Wikimedia
To put it mildly, the English language has a tricky spelling system. It’s confusing, illogical, archaic and often just bizarre. Case in point: in standard British English the letter o can, depending on context, represent nine out of the twelve monophthongal vowel sounds (by my count). We have a seriously complicated way of spelling things.
Of course, given how many accents our language has, some spellings make “sense” in certain accents where they don’t in another. The o in not is logical in the accent of Edinburgh (where it’s a pure o sound); less so in Detroit (where it’s closer to the a in cat). This begs the question: do children with certain accents have an easier time learning how to write than others?
This is worth some investigation, since one of the most problematic vowels in English spelling is also involved in one of the language’s largest accent divisions. This would be what we call the STRUT vowel: the short vowel in such disparate words as strut, cut, fudge, tough, won, one, ton, love, above, blood and run. It’s a challenge for a beginning English reader to see how one sound can be represented in so many ways.
But this is more of a challenge in some accents than others. In particular, this vowel doesn’t technically exist in the accents of Northern England. There, these words use the vowel in words like foot and put (ʊ). Hence the goofy transcriptions you sometimes read in novels set in the region (“Are you all right, loov?“)
You might say, then, that there is more logic to the spelling of these words for someone from Manchester than someone from London or America. From the Northern English perspective, o, u, oo, and ough look more closely related.
Is it easier, then, for a child from Northern England to learn to write these words than an American? (I struggled mightily as a child to grasp which word “rough” represented.) I’m probably overreaching, but it would be interesting to construct a study judging the speed at which children of different linguistic backgrounds learn to read and write (if this has not already occurred).
In the broader spectrum of language, of course, orthographies (spelling systems) are not alike in ease of use. Does this also apply to variations within a language? Is it possible certain accents are conducive to spelling and reading? While other accents impede literacy?
There is research out there on dialect and spelling acquisition/mistakes–a student in our department did her (undergraduate) dissertation on the topic, using three UK regional accents. Now, I don’t have her bibliography, unfortunately, and it’s hard to get the right search terms on Google Scholar without getting a lot of junk, but there is, for instance, this:
http://onlinelibrary.wiley.com/doi/10.1111/j.1467-8624.1997.tb01937.x/abstract
That’s great, Lynne! The mere abstract gives me some good stuff to mull over. I didn’t get into non-rhotic vs. rhotic accents in my post, but I can only imagine the trouble r-lessness can create. Especially when the trap-bath split comes into play.
It would be interesting to look at different types of non-rhotic accents, as well. In Liverpool English, for example, would “hurt” more likely be “het?”
“he o in not is logical in the accent of Edinburgh (where it’s a pure o sound); less so in Detroit (where it’s closer to the a in cat). This begs the question: do children with certain accents have an easier time learning how to write than others? ”
But this would imply that the symbol ‘o’ is somehow inherently connected to the sound [o] (and likewise letter ‘a’ to [a]), and that this magical connection is readily felt by all speakers of all accents. So that a speaker whose sound value is further away from the archetypal vowel posited by the symbol would be slightly more confused when he/she first learns the spelling. But is of course completely untrue.
It is a different matter with phonemic mergers/splits, and nonrhoticity. I can see how they could influence a child’s ability to master English spelling.
I was just about to post the same thing!
The last bit about phonemic mergers/splits and nonrhoctiy reminds me of certain eggcorns, like those dealing with rhocity and cot/caught mergers http://eggcorns.lascribe.net/
I knew that sentence was going to get me into trouble! However, I stand by it, because I’m not talking about o in general, but rather o in a particular context (“not” and other LOT words).
‘O,’ uncombined with any other vowel sounds, indicates two separate phonemes of English: the vowel in ‘rod’ and the vowel in ‘rode.’ That ‘-e’ indicates the preceding ‘o’ should be ‘long’ seems a simple rule to adult eyes, but is less so to a child with little experience reading. This is especially true when you include other ‘long o’ words like ‘post,’ ‘poll,’ ‘toll,’ and pre-rhotic ‘o’s in ‘port’ and ‘afford.’
Now in many Scottish English speakers, the vowels in ‘not’ and ‘note,’ while separate, are quite close. As the chart of F1/F2 values on page 4 of this study handout indicates, they actually overlap in some speakers. Contrast this with the accent of someone from Detroit, where the two phonemes are pronounced in entirely different regions of the vowel space, and you’ll see what I’m getting at.
I have no idea if Scottish children actually have an easier time grasping this distinction than kids from Michigan. But neither would I be surprised if that were the case.
do children with certain accents have an easier time learning how to write than others?
Children with a southern Irish regional accent ought to have the easiest time.
Scottish/northern Irish children would probably be a close second, but they are likely to have the cot-caught merger which might be a handicap.
In Ulster, it’s kind of mixed. For example, I make a distinction between cot and caught, but I don’t make one between Don and Dawn, so that may have been more difficult for me than for someone from the Republic. However, in other parts of Ulster, like Derry, for example, there is a complete merger. I understand that one has to make generalisations sometimes though.
I would also think that US children would generally be better off than English children. The rhotic-nonrhotic distinctions are all very useful for spelling, especially lettER/commA and NORTH-FORCE/THOUGHT.
American children, especially if they are cot-caught splitters, are probably almost as fortunate as Irish children. The father-bother distinction isn’t particularly important for spelling, because of the small size of the PALM set. The loss of the Mary/marry/merry distinction may be more problematic: “marry” words are fairly consistenly spelled -arry, and “merry” words are fairly consistently spelled -erry.
“The rhotic-nonrhotic distinctions are all very useful for spelling, especially lettER/commA and NORTH-FORCE/THOUGHT.”
I think someone may have mentioned this either here or on a another related blog, but for many Americans (like me) and probably even more so for Canadians, the vowels in NORTH-FORCE and THOUGHT don’t sound anything alike. I’m not talking about the presence vs. non-presence of /r/; I’m talking about the sound of the actual vowel in each set. I don’t think of those as being the same vowel at all and I don’t even have this “cot-caught merger” you write of. This caused me and other students in my introductory linguistics class some problems when we were learning phonemic transcription, because the book showed NORTH-FORCE and THOUGHT as being together in one set. A lot of my classmates had what I call a “WTF look” on their face when they saw that.
“The father-bother distinction isn’t particularly important for spelling, because of the small size of the PALM set.”
That’s true. Also many (North) Americans like me pronounce the /l/ in all traditional PALM words spelled with alm. The resulting vowel is very far back for me and maybe even a bit rounded. It sounds quite different from the vowel I use in father.
I think someone may have mentioned this either here or on a another related blog, but for many Americans (like me) and probably even more so for Canadians, the vowels in NORTH-FORCE and THOUGHT don’t sound anything alike.
Yes — but the crucial point is that for the great majority of English and southern hemisphere speakers, they do sound alike, with pairs such as pawn-porn and caught-court being homophonous. Such speakers are at a considerable disadvantage in spelling: is /’ɔːfəl / spelled “awful” or “orful” (the latter has 86,000 ghits)?
There’s a show called Pawn Stars here in America. I’m not sure who came up with the title. It must have been some Englishman.
Agreed. As the link that Lynne provides above shows, it’s pretty clear that r-lessness can get children into all kinds of orthographic trouble.
Hence the goofy transcriptions you sometimes read in novels set in the region (“Are you all right, loov?”)
For whatever it may be worth, I usually see midlands and northern English “love” (STRUT-FOOT merger) respelled as “lurve” – and occasionally as “luv.” I haven’t come across “loov” – but now that you’ve brought it to my attention, I probably will.
Judging from the high rate of creative spelling seen in American postings on the internet, the idea that Americans have an advantage in learning to spell strikes me as a non-starter. Otherwise I agree completely with Andrej.
Actually it seems that many North Americans, especially younger ones, have completely lost the distinction between intervocalic /t/ and /d/, giving rise to misspellings such as “suddle” (58,000 ghits).
This may well counterbalance any advantage gained from retaining the the rhotic/nonrhotic distinctions.
Australians may well have the shortest end of the stick. Non-rhotic with some t-voicing and the weak vowel merger. Does anyone know whether Australians find spelling particularly difficult?
Really good point about Aussies. They have the intervocalic t/d thing like we Americans do, and if you take a look at this vowel chart, they have somewhat less of a TRAP/START and FLEECE/KIT distinction than for most accents.
Re the intervocalic /t/ voicing – I once had a brief argument with a fairly well-educated, intelligent American thirtysomething about ‘deep-seeded’, which she claimed was a perfectly fine alternative to ‘deep-seated’. Fortunately, we had teh Internets nearby 🙂
http://itre.cis.upenn.edu/~myl/languagelog/archives/001815.html
It’s an interesting idea this and I have no idea what the answer is but I can relate something that I have noticed. As you say many languages have very phonetic spelling systems so it is very hard to go wrong spelling a Dutch or Polish word once you know the rules of orthography.
In my opinion learning to spell English is somewhat akin to learning Japanese Kanji even though that might seem far-fetched. What I mean though is that there are many words that you have to learn as individual entities and then once you know it you don’t think of it as remotely remarkable. So spelling Worcester is natural because it has a placeholder in your brain and not because of its pronunciation. Of course some English words are phonetic but I imagine that English speaking children abandon linking how they pronounce words to how they are spelt from a very early age. That’s why English speaking countries have spelling bees, a spelling bee for a Spanish speaker would be largely superfluous.
To cut a long story short I think that English might have more in common with Chinese and Japanese than we think. We learn the right ‘letter set’ for a word in a certain context which is why we can deal with so many homophones. That’s the same way as Kanji work. You know the sound but you have to know the context to find the right Kanji. We learn the ‘correct’ pronunciation of an English word and match it to the right letter set. Our starting point is not our own ideolect (or is it and I am just waffling….?).
That’s not far-fetched at all! In his book Empires of the Word, linguist Nicholas Ostler mentions that, in some sense, it might be easier to look at English as having a character-based writing system similar to East Asian languages. Just because we use a writing system rooted in the Latin alphabet doesn’t mean that our system of writing could be described as particularly “phonetic.”
Thanks for that reference trawicks, I must look that up. I thought that my observation was a bit off the wall largely because people often cannot understand how character based writing systems evolved. When I try to explain the beauty of the Japanese writing system and the extra expressive dimension it gives I am met with bemused looks.
Not sure if it matters in relation to this particular discussion, but it is interesting to note that the ‘STRUT/FOOT’ merger is actually a ‘FOOT/STRUT’ split. That is, there was originally only the FOOT vowel in English, then this split into two different vowels in most varieties.
I recently did some research into the way in which non-native speakers’ STRUT vowel changes towards FOOT the longer they spend in the north of England. One of the things I found was that the speakers were more likely to use a northern-influenced vowel when reading a list of words than in free speech. On closer inspection, it turned out that there was quite a clear orthography influence at work, with the letter ‘u’ triggering a more FOOT-like production, similar to their first language.
I actually think it’s quite germane to the discussion. We often forget that our spelling system was once much more logical and orderly than it is now. Alas, we just couldn’t keep those vowels still!
Historically speaking, there wasn’t even a FOOT-STRUT split, because “foot” had the GOOSE vowel, as its spelling suggests, in Middle English,
The PUT-CUT split would be a more felicitous name.
Interesting. So FOOT became a PUT word later then? Is this related to how some northern English and Irish accents have /uː/ in cook, book and maybe a few other words?
Very roughly, words spelled oo had GOOSE in Middle English, while words spelled u had PUT-CUT on Middle English. There are exceptions — for example “wool” was a PUT-CUT word and was actually spelled “wull”.
The accents that have GOOSE in the -ook words are therefore conservative in this respect.