By Anatoly Liberman
The world has solved its gravest problems, but a few minor ones have remained. Judging by the Internet, the spelling of whoa is among them. Some people clamor for woah, which is a perversion of whoa and hence “cool”; only bores, it appears, don’t understand it. I understand the rebels but wonder. Woah looks like a garbled Hebrew name, though, on the other hand, why couldn’t there be a maiden called Woah? Quite possibly, she even existed and only escaped the annalists’ attention because she spent her childhood and youth in a cloister known as W.O.A.H. (“World’s Oldest Association for Happiness”), a precursor of YMCA. When she grew up, she found a spouse and became (“morphed into”) Whoa, because, wherever she went, she aroused great surprise. Stranger things happen in married life. I am neither a biblical scholar nor a marriage counselor; my concern is solely about the digraph oa in English. How did this combination of letters arise and why does it occur in so many words?
I will pass by cocoa and boa: both are too exotic to cause trouble. But why is the past of ride spelled rode, while the path used for riding is called road? Let us begin at the beginning. Middle English had two long o’s. One resembled aw in Modern Engl. law, paw, saw, the way those words sound in the varieties of English in which Shah and Shaw are distinguished. It went back to long a (approximately as in Modern Engl. spa) and was an open vowel, that is, during its articulation the lower jaw dropped quite a bit. Another long o continued Old English long o and was less open. In some dialects, a third o appeared, the product of short o lengthened before single consonants. It got in between the two older vowels. The space became too crowded, and multiple mergers took place. As a rule, the third o, in the few areas in which it arose, merged with its close neighbor. Still later all o’s changed under the pressure of the Great Vowel Shift.
While the two o’s coexisted, they were not confused in speech, and scribes needed different symbols for them. They introduced the digraph oa for open o: the letter a could signify the openness of o because a is the most open vowel in any language. The symbol for long close o remained the letter o. The device looks awkward, but its use can be justified. The digraph was not an invention of the English Middle period: oa had sufficient currency in Old French, and the scribes (of French descent or educated in the French tradition) followed the only model they knew. But there may have been another reason. While long a was becoming long open o, the spelling oa reflected the ongoing change or the scribes’ uncertainty (a? o?) quite well.
Here are some examples. The digraph oa in boat, goad, load, loan, oak, soap, and toad, among many others, designates the vowel that once was open long o (from long a). Road also belongs here. Old close long o became long u (pronounced as in Modern Engl. pooh–pooh); hence boot, doom, school, and so forth. These words will no longer interest us. As noted, short o, lengthened in Middle English, tended to merge with its close partner, and orthography did not react to lengthening: words with it were spelled with o in both Old and Middle English, changed under the influence of the Great Vowel Shift, but retained their visual image through the centuries (as, for instance, in spoke)—let us add: sometimes! After open and close long o merged, people could not know which is which, and spelling became chaotic.
Take the word mole “spot.” In Old English it had long a. This means that today we could expect moal. No such English word exists, though the omniscient Internet reminds us of Moalboal, a fourth class municipality in the province of Cebu, Philippines. Mole “animal name” had a short vowel in Old English; consequently, the spelling of its present day reflex (continuation) is regular. Opponents of Spelling Reform keep repeating that, though our spelling is hard to learn, at any rate, it preserves the venerable past of the English language. This argument makes no sense even in general terms (why should the spelling of a modern language be a faithful transcript of the past?), but, to make matters worse, spelling distorts the history of English as often as it reflects it. The case of road is characteristic. This noun and the past tense of the verb ride go back to the same base and were homonyms in Old English, namely rad (with long a). It follows that both should be spelled road. Either under the influence of forms like spoke or for some other reason, rode acquired a wrong shape. Home and bone are also “wrong”: their original forms (ham and ban, with long a) predict modern hoam and boan, like foam and groan. By contrast, soak should be spelled soke, but it is not. To restore fairness, English has a legal term soke “right of local jurisdiction,” which is spelled “correctly.”
The plot thickens when we encounter late borrowings, most often from French. Consider coat and coast. What would have happened if they were spelled coste and cote? Coste, with –e after two consonants, does not look ugly (compare haste, paste, and waste versus waist); cote (as in dovecote and sheepcote) is also fine. Moat had o lengthened before a single consonant and therefore should have been spelled mote, like mote “a speck of dust.” Here opponents of Spelling Reform tell us that homophones should be distinguished in writing. Really? Are they seriously inconvenienced by fan (for winnowing) ~ fan “admirer,” poach (as in poached eggs) ~ poach “trespass,” and dozens of others?
Some dictionaries list dote and doat and explain that dote “to be silly or weak-minded” can be distinguished from doat “to bestow excessive fondness.” Perhaps it can, but I am not sure that someone old but not yet in his (or her or better “their”) dotage will nowadays write a doating parent. The most puzzling word with oa is broad. In Old English, brad rhymed with rad (both had long a). Then, quite regularly, both changed their long a to long open o. Road went further and farther (roads usually do) and acquired a diphthong by the Great Vowel Shift, while broad stayed with its long undiphthongized o. At least two conjectures have been offered to explain this anomaly. According to one, broad, meaning what it does, was pronounced with an intonation that preserved long o intact. This is possible but rather hard to believe. The other explanation has it that broad is a northern form; in the North, vowels stayed away from many changes familiar to those who speak Standard English. But no solid evidence testifies to broad having been particularly favored in the North. Finally, I am pleased to report that there is absolutely no need to spell hoarse and hoard with oa, even though the vowel before r was open. Compare horse and porcelain. Yes, and I have almost forgotten gloat, a probable remnant of eighteenth-century slang. Wouldn’t glote satisfy us?
Read and weep, and when you have dried your tears, look up the origin of school and shoal “large number of fish” in some good dictionary.
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 articles via email or RSS.
Subscribe to the OUPblog via email or RSS.
Subscribe to only language articles on the OUPblog via email or RSS.
Image credit: Chile – twisting switchbacks at the border crossing. Photo by McKay Savage, 26 February 2012. CC-BY-2.0 via Wikimedia Commons.
But horse and hoarse were historically distinguished, and remain distinct in some varieties (Irish, Scottish., some parts if the USA). Google “horse hoarse merger”.
I think there is no need for a special explanation of broad. Sociolinguistic study shows that sound changes don’t happen all at once, but word by word by word, and some words resist to the end. Nothing else can explain why we say you (stressed) instead of yow (to rhyme with cow), and so many in Ireland and elsewhere say me for my. (I do not believe the OED’s idea that Old Northumbrian iw flew to London on the back of the south wind. In Southern Scots, however, you and me are regularly yow and mey.) Another good example is Hawaiian tūtū ‘Grandma’, despite general Hawaiian [t] > [k] (President Obama called his grandmother Toot.)
Don’t be too hasty to merge horse and hoarse in writing. The NORTH-FORCE merger in speech didn’t happen until the 1970s in New England and the American South, and has never happened in Scotland, Ireland, Eastern New England, most of the Caribbean, and India, nor among AAVE speakers. That’s a lot of anglophones. However, horde belongs to FORCE and would be better off as hoarde; for that matter, pork would be better as poark to distinguish its vowel from that of fork.
I admire your efforts to find logical explanations for English spelling conundrums, but since even the spellings of highly educated people like Sir Thomas More and Elizabeth I were highly inconsistent, i suspect they are mostly random.
It’s terrible that they continue to be tolerated, despite plenty of evidence that they give children a hard time and leave too many adults functionally illiterate:
http://improvingenglishspelling.blogspot.co.uk/2013/08/why-english-spelling-needs-modernising.html
So, is this why read is read read and read is read read? (“reed”, “red” “reed”, “red” “red” “red”)?
Yes, Andy. How can anyone not be more than ready to redd out the whole mess?
This is a delightful piece. I’m sorry to be so late coming to it, but most of the points I would have made have been more than adequately dealt with in these comments.
But when you say “I am pleased to report that there is absolutely no need to spell hoarse and hoard with oa, even though the vowel before r was open”, it almost sounds as though you are now declaring the horrid hoary NORTH-FORCE question closed in favour of the merger. That would indeed be premature, and here I’m surprised at John Cowan’s uncharacteristically circumspect use of the word “hasty” for it! What are the hapless unmerged to expect next? An etymological argument for unifying the spelling of “four” and “forty”? A reformed spelling must surely respect the perfectly respectable oppositions that do survive. That of course is what makes these projects so complicated.
And for the unmerged, these ‘oar’ spellings exhibit one of the few consistencies there are. The irony is that the poor scribes who introduced the digraph oa for open o could not have foreseen that in this context it represents the closer of the distinctive realizations.
In your example “whoa”, it may well have been intended to represent an open vowel as opposed to “woe” at some point, but I doubt if anyone could adduce any opposition now, unless that were the ubiquitous spelling pronunciation rhyming with “boa”. What fascinates me about this example is that the form “woah” is obviously a respelling of the reinterpretation of the original respelling “whoa” in order to represent this spelling pronunciation!
I have no evidence for this, but i suspect that one of the worst dilutions of English spelling consistency was implemented quite deliberately.
When after the end of the 100 years war with France in the 15th century, Chancery clerks were obliged to switch from the previously superior French to common English for official documents, they must have been very annoyed. Might this be why they decided to make English spelling less penetrable? Were they trying to preserve their higher status?
Why else would they have wrecked the earlier much more regular spellings of Chaucer, such as ‘lern, erly, frend; beleve, seke, speke’, and mainly by using one grapheme for different sounds (e.g. heal, head, heave, heaven, treat, threat)?
‘so many in Ireland and elsewhere say me for my‘
In Ireland, if not elsewhere, me for my (and be for by) are only used for the weak form; the strong or citation form retains the diphthong.
[…] The digraph oa and a hoarse horse […]
Mallamb: It took me a while to decipher that “horrid hoary”, since for me “horrid” is not NORTH=FORCE at all, but LOT + /r/, i.e. /hɑrəd/.