By Anatoly Liberman
Confusing as English spelling may be, it has one well-publicized, even if questionable, merit: it tells us something about the history of the language. For example, sea and see were indeed pronounced differently in the past. This fact is of no importance to a modern speaker of English but can be put to use in a course “Spelling as Archeology.” In other cases, modern spelling only puzzles and irritates. For example, most of my undergraduate students believe that the preterit of lead is lead (like read ~ read), though they never misspell bled and fled. We are heirs not only to the pronunciations of long ago but also to the absurdities of what may be called learned tradition. The great lexicographer Samuel Johnson (his dictionary was published in 1755) preferred the spelling skeptic, and this is the standard variant in American English, while in England sceptic prevailed; great reverence for Johnson did not help. When we compare sceptic and ascetic, we cannot help wondering why sc is pronounced as sk only in the first word or rather why those two words are spelled alike if they are pronounced differently. Even English speakers long for some consistency when they write. Years of apprenticeship should have weaned them from this craving, but humans are incorrigible rational beings.
English words of Scandinavian origin begin with sk- rather than sh-. The word Scandinavia never ceases to remind us of this. Therefore, we are not surprised to learn that skin is a noun of northern descent, whereas shin is native. Skirt and shirt are doublets: they go back to the same etymon (source), but only shirt is “truly” English. Since sk- and sc- are used almost interchangeably in modern words, the picture comes out blurred. Besides this, the sk- ~ sh- rule is far from being universal. Scale “weighing instrument” is indeed a northern word, but scale (in playing scales), scale “lamina,” and scale “climb” are from Romance. Skeleton is ultimately from Greek; skates is from Dutch. In the north of England, the Scandinavian influence was more common than in the south, for the Vikings held sway there. As a result, some words may be English but pronounced with sk-; skedaddle seems to be one of them. As pointed out, skirt and shirt are doublets; in contrast, scatter and shatter are not. Neither are skit and… (I probably have no need to supply its fictitious English partner). Such facts are curious but do not inconvenience spellers. Trouble begins when etymology gets the better of common sense, and this is what I called a learned tradition.
Few words are more offensive to the eye than scythe. It has an ancient root discernible in the borrowed verb dissect (sec- means “cut”), and the early Old English for scythe was sigdi. The g at the end of sig merged with the preceding vowel because it developed into a sound approximately like the one designated by y in Modern Engl. yes. This was a regular development in some positions: compare Engl. say and German sagen. The word ended up with a pronunciation resembling today’s seethe and eventually yielded sithe. But this was not good enough for Renaissance scholars. They associated scythe with scissors. Now, scissors goes back to Middle English sisoures, from Old French cisoires. Its root (again “cut”) has been preserved, among others in Engl. (de)cide, from Latin (de)cidere “cut off, cut the knot; determine.” Note that c (or k) does not appear after s in any of the old forms discussed so far. 16-th century philologists believed, wrongly as it turned out, that sisoures traced to Latin scindere “cut, cleave,” which English has in rescind and the bookish word scission (both, as well as decide, naturally, borrowed). Thus, we are dealing with three roots, all of which mean more or less the same and which can be found in dissect, decidere, and scindere.
It would be unreasonable to accuse the linguists who lived long ago of mistakes they made while trying to explain the origin of English words. They did not have the benefits of our etymological methods. One could only wish that they would have shown some restraint. If they had just left their language alone! But they did not. They wanted words to have an appearance worthy of their imagined ancestry, so that they began with scindere, inserted c in what is now scissors, and in their enthusiasm added c to sithe. But scithe needed a finishing touch to make it look really dignified. This is how scythe arose, along with Smythe, Wylde, and the rest. Incidentally, scissors is disfigured in two places, for -ss- is not fit to designate the pronunciation z. Yet there are other words of this type: dissolve, possess, hussar, dessert, hussy, and the almost forgotten hussif “case of sewing necessaries”; they also have the misleading letter group -ss- in the middle (hussy and hussif are occasionally spelled huzzy and huzzif; both go back to a word for “housewife”).
The history of scythe is stuff for a tragedy in five acts: sec- becomes seg- in Germanic and finds what it hopes to be a safe haven in segdi, which yields sigdi; g is weakened and lost, and, as result, i is lengthened (it has early premonitions that it will one day become a diphthong, but since this is the fate of every long i, it takes the impending disaster in stride); etymological fanatics get wind of something called the cutting edge, connect sisoures with scindere, and produce scissors with ss in the middle, apparently to blunt the effect of the misdeed; it occurs to them that sithe belongs with scissors, and they insert c into it; after so many operations the tool becomes useless, and in the vain hope of saving the situation the villains replace the vowel letter i with y. Now no one can recognize the initial product. Act drop. The tragedy has an epilogue, a farce needed to produce an anticlimax rather than catharsis. In the 18th century, a strange pronunciation of sigh spread to London: people began to say sithe. If that fashion had stayed, scythe would have acquired a full-grown mourner.
Numerous English words that are spelled with sc- before i and e were never pronounced with sk. Such are ascend, descend, ascetic, science, rescind, and scintillate, to mention a few. They were sacrificed on the altar of etymology. Even scent, whose origin could not be in doubt (Latin sentire, Old French sentir), got sc- in the 17th century. Scion is obscure. Schedule, schism, and scheme were the subject matter of a special post, and here I will not dwell on the uses and abuses of sch-. Sceptic had a good chance to end up among ascend and ascetic (the letter c in French sceptique is mute), but it conformed not only to the Greek spelling but also to the Greek pronunciation of its etymon. Hence skeptic, at least in some parts of the English speaking world. Such are the trials and tribulations of skeptics and ascetics. The wheels of English spelling spare no one.
Anatoly Liberman is the author of Word Origins…And How We Know Them. His column on word origins,
The Oxford Etymologist, appears here each Wednesday. Send your etymology question to firstname.lastname@example.org; he’ll do his best to avoid responding with “origin unknown.”