Language is conventional, not logical. But try telling that to the Greeks.
Like the rest of us, they weren’t happy to think that words are made up of arbitrary combinations of sounds or letters. Surely there must be some logic to etymology, they thought, just as we do today. Even in the absence of historical evidence, or despite it, we think that just by exercising our reason and imagination we can figure out where words came from.
In Plato’s dialogue of Cratylus, for example, Socrates feels inspired to provide the logic behind the Greek words for “man,” “body,” and “soul,” among others. In the translation by Benjamin Jowett, here is Socrates’ etymology for “soma,” the soul:
That may be variously interpreted; and yet more variously if a little permutation is allowed. For some say that the body is the grave (“sema”) of the soul which may be thought to be buried in our present life; or again the index of the soul, because the soul gives indications to (“semainei”) the body; probably the Orphic poets were the inventors of the name, and they were under the impression that the soul is suffering the punishment of sin, and that the body is an enclosure or prison in which the soul is incarcerated, kept safe (“soma,” “sozetai”), as the name “ooma” implies, until the penalty is paid; according to this view, not even a letter of the word need be changed.
And “helios,” the Greek word for sun:
The origin of the sun will probably be clearer in the Doric form, for the Dorians call him “alios,” and this name is given to him because when he rises he gathers (“alizoi”) men together or because he is always rolling in his course (“aei eilein ion”) about the earth; or from “aiolein,” of which meaning is the same as “poikillein” (to variegate), because he variegates the productions of the earth.
It’s akin to Adam’s assignment in the Garden of Eden to name the animals God had created, “and whatsoever Adam called every living creature, that was the name thereof.” But Socrates invokes logic, imagining the “primeval givers of names” deducing names from the properties of things.
Logic like that can lead to the extreme of deriving words from their opposites. The most famous example is from an ancient Roman grammarian who decided that “lucus,” meaning a dark wood, must come from “luceo,” meaning to shine. In other words, it is called “lucus” because it isn’t. In honor of this absurdity, the phrase “lucus a non lucendo” (bright because not bright) has become the designation for an etymological contradiction.
But we really can’t blame the ancients for a practice that continues unabated today of imagining where a word comes from without bothering to look at its history. Linguists call this “folk etymology,” but often it’s cultivated folk who decide where a word logically must have come from.
A good modern example is America’s greatest word export, OK. A word as odd as that, and with worldwide success, invites imaginative explanations, and there are plenty of them.
It happens that, thanks to the meticulous scholarship of Allen Walker Read in the 20th century, we know with unusual certainty exactly when and where OK had its beginning. That was on March 23, 1839, on Page 2 of the Boston Morning Post, in a joking misspelling of “all correct.” From that starting point a thick trail leads to its uses and meanings today.
But that evidence hasn’t stopped savants from imagining a more dignified origin. Surely it came from the Choctaw Indian language, or a 2,000-year-old expression in Greek, or the initials of a biscuit maker; or from Latin, French, German, Finnish, Scots, or the Wolof language of Africa. You can read about them all—and why they’re all wrong—in my book about OK.
Not that anyone will be deterred from claiming otherwise. I suspect the anti-Bostonians (akin to the anti-Stratfordians who are sure Shakespeare didn’t write his plays) will assert themselves in the comments to this post.