yet, as Steve Metcalf argued to me in this special episode of A Million Little Gods on changing moral standards: “One interesting question is, if we determine constancies across that many paradigm shifts, and that’s why we still read Homer, is there a feature of human nature that is a constant, or functions as a constant, that allows us to inherit his worldview in some way that’s still sensible with our own?”
I think there are constancies. I believe in a moral law that manifests itself in human practical decision-making, and I think it has always been at work. All of that notwithstanding, however, I have to admit that the level at which our moral lights guide us and how they dovetail with our use of language is opaque. To flip Gertrude Stein’s famous turn of phrase (on which President Biden recently riffed) on its head: I’m sure there’s a there there. But the “moral law within” lies far deeper than I can peer. I do, however, have an intuition about how it expresses itself in social reality: I posit that our inheritance—not only of morals but also of words and phrases, which are sometimes the lexical correlatives of morals—is not simply stemmatic or phylogenetic. Both sets of phenomena—words and phrases on the one side, sentiments on the other—double back on themselves. They reticulate and re-polinate and can’t be clearly traced back from twig to branch to bough to stem.
That’s true of Samuel Johnson’s relation to Cicero. Far be it from me to begrudge Dr. Johnson the chance to indulge in a little anxiety of influence. But when he pays Cicero a backhanded compliment, calling his own Rambler essay a “humble version” of Cicero’s diatribe against ambition, he’s obfuscating the massive debt he owes.
What Johnson inherits from Cicero isn’t his methods or ideas, but his sensibility. We think of Samuel Johnson as the defining figure of eighteenth-century London. Entire tour guides are designed around tracing the material remains of his milieu. The etymological understanding of English on display in his Dictionary, his magnum opus, is incomplete, to say the least, but it was a massive contribution. And Johnson was a brilliant essayist and occasional poet, a sage purveyor of public wisdom, a critic of style, and an arbiter of the “goodness” of language. The border between those roles was absolutely porous in Johnson’s eyes, and in that way he is consummately Ciceronian.
The other work, besides the Dictionary, that solidified his reputation for acuteness of taste and expensiveness of scope was his multivolume series The Lives of the Most Eminent English Poets. Like his Roman forbearer, his critiques are usually general—elegance or majesty here, dullness or baseness there, sort of like a Mad Libs book that’s restricted to only using words that make sense in context (which is a pretty good description of how large language models like ChatGTP work, by the way10)—until he strikes upon a subject he either geeks out over or hates.
The twentieth-century scholar and critic William Empson sometimes channeled Johnson’s mode of social-critique-embedded-in literary-commentary, but he far outshined Johnson in keenly parsing the use of language as a medium. However there are moments in the Lives when Johnson rises to Empson’s level of discernment. My favorite example of such a moment is in the Life of Pope, when he addresses the famously clever, self-referential stanza from the “Essay on Criticism,” in which Pope tries to mirror the points he’s making about phonetics and meter in the form of the very lines in which he makes them. “The sound must seem an echo to the sense,” Pope writes, and then proceeds to effect exactly that idea in the lines that follow.
As Paul Fussel notes,11 Pope changes out the standard iambs for spondees at the metrical feet “/too la/” and “/move slow/,” which makes for a “cacophony.” And when Pope places several sibilant post-alveolar fricatives [s] beside each other in “/Ajax Strives some/,” we have to struggle to get them each out distinctly. Together these stylistic choices supposedly wear us out, and it “makes us one with Ajax.”
Johnson acknowledges this slowness of Ajax and the swiftness of Camilla, and even points to a line about Sisyphus laboring with the rock from Pope’s translation of the Odyssey, where Pope achieves the same effect even more forcefully:
“Who,” Johnson asks, “does not perceive the stone to move slowly upward, and roll violently back?” But then he offers a self-composed parody of those lines with exactly the same number and types of metrical feet, only it doesn’t seem to have any of the lumbering slowness violently interrupted by sudden rapidity:
One has to conclude, Johnson argues, that though they may seem to be influenced by choices of words and meter, in fact the aesthetic effects of such lines are really mostly achieved by connotation: “[I]n such resemblances the mind often governs the ear, and the sounds are estimated by their meaning.” Johnson admits that these aesthetic effects are real, but he’s deeply suspicious of pursuing them through the cultivation of technical prowess: “[W]hen real [they] are technical and nugatory, not to be rejected and not to be solicited.”
Johnson voices his suspicion of close reading and metrical tinkering even more loudly in Idler Nos. 60 and 61, where he invents an earlier (more absurd, less murderous) version of Tom Ripley: Minim the Critic, a Brewer’s son who comes into good fortune, resolves “to be a man of wit and humour,” and goes about infiltrating high society by hanging around coffeehouses and learning the art of gossip until he gets the hang of the critical trade. “He is the great investigator of hidden beauties,” Johnson describes Minim, as his skills at imposture grow stronger, “and is particularly delighted when he finds ‘the sound an echo to the sense.’”
Ironically, in order to clarify his disdain for trifling close inspection of technique, Johnson has to turn to just such close inspection in The Life of Pope. And it’s doubly ironic that, in so doing, his critical skills rise to the level of Empson. It’s paradoxes like those that make us humans hard to parse and still separate us—so far, at least—from the large language models.
But Johnson’s greatest contribution was to lexicography. His Dictionary embodies one of those historical moments—like the period between the two English Civil Wars, when the Levellers and Diggers somehow carried within them the germ of both modern laissez-faire liberalism and proto-communist socialism—when things we now hold to be diametrical started out as the selfsame thing before breaking apart like an unstable element.
As Michael Adams argues in his excellent retrospective piece for the magazine Humanities, “What Johnson Really Did,” Johnson created the arch parody of the authoritative dictionary (think of McSweeny’s The Future Dictionary of America) before there was an authoritative dictionary. His self-deprecating definition of Lexicographer is legendary:
He also established the authoritative dictionary he was parodying. As Adams explains, Johnson didn’t invent most of the features of the modern dictionary—he codified them: Johnson didn’t invent the division of words into senses, but he insisted on it. He didn’t invent the quotations dictionary, but he realized that adding them would make dictionaries a source of literary pleasure and, more importantly, established quotations as the necessary attestations of word senses. He didn’t invent the idea of a conservative prescriptivism that would shore up the language in the face of consistent change, but he found ways to make the case so subtly, in language so urbane, that we still defer to him even today (as I did in Part I).
“This, my Lord,” Johnson wrote in 1747 to Philip Dormer, Earl of Chesterfield, whom he was asking to be his patron in the creation of the dictionary,12 "is my idea of an English dictionary;"
a dictionary by which the pronunciation of our language may be fixed, and its attainment facilitated; by which its purity may be preserved, its use ascertained, and its duration lengthened. And though, perhaps, to correct the language of nations by books of grammar, and amend their manners by discourses of morality, may be tasks equally difficult, yet, as it is unavoidable to wish, it is natural likewise to hope, that your Lordship’s patronage may not be wholly lost; that it may contribute to the preservation of ancient, and the improvement of modern writers;
[ . . . ]
We are taught by the great Roman orator,13 that every man should propose to himself the highest degree of excellence, but that he may stop with honour at the second or third: though, therefore, my performance should fall below the excellence of other dictionaries, I may obtain, at least, the praise of having endeavoured well[.]14
Oh, he got plenty of praise. Until he didn’t. Embedded within his own methodology (the insistence on enumerating different senses of words) was the seed of an idea that would topple his urbanity and his Ciceronian authority. For his vaunted milieu, the hierarchy of the eighteenth century, was already in twilight—and one shifting paradigm was the rejection of the idea that “the preservation of ancient” could also serve in “the improvement of modern” language.
. . . the more they stay the same.
The characteristically taut and readable formulation of the point in John McWhorter Times piece, which I started this essay with, is instructive:
The central point is this: The fit between words and meanings is much fuzzier and unstable than we are led to suppose by the static majesty of the dictionary and its tidy definitions. What a word means today is a Polaroid snapshot of its lexical life, long-lived and frequently under transformation.
The Polaroid snapshot simile is good, but he didn’t make it up ex nihilo. A little over 100 years ago, the Swiss linguist—and, along with Charles Sanders Peirce, one of the founders of semiotics—Ferdinand de Saussure proposed that historical philologists should stop trying to braid together their long-since frayed beliefs about how meaning happens in language. His alternative model took hold and now serves as the basis for McWhorter’s snapshot comparison.
The fraying began as early as the sixteenth century, when travelers to the the Indian subcontinent began noticing patterns of similarity between Indo-Aryan, Persian, and European languages. In 1647, a smartypants ahead-of-his-time Dutch scholar named Marcus Zuerius van Boxhorn conjectured that most of the European languages, plus Turkish and Persian, were all related and probably had a common ancestor. In the nineteenth century, scholars like Rasmus Rask and Franz Bopp started noticing consistent patterns of morphological and phonological change in the written records between languages. These patterns indicated the changes didn’t have anything to do with the words’ meanings; you could often just explain the changes by considering their “sound environment” (i.e. the other sounds in the word that surround the particular vowel, diphthong, or consonant exhibiting the pattern of change). Sometimes the patterns were so strong and frequent that you could reliably pick out any word with a certain phonological feature in the right sound environment and be sure it would have changed correspondingly over time: a sound law.
Tracing such laws was the groundwork for a giant leap in comparative linguistics. Scholars, in particular August Schleicher, pieced together like a puzzle the innumerable patterns to reconstruct what must have been the language spoken 5000 years ago by the Kurgan-building people of the Pontic steppe of what is now war-torn Ukraine: Proto-Indo-European.
I can think of a few mind-boggling collective human accomplishments, but the reconstruction of PIE certainly counts as one of them. And it didn’t take long for scholars of the other language families to follow suit. The comparative historical tracing of all languages has been a tremendous success, but it came at a price that shouldn’t be too hard to adduce: As I noted in “The Awful Other - Part 4,” There’s “something counterintuitive about the inclusion of linguistics as one of the four fields of anthropology.”
We tend to think of language as deeply personal and bound up with our conscious experiences, and it seems awkward to regard the idiosyncratic expressions of people’s minds as natural phenomena to be described and cataloged.
Certainly when they are written, but even when they are fleeting, temporal oral phenomena—curiously intoned, ambiguous in duration, with an onset that becomes surprisingly hard to pinpoint, the closer you inspect it—words are material culture. But they are also semantic signs. It is in the latter role that we find it hard to accept that words change. We want to poke our fingers into some etherial space in our communal head where the words are, point at them, and say “I’m talking about this thing here, not that thing there.” And yet at the end of the nineteenth century, one scholar showed that, ironically, it is specifically because language is a mental and social phenomenon that we can’t think of words as permanent.
The American philologist William Dwight Whitney, whose birthday was 194 years ago yesterday,15 was a pretty impressive grammarian of Sanskrit. But his lasting contribution to linguistics was to call into question some general assumptions that guys like Rask, Bopp (Whitney's own teacher), and Schleicher had made in comparing languages. They had assumed that by looking into ancient languages like Sanskrit, Greek, and Latin, they could see complexities that had been “lost.” And they believed that when, by means of stemmatics, they reconstructed even more complexities and brought them back to light, they were getting back to the language.
It’s revealing to read a letter describing Sanskrit grammar with approval as “of a wonderful structure; more perfect than the Greek, more copious than the Latin.” Once it had been resconstructed, PIE was treated as an organism, or at least as a species, the individual tokens of which had lost something important over time. That was exactly the parallel August Schleicher drew in his book Die Darwinische Theorie und die Sprachwissenschaft (Darwinian Theory and Linguistics): languages were “organisms of nature.” Despite wrapping themselves in Darwinian robes, the nineteenth-century philologists were the successors of Cicero and Johnson. As Stephen Alter notes in his biography of Whitney, their fundamental assumption was that the loss of grammatical features like inflection markers was just that—a loss:
It therefore seemed reasonable to conclude that the entire Indo-European family had undergone a loss of morphological structure over time. This was the view not only of the first generation of comparative philologists—led by Friedrich Schlegel, Franz Bopp, and Jacob Grimm—but also of their midcentury counterparts, includ- ing Ernest Renan, Georg Curtius, August Schleicher, and Max Müller. Curtius reaffirmed the idea as late as 1871: ‘‘That the full forms are prior to the weaker forms is the basic, hardly disputable assumption underlying all of comparative grammar.’’16
Leaving aside the fact that this was a pretty fundamental misreading of Darwin, it also eventually couldn’t face the onslaught of empirical evidence to the contrary. Languages don’t always lose complexity over time. They develop new and different tools that are fit to deal with the problems they face. But the nature of those problems is very different, David Whitney insisted, from that of a nut too hard for a bird’s beak to crack. They exist intersubjectively: in the nexus in and between the minds of people.
This is where Saussure enters the picture. As he writes in his Cours de linguistique générale, “A first impetus was given by the American scholar Whitney, the author of Life and Growth of Language (1875).”
Shortly afterwards a new school was formed by the neogrammarians (Junggrammatiker), whose leaders were all Germans: K. Brugmann and H. Osthoff; the Germanic scholars W. Braune, E. Sievers, H. Paul; the Slavic scholar Leskien, etc. Their contribution was in placing the results of comparative studies in their historical perspective and thus linking the facts in their natural order. Thanks to them, language is no longer looked upon as an organism that develops independently but as a product of the collective mind of linguistic groups.17
The obvious subsequent question to ask is, “Well then, what are words?” Saussure’s answer is complex, and it would take vastly more space than I have left in this essay to do it justice. But I’ll do my best to explain the part of the Cours that licenses McWhorter’s conviction that words are “snapshots” in time. And then I’ll venture to offer a bit of redemption to the likes of Confucius, Plato, Cicero, and Johnson.
The first thing to point out is that Saussure aims to explain something much bolder than simply “words” in the Cours. As the citation on the plaque at the beginning of this section states, he wants to examine the “life” of signs in society. That probably seems like a weird thing to say. To clarify, “life” here isn’t the phenomenon that happens among some forms of matter that we label biological.18 He's got a different sense in mind: the nexus of social interractions of signs. Now you might be thinking, “Okay, I can accept that words signify concepts, and so sure, we interact with each other using our concepts. But, really, come on, it's us people doing the living and interacting.”
If you’ve gotten that far, my job’s basically done. Saussure’s next move is (at least he was convinced) a checkmate: Indeed, we have our subjective minds, and indeed we speak about the concepts in our minds; but speech (la parole) is not language (la langue). Language is an infinitely complex structure that is held together by the bonds of interrelation between signs. And that structure is always in flux. (Sorry Plato.)
Speaking of chess, one of his most famous analogies was to that game: At any point, there is a nexus of relationships between the individual figures defining the state of the game. If you move one piece, not only does that one individual piece change position, nor one relationship change—all of the relationships change, and therefore the whole structure. And here’s the kicker: The mere fact of their being abstract doesn’t stop signs, relationships, or the language structure from being real factual things in the world.
This had major implications for the science of linguistics. We can simply assume that the signifier (signifiant, the physical form of the sign, such as a word or sound) and the signified (signifi, the concept or meaning that the sign represents) is arbitrary and conventional. The meaning of a sign is determined by its relationship to other signs within the system. All of the “meaning” happens within the system, not in the signifiers.
Now the field of semantics, which studies how meaning occurs in language, isn’t hemmed in by Saussure’s structuralism. There are a lot of big ideas about that floating around. But Saussure did establish an epistemic virtue for all branches of linguistics: You can study language from a synchronic perspective, or you can study language from a diachronic perspective, but you shouldn’t cross the streams. That would be bad.
The diachronic perspective—studying language change over time—helps us understand how languages evolve in their morphology and organization, and how they’re shaped by historical and cultural factors. It can also reveal how linguistic structures have been influenced by speakers’ cultural backgrounds, as well as by social and political factors.
Meanwhile, the synchronic perspective—studying language as it exists at a specific point in time—gives us McWhorter’s vaunted snapshot of the linguistic system at a particular moment. It helps us understand the internal structure and organization of the language, and it also makes it possible to study language in use and how it functions in communication.
Misconstruing the two leads to mistakes like those made by the nineteenth century philologists, who confused diachronic change for synchronic change: it ends in poor value judgements. You’re better off just assuming that the whole system has changed, and not get upset when toilet goes from the French word toilette, “a cloth or bag for clothes” → “a fine cloth cover on the dressing table for the articles spread upon it” → “the articles used to get dressed, like bottles and combs and brushes and mirrors” → “the act or process of dressing” → “a dressing room” → “a dressing room with a lavatory attached” → “a lavatory” → “the porcelain plumbing fixture in a lavatory.” It all ends with a cloth, anyhow. (🧻)
And yet. And yet! And yet I can’t help but find the expansive, deep-time array of diachronic changes that I traced in the Etymological Interlude endlessly fascinating—synchronically. Some etymologies are self-evident, almost tautological. You don’t need to know them because you sort of already knew them anyhow. But other etymologies are almost too delicious. Learning them is an experience almost like serendipity; like having an unfocused craving, but then finding something that perfectly satisfies it.
When I consider that shit and shed and science and conscience and schizophrenia and scissors all trace their heritage to the PIE root *skei-, I feel like I know something important about what (or how) each of those words make meaning happen now. It makes me think maybe our ancestors really did separate superstition from religion (maiores nostri superstitionem a religione separaverunt) and Saussure can go jump on an egg or suck on a lake.
And not only our ancestors, but also philosophers feel the pull toward the dense mass of certain words’ ancestry. When I was researching Best Behavior - Part I, I discovered that Christine Korsgaard had written an entire essay, “Conscience,” in which she appeals to that word’s etymological components to define it as “knowing something in common with someone,” in comparison with the Greek word it was a translation of, συνείδησις (synedeisis), which could be defined by its components in the same way, or as “knowing something in common with oneself.” She traces the diachronic addition of senses to those words, and that of the related synderesis, and eventually appeals to their etymologies to undergird her grounding of universal moral law in the integrity of the self: you know the moral law in common with yourself. As I’ve already said, I’m not on board with that. But that doesn’t mean that Korsgaard hasn’t hit on a way in which the word conscience and its component parts make meaning across time.
I have an intuition that this power has something to do with the notion of “conceptual metaphors” laid out by Geroge Lakoff and Mark Johnson in their pioneering works in cognitive linguistics, Metaphors We live By and Women, Fire, and Dangerous Things. Sort of like conceptual synesthesia, they contend, we understand one domain by analogy to another. A conversation, e.g., we think of as a journey, with a beginning and a destination—some place you’re headed—and impasses. Sickness is war, and so we win or lose the battle with cancer. When talking about communication itself, at a meta level, we use the so-called conduit metaphor. We place ideas into containers like sentences. We try to get across ideas. Words are containers of thoughts: paragraphs full of ideas, pregnant with meaning. These examples all sound like things we do with words that already exist, but I contend it’s the realization that etymologies uncover conceptual metaphors buried in words’ morphologies that gives us that sense of serendipity I mentioned earlier—it’s metaphors all the way down.
Sometimes that feeling accompanies particular sounds. The idea of a phoneme (which can be combined with others to make meanings) doesn’t work here. These sounds don’t signify; they’re more like iconic embodiments of a concept. Think of [gl], which suggests some sort of luminescence in many Germanic languages: glimmer, glitter, glow, gleam, glare; or [sl], which can suggest a lack of friction (slick, slide), though not always (slow). These are called phonaesthemes, and they’re basically Plato’s “appropriate sounds” from the Cratylus.
Back in 2017, the writer Sam Sanders posted a wonderful piece on the Awl, “Dillweed (As An Insult),”19 in which she contemplates the herb dill and the lesser known dillweed, but goes on long tangents about what she is really interested in: Why would dillweed be an insult?
This sparked etymologist Ben Zimmer's interest, and he took to Twitter to carry out a lengthy diachronic excursus of dillweed and its insult cousins. You’ve got your dillwad, dilly, dillhole, dipstick, dipshit, dipwad, dimrod, dickhole, dickhead, etc.
A naughty but wonky debate ensued (this was when Twitter was good) about what connected those words. But the answer was easy: [dɪ] is a phonaestheme. First of all, the phoneme [d] by itself can sometimes phonaesthetically imply stupidity: dumb, dunce, duh. [dɪ], on the other hand, seems to imply stupidity, testosterone, and penises, all at once.
I have a theory—basically unfalsifiable, so you can take it or leave it—that we sometimes recognize phonaesthemes that started off in PIE and still carry the same power to this day. *Skei- is certainly a root and was inflected to make words in PIE, but I think the association we have with cutting and splitting when we hear [sch], [sc], and [sk] today aren’t institutional memories of a PIE word. That sound is simply a phonaestheme for us—and maybe for those Dills building mounds in the Pontic Steppe 5000 years ago, too.
The Dutch linguist Yoïn van Spijk often posts cool etymological infographics from his Tumbler on Twitter, usually showing how words that you just assume have to be related aren’t. Here’s one disproving any relationship between the latin word fluere (flow) and the English word flow. 🤯
The user NeuroPolarbear asked the same question that was on my mind: [D]o we think that PIE *plew and *bhlewH are not themselves related? Van Spijk maintained that there were no attestations of *p-*bh correspondence among PIE words. Nor were there correspondences between other pairs of voiceless stops and voiced aspirated stops, which would indicate that some sound law was at play.
Things took a more specialist turn when the renowned linguistic historian of Arabic and Berber Marijn van Putten said he wasn’t so sure of that. He noted that he’d indeed found such correlations. eg.:
karp- ~ gʰrebʰ- "to grab"
ḱerd ~ gʰrd- "heart"
His explanation was that this is an example of consonant gradation, a form of consonant mutation that occur almost exclusively among words in the Uralic family of languages (e.g. Finish, Hungarian, the Sami languages, and Samoyedic languages like Naganasan).
Depending on your linguistic knowledge, that might seem weird to you, since you know that Uralic is a completely different family from Indo-European, with its own equivalent of PIE, Proto-Uralic. But Marijn van Putten is a student of the venerable historical linguist Frederik Kortlandt, who contends that before PIE or PU there existed a proto-language supergroup (like the Traveling Wilburys of the Proto-languages)—Proto-Uralic.
I was curious about van Putten’s contention that his examples were incidents of consonant gradation, since I understood that phenomenon to only happen in the middle of words (word medial situations). As far as I understood it, the gradation happened because of the stress placed on the syllable that came before it, while van Putten’s examples were at the beginnings and endings of words.
He said he was working under the precedence of Kortlandt, who “cites *tekm *dʰgʰem- 'earth' as an example[,]” and “seems to expect consonant gradation to happen in pre-stress syllables.”
Then he extrapolated a more etiological clarification from Kortlandt’s explanation, noting that this possibly indicates a sound law at work in consonants in proterokinetic versus hysterokinetic nouns.
That’s fascinating and maybe true, but beyond my pay grade! But, respectfully, sound laws, schmound schmaws, I say! I don’t think you need to show a phylogentic relationship. Flow and fluere are related in a different way: [fl] is a phonaestheme associated with moving liquid, just as *pl and *bhl were waaaaaaay back in the day. Or maybe there’s no there there. That’s fine, too. I can go with the flow.
p. 41–5
More on that later.
p. 31
Meaning “provincial,” but you can clearly see the vector by which that word comes down to us English-speakers in our current sense.
See note 4. The scare quotes are for your sake, not Cicero’s. He meant “provincial.” Of course, he also bears a good share of the blame for it meaning something so very different now.
Compare that to “elect,” and suddenly the association of reading with choosing becomes less opaque.
Think of “ligament,” but also “rely” (since, in the change from Latin to Old French—whose relier is the source of our word—[g] was either dropped after certain vowels or changed to [y], like here.
I repeat: More on that later.
The “Dream of Scipio” (Somnium Scipionis) is most of what still remains of part 6 of De re publica, which was Cicero’s adaptation of Plato’s Republic. Cicero’s version is a defense of the form of government that had held fast in Rome for centuries before certain people (*cough* Julius Ceasar *cough*) threatened its survival. Appropriately enough for our topic, the only reason we call Plato’s most renowned dialogue the Republic (in Greek it’s Πολιτεία, or Politeía, literally meaning “the consitution” or “the way things are set up”) is because we’ve historically adopted Cicero’s term. Ironically, when it comes to Cicero’s own book, we’ve decided that the associations with the word “republic” in modern English are so misleading that we must instead call it “De re publica”: “The public matter.”
As for the “Somnium,” that tells the story of a dream vision had by Scipio Aemilianus (who serves as the Socrates for Cicero’s dialogue), in which he is astrally projected to a higher plane and guided by the spirit of his grandfather-in-law, Scipio Africanus, to look down on the earth and see it for the simple and meager thing it is. The only reason we know of the “Somnium” at all is that much of Cicero’s text was transcribed along with a commentary on it by the Neo-Platonist writer Macrobius. That commentary was massively influential on medieval culture, and its impact is still felt today. It is the source of the Ptolemeic model of the universe that dominated the cosmology of the middle ages. And everyone from Chaucer to Dante to George Lucas is indebted to Cicero (by way of Macrobius) for the “dream vision” trope of the “Somnium”:
More on that in Part 3!
p. 36. Fussell’s book is a little too assertive for my current tastes, but it’s how I learned to think confidently about poetic meter as an undergrad.
By the time the Dictionary was in galleys, Johnson wasn’t so happy with the good Earl—who apparently left him hanging for most of the writing process and wanted to take credit right when it was about to go public. As the curators at the of the British Library note, this was a turning point in the publishing industry. Johsnon pointed the way for writers to profit independently from their publications. The bitter letter Johnson wrote to Chesterfield reads like the “About” page from a present-day writer leaving behind a traditional columnist job to start a Substack.
You know the one.
Paragraphs 72 and 74
And he’s a role model of mine, as a fellow wordy American with one foot always in Germany
p. 216
p. 5
From that piece, I discovered that, “According to Moscow-based correspondent Michele A. Berdy of The Moscow Times, Ukrainians are often just referred to in Russian as ‘Dills,’ and the nation ‘Dill-Land.’” We should send more tanks.