By Anatoly Liberman
Why don’t good and hood rhyme with food and mood? Why are friend and fiend spelled alike but pronounced differently? There is a better way of asking this question, because the reason for such oddities is always the same: English retains the spelling that made sense centuries ago. At one time, the graphic forms we learn one by one made sense. Later the pronunciation changed, while the spelling remained the same. Therefore, the right question is: What has happened to the pronunciation of the words that give us trouble?
Everybody has heard that vowels can be short or long, but those who have not studied Greek or Latin should beware of such terms. According to the formulation English speakers learn pat, pet, pit, and pot have short vowels, as opposed to pate, Pete, spite, and spoke. However, those are conventional terms inapplicable to today’s pronunciation. Some idea of what length or at least duration means can be gained from words like father and spa. They indeed have “long a.” Flaw and saw have “long o” in the pronunciation of the speakers who distinguish between Shah and Shaw; too has “long u.”
In ancient Indo-European (which includes Germanic), a long vowel functioned as a combination of two short ones. Then, not earlier than the thirteenth century but hardly later than the fifteenth, the situation in English changed, and vowels became indivisible, with their duration depending on how the postvocalic consonant adhered to them. If the contact was close, the vowel was pronounced short, but if it was loose or if there was no consonant following a vowel, length prevailed. Although untrained speakers of British English rarely associate the difference between bid and bead, cod and cord, let alone bud and bard, with vowel duration, the difference is significant. Speakers of American English have still more trouble here. Even Germans, in whose language the situation is roughly the same as in English, complain that the recent spelling reform disturbs them when they have to decide whether the previous vowel is short or long.
Linguistic revolutions occupy a few pages in elementary textbooks. In real life, they entail vacillation, acceptance of a new norm, returning to the conservative variant, and multiple inconsistencies. This is what happened in English. The choice of the vowel-consonant contact was not regulated by strict rules, though some tendencies can be detected even after so many centuries have passed. In Old English, good, hood, and mood had the same vowel (approximately as in Modern Engl. saw) and formed a perfect rhyme. That state of undisturbed harmony was ruined by the introduction of the new regime: every word had to choose its type of contact.
Apparently, the choice was often arbitrary. Good and hood acquired close contact, and the vowel shortened, while mood ended up with a loose contact and, consequently, a long vowel. Some words develop fast, others are slowcoaches. When the so-called Great Vowel Shift struck (this happened in early Modern English), the original vowel of good, hood, and mood (to repeat: it resembled aw in Modern saw) acquired the value of Modern English too. As is well known, in the north of England both words in the phrase shut up are pronounced with the vowel of put, that is, shoot oop, unlike what happens in the south. But Standard English also lacks consistency: put and cut do not rhyme; bull and bulb have different vowels. It seems that the vowel of blood changed from what we today hear as aw in saw to “long u” as in Modern woo so early that it became short in time for changing to the vowel of Modern cut. Therefore blood (similarly, flood) now rhymes with bud. The spelling of good, mood, and blood takes no cognizance of the changes that have happened in those words. On paper they still occupy the same chamber as they did at the court of King Alfred.
Scholars have been describing English pronunciation since the seventeenth century, and we have detailed evidence of the great vacillations in the way all such words were pronounced in the not too distant past. It always comes as a surprise that some variants, vastly different from those we now hear, existed not a thousand years ago, but in the days of Fielding, Dickens, and even later. In Middle English, friend rhymed with fiend. Yet, unexpectedly, close contact set in in it before nd, a consonantal group that at an earlier epoch caused lengthening (that is why bind and bound rhyme with lined and crowned: without lengthening they would have rhymed with dinned and the first syllable of Tundra). Close contact meant shortening. The two words are still spelled alike, though only fiend has retained loose contact.
With time, the visual image of the most familiar words becomes so familiar that we stop noticing the incongruities. Said, plaid, says, done, gone, one, and dozens of their linguistic cousins are hieroglyphs to be learned individually. Our reverence for the past will probably never allow us to switch to sed, sez, plad, or, God forbid, nun (the latter for none). What a devilish plan to separate friend and fiend, even though living speech has separated them quite successfully! Frendship will certainly be less precious without a superfluous letter.
By way of conclusion, here is a curious example of interplay between shortening in a dissyllabic word and folk etymology. Once the word Christscrosse existed. It meant “figure of a cross.” In compounds and long words the first vowel was regularly shortened, which explains the difference in the modern pronunciation of holy and holiday, south and southern, white and Whitsunday, Christ and Christmas. The first vowel in Christcrosse developed as in Christmas. That word also denoted “alphabet,” because the figure of the cross was used in front of the alphabet in hornbooks and primers (early primers consisted of a single page protected by a sheet of transparent horn). The word lost its religious significance to such an extent that the pronunciation and spelling crisscross obtained and the compound was understood as a reduplication like trictrac, flip-flop, and so forth; hence the change of sense.
Those who were impressed by the story told here may ask themselves how they pronounce the word primer. Does it rhyme with dimer or dimmer in their speech? Cogitation and an attempt to square accounts with their phonetic conscience will allow them to spend a pleasant week in anticipation of the next post (which will have nothing to do with spelling).
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 here, 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.
View more about this book on the 
Flaw and saw have “long o” in the pronunciation of the speakers who distinguish between Shah and Shaw;
“Long O” usually describes the vowel of “GOAT”, which is quite different from that of “flaw” and “saw” in every accent I’ve ever heard or seen described.
Put and cut do not rhyme because an initial labial protected put and bull from the effects of the FOOT-STRUT split; this is a systematic rather than a sporadic change. In Scotland, this protection was ineffective, which is why we now have both put and putt in English, the latter being originally the Scots cognate of the former.
I do not know why dictionaries don’t seem to report my own pronunciation of bulb, though: I make it FOOT, as the above rule would suggest, and listening to m-w.com’s recording, I hear FOOT in that as well, though they list the pronunciation as STRUT. In any case, there is no phonemic contrast here, so “ayther will do” (as the Irishman said when asked whether to say either with FLEECE or PRICE).
But as for frendship, it is sacred however you spell it (within reezon). A rose by any other name would, in fact, smell as sweet.
Reverence for the past is indeed likely to prevent the ridiculous letters in words like ‘said, plaid, done’ ever being normalised. This is mainly because most people either don’t know or don’t care that they are the reason why learning to read and write English takes an exceptionally long time and entails high levels of very costly functional illiteracy.
Most short vowels are spelt as in ‘sat, set, sit, cot, cut’. Unfortunately, they have irregular spellings in some of the most frequently used words. The number of words with different spellings for short a, e, i, o and u is relatively small (3, 67, 53, 33 and 68) compared to their regular use in at least 466, 301, 421, 375 and 308 words respectively. But because they occur is some of the most often used words, they act like constant spanners in the basic workings of the English spelling system and impede children’s progress with learning to read and write disproportionately.
Merely improving the spellings of ‘friend, said, says, head, any, many; one, once, other, another, brother, mother, come, coming, some, money, sometimes, something, does, Monday, young’ and cutting the surplus from ‘have, give, live, are, gone, were’ would already make initial English literacy acquisition much less confusing than it currently is.
The most baffling aspect of English spelling, however, is consonant doubling. Young children are taught that their function is to keep short vowels in multi-syllable words short – to distinguish ‘dinner’ from ‘diner’. They learn this as the reason for doubling the final consonant of the likes of ‘fat’ and ‘fit’ when adding suffixes to them. They spend many hours on exercises designed to teach them when to double consonants (fatten, fatter, fitter, fitted, fitting), when not to (dined, rained, leaked) and when to drop vowels (dining, joking) before suffixes.
Unfortunately, the ‘open long’ (late, here, bite, wrote, cute) and ‘closed short’ (latter, herring, bitter, written, cutter) vowel spelling system for a, e, i, o and u is completely random when it comes to root words of more than one syllable (ballad salad, mellow melon, billion bilious, robber robin, muddy study). Only 381 common words use doubled consonants after short, stressed vowels, while at least 439 don’t, and 153 use them for unfathomable purposes (apply, approve).
Adopting regular, purposeful use of doubled letters would be one of the simplest ways to make learning to write English much easier and less time-consuming than it currently is. All this would require would be for regular, functional doubling (bannister, creddit, hiddeous, holliday, studdy) to be regarded as acceptable as its current inconsistent usage. Amending some of the irregular spelling for short vowels at the same time (reddy, wimmen, hunny) would help even more.
I would add to what Masha stated so accurately on the cost aspect of the matter.
There is a human cost to this, but there is an economic cost as well! While major economies around the Commonwealth are faltering, lots of money and resources continue to be pouring in literacy, simply because the spelling is so unreliable and irregular that kids and teachers must spend inordinate amount of time and money (more than other countries that have more “logical” languages) to mitigate the flaws. Canada’s success on international tests is in fact a very costly affair:
“Ontario’s debt drag
The province’s total debt (borrowing without taking into account assets) sits at $257.5 billion. Last year it was $236.6 billion.
Servicing that debt costs the province about $10 billion a year. Debt servicing is the province’s third-largest annual expense and one of its fastest growing expenditures.
The government spends more on servicing the debt each year than it does on colleges and universities.
Each one per cent rise in interest rates boosts the debt servicing cost by about $467 million.
A line from a finance ministry backgrounder puts the problem into perspective: “If no action is taken to balance the budget, Ontario would pay almost as much to service debt in 2017-18 as it spends on education today.” (http://www.cbc.ca/news/interactives/budgets-provinces/, 2012)”
Canada’s combined (federal and provincial) Debt-to-GDP was 57.9 per cent in 2010-2011
‘With time, the visual image of the most familiar words becomes so familiar that we stop noticing the incongruities. Said, plaid, says, done, gone, one, and dozens of their linguistic cousins are hieroglyphs to be learned individually. Our reverence for the past will probably never allow us to switch to sed, sez, plad, or, God forbid, nun (the latter for none). . . Frendship will certainly be less precious without a superfluous letter.’
But what if you are one of the millions who do not have reverence for the past, because they do not know it?
If it wer not for spellcheckers, we could cut out all the superfluous letters that do nothing for pronunciation or meaning and often only mislead – and all the millions in Anglophone countries and overseas would have more chance of lerning English.
2011, Yule, Valerie ‘Recent developments which affect spelling. On the possibility of removing the unnecessary difficulties in English spelling, while leaving the basic appearance of English print intact.’ English Today, 107, vol 27, No 3. Sept 2011, pp 62-67 http://journals.cambridge.org/repo_A839oLF6
[…] who pointed out that “long o” occurs in words like goat rather than law, is advised to reread my post. There I explain why the school definition of “long a, o, e, i” cannot be applied to language […]
[…] If you wanted to learn a little bit more about the evolution of the English spelling system and how it has evolved over the years, the Oxford University Press’ blog has a great post on it here. […]