By Anatoly Liberman
This is a story of the names of two letters. Appreciate the fact that I did not call it “A Tale of Two Letters.” No other phrase has been pawed over to such an extent as the title of Dickens’s novel.
The letter H.
The name aitch stands out. We observe the usual vocalization of letter names in Engl. B (bee), K (kay), S (es), German H (ha; at the end of the tenth century, ha was still the name of H in Latin, and in the twelfth century an Icelandic grammarian called H the same), and Italian F (effe). But aitch? What does H, which designates aspiration, have to do with ch, a sound belonging with affricates (to use a special term), that is, with kh, ts, dz, j (as in jay), and pf? And why ai-? Strangely, the answer is not quite clear, though the name of any letter can be expected to have a transparent origin.
The older explanation presupposes that at one time Latin had the form accha or ahha, with aitch being its descendant. But as long ago as 1892 a more convincing etymology was proposed by the American linguist Esther S. Sheldon. It was developed and improved by Horst Weinstock, a contemporary German historian of the English language and an expert in old scripts.
The reconstructed name ahha looks like ha with a and ha reversed, h doubled for an unexplained reason, and another a added for vocalization. Yet, as we have seen, long after the collapse of the Roman Empire the name was still ha in the Germanic part of medieval Europe. If the reconstructed ahha arouses suspicion, accha inspires even less confidence, for where did c come from? Yet its legitimacy cannot be doubted, for Italian acca, Spanish ache, and English aitch presuppose the sound of k as its source (English ch in old words always goes back to k: so church, child, each, birch).
An alphabet is a sequence of letters, and rules exist in many cultures to help learners memorize it. Songs like A, B, C, D, E,—you are in the jungle in a coconut tree or X, Y, Z, now we know the alphabet are well-known. It appears likely that medieval Romance learners, while memorizing the part of the alphabet beginning with H, skipped the vowel I (J did not exist at all), so that H and K found themselves in close proximity. Their closeness produced the sound complex HAKA. Since initial h was lost nearly everywhere in the Romance speaking world, haka became aka, of which Italian acca is a natural continuation. It should also be borne in mind that under no circumstances could ha in its capacity as a letter name survive in Romance, for, once h- disappeared, ha (H), merged with the name of the vowel A. The involvement of K in the merger caused minimal trouble, because in the Romance languages K had very low frequency. All the work was done by C; since early times, K appeared only in words of Greek and Germanic origin. Paleographers (students of ancient writing) discovered that alphabets in many Roman manuscripts lack the letter K. (In so far as I never miss the chance of lambasting English spelling, I might as well give a fling to it here too. If we really need both C and K in English, we could at least abolish the difference between skate and scathe and agree that sk would do well for all cases, even—horribile dictu—for skool, that is school.)
Perhaps especially telling is the Portuguese name agá, with its stress on the second syllable. We can assume that in reciting or chanting the alphabet stress was not fixed and vacillated according to the requirements of intonation and rhythm, as it sometimes does in English: compare Ténnessee Wílliams but the státe of Tennessée, fífteen books but Room Fiftéen, goodbye but The Góodbye Girl. Agá is a similar variant of ácca. Nor is the case of Portuguese isolated: accá turned up in Langobardian and Franconian Romania. In Anglo-Norman, the letter H was called hace, which should have yielded Engl. haitch; compare the noun ache: it was pronounced aitch for a long time. The loss of initial h testifies to the influence of Central French. Few etymologies are perfect. The etymology presented above rests on the suggestion that at one time speakers recited the alphabet, with K immediately following H. This reconstruction may not satisfy everybody, but at the moment it is the best we have.
The letter Y.
Some time ago, I wrote a post about the uselessness of the letter Y in English. One of our correspondents asked me about the origin of that letter name. His query suggested the entire idea of my “Alphabet Soup,” which shows that the more questions and comments I receive the better. It turned out that the etymology of wy (the letter name) is even more obscure than that of aitch (h). With very few exceptions (the OED being one of them) dictionaries do not touch on the enigmatic name, and the OED admits that the riddle has not been solved. My etymological database contains only one relevant citation. On 25 October 1895, Skeat gave a talk to the Cambridge Philological Society titled: “On the Origin of the Name of the Letter y and the Spelling of the Verbs ‘build’ and ‘bruise’.” This is what he said. In Anglo-French, the letter was called wi. This is known from a manuscript written about 1210. For rendering Old Engl. y, whose value, when the vowel was long, must have been that of Modern German long ü, as a rule, the scribe used the letter u, but in at least seven instances he substituted the digraph ui for u (herein lies the connection between the main subject of Skeat’s talk and the verbs bruise and build).
I think it will be more profitable if I quote rather than retell Skeat:
“Since in those days the vowel u was not pronounced (as now) like the ew in ‘few’, but like oo in ‘cool’, it follows that the symbol ui must have been called oo-i, or in rapid speech wi [with long i] (formerly sounded as we, not as wy). That is, the name wy denoted ui, a symbol used in Southern English of the thirteenth century to represent the sound of Old English y. If we reverse ui, and write iu, which (pronounced quickly) gives the sound of the ew in ‘few’, we get the present name of the letter U; and it is well known that the modern sound of u in ‘cure’ arose from the Old French u, which was pronounced very like the Anglo-Saxon y. That is, u-i (= wy) gives the name of the Old English y, and i-u (= eu) gives the name of the Old French u sound which resembled it. It follows that the true U, as heard in ‘ruby’ has no name at all in modern English. It ought to be called oo. This result is fairly proved by the fact that two verbs with the spelling ui (for A[nglo]-S[axon] y) still survive in Modern English. These are ‘build’ from A.S. byldan and bruise from A.S. brysan [with long y]. These spellings are the more interesting from the fact that they have never been either understood or explained till now.”
The best etymologies are usually simple, while Skeat’s explanation is rather convoluted, but, obviously, wy could develop only from wi, with long i, that is, from a form sounding like modern wee, and ui is a good candidate for its etymon.
In the picture, you can see the Old Icelandic runic alphabet, the so-called futhark (f-u-th-a-r-k are the first letters of the sequence). The letter for ha (for the purposes of memorization it was called hagl “hail,” noun) is the first in the second row.
Anatoly Liberman is the author of Word Origins…And How We Know Them as well as An Analytic Dictionary of English Etymology: An Introduction. His column on word origins, The Oxford Etymologist, appears on the OUPblog each Wednesday. Send your etymology question to him care of [email protected]; he’ll do his best to avoid responding with “origin unknown.”
Subscribe to Anatoly Liberman’s weekly etymology posts via email or RSS.
Subscribe to the OUPblog via email or RSS.
Image credit: The Elder Futhark table (screencap) via Wikipedia.
It appears likely that medieval Romance learners, while memorizing the part of the alphabet beginning with H, skipped the vowel I (J did not exist at all), so that H and K found themselves in close proximity.
Is there any independent evidence for this omission of I? It sounds implausible.
You may have already dealt with this, in which case I apologise. I have always wondered why the letters of the alphabet are in that particular order – ABCD etc. The same goes for the runic alphabet – futhark. Who decided on this order? Has it to do with the use of letters as numbers in the Greek / Hebrew alphabets? But if so, wouldn’t that apply only to the first ten letters? Thank you for any light you can throw on this.
Having lived in the Netherlands for some time, I have often wondered if our ‘building’ and ‘bruise’ came from the Dutch – obviously not. The Dutch pronunciation of ‘ui’ can only be learned by hearing it!
@Catherine Sommer: to the best of my knowledge there is no known rationale for either ordering. That of the Roman alphabet (ABCD) derives, via Old Italic and Greek, from the order of the Phoenician alphabet, which is also reflected in many of its other descendant scripts such as Cyrillic, Hebrew, Arabic, Georgian and Armenian).
The Futharc runic alphabet was also derived from Old Italic, so it’s a sister of the Roman alphabet. However, it changed the inherited alphabetical order for reasons that are again unknown, AFAIK.
I used to wonder why the usage of y was so messy in English (system, cycle, fly, gently). I now know that it’s invariably pointless to ask logical questions about English spelling: http://englishspellingproblems.blogspot.co.uk/2013/07/regular-and-unpredictable-spellings-in.html