|
|
........ |
Does English
have
a dyslexic orthography? by
Steve Bett, Ph.D.
Italian words are spelled the way they are pronounced, unlike many words in English and French. Since there is no tricky relationship between sound and symbol, Italian speaking dyslexics can cope better with their reading disability. [more] It is not the mother tongue that is tricky, it is the writing code. The following article defines the problem in terms of code overlap and polyvalence: factors which prevent some people from mastering reading and writing and most people from mastering English spelling. Problems that could be postponed and nearly eliminated through the use of a more phonemic writing system as an initial teaching alphabet [i.t.a.] or a parallel pronunciation guide notation.. The following article addresses these questions:
There have been hundreds of speeches and articles on our declining literacy rates and the problems that some children have in learning to read and write. Some of these articles suggest that the fault is with the learners [some are dyslexic or have other learning problems] or the teachers [some teachers are poorly trained or ineffective]. Teaching methods, television, and society in general have also been faulted. Only recently have some researchers suggested that part of the blame might be with the needless complexity of the spelling code used for English. If dyslexia were found only in children trying to learn English, should the dysfunction term be applied to the children or to the overly complex code we ask children to learn? According to one study, referred to by Richard Wade, only 17% of the general population over 16 can spell the following list of words correctly: accommodate. Dr. Yule's sweet 16 spelling test has been admiinistered to dozes of different groups over the past fifteen years. The overall average is in the 50% range. The average person is able to spell about 50% correct. This is consistent
with Bett's claim that English is less than 50% predictable. Some
have made claims that it is 85% predictable.
Irregularity in the spelling code brings on the symptoms of dyslexia Several researchers at a 1999 conference on dyslexia reported that English spelling and lack of regularity in phonological patterns is indeed an obstacle to reading and spelling efficiently. One researcher contended that "English is a dyslexic language." With its many irregularities and vast vocabulary, he said, English actually causes greater numbers of dyslexics than other languages do. Two of the keynote speakers agreed with this conclusion. http://www.ldonline.org Until this time, the code is rarely identified as the culprit because most people assumed that the code or orthography is part of the English language. Even in the recent press releases on the article in Science, the code was confused with the language. Once the writing system is identified as a code, its relationship to the language can be seen as a historical accident. Changing the code used to visualize speech would have no effect on the language or pronunciation other than the curious one such as those who try to pronounce the b in debt, the p in cupboard, or the e in vegetable]. Its the code not the language . . . There is nothing wrong with the English tongue or the roman alphabet. It is not the mother tongue that is tricky, it is the writing code. There are over 40 phonemes to be represented. What the 26 Latin characters lack in number can be made up through the use of markers, diacritics, and digraphs. This is not an ideal solution but it is quite workable. Linguists such as Sweet have commented, "There is nothing wrong with the Latin letters, only our irrational use of them." Mark
Twain said, "Our alphabet does not know how to spell." 1
Traditional spelling matches pronunciation guide spelling less than 40%
of the time. This was not always the case. The 10th century Saxon
alphabet and the Old English writing system had few of the code overlaps
and little of the polyvalence [multiple
values] we see in today's English orthography. (Scragg,
1974, Coulmas, 1990)
Notice that all phonemic notations spell different sounds with different letters and the same sound with the same letter. Words that rhyme are spelled the same. Ear rhymes become eye rhymes. [tee, see, kee, mee, seej, seez] [more] McLuhan wrote that the alphabet, unlike previous writing systems, could be mastered by anyone in a few hours. By an alphabet, McLuhan meant a consistent set of correspondences between the simple sounds [or phonemes] of the language and the written symbols used to represent these sounds [graphemes] (See the Unifon alphabet below). What could be mastered in a few hours was 20 to 40 sound signs - the exact number depending on the language being represented. English and most northern European languages need at least 40 sound signs or phonograms. [More on testing McLuhan's claim].
ALPHABET: A type of writing system in which
a set of symbols
The task of associating 40 sound categories with 40 graphic shapes is not that daunting. According to Flesch, Russian school children are routinely introduced to the Cyrillic 39 character alphabet at the rate of one letter per day. All children master the task within 4 months and from then on are able to relate the symbols on the page to sounds they have already mastered as native speakers of Russian. In 4 months, Russian school children display a level of proficiency that it takes the average English speaking children nearly 4 years to achieve. This is the penalty for having a confusing and inconsistent orthography. Complicated tasks take longer to master than uncomplicated ones. The problem with English orthography is that it is needlessly complicated. Normally the invention of superior codes replace inefficient codes. For example, The Hindu-Arabic numbers replaced roman numerals [more]. Language codes seem to be more resistant to such innovation and more tradition bound. Alphabets that
visualize the mother tongue's phonological code are the easiest to learn
Aelfabets
thaet vizyualaiz the muthher tunngz fonological cowd aar the iesiest tu
lern.
All languages are 100% phonemic. Differences in meaning are indicated by differences in sound or articulation. Writing systems are phonemic to the extent to which they visualize that phonological code. Most languages that use the roman alphabet are over 80% phonemic. Some like Italian and Finish are close to 100%. English is only 40% phonemic. English spelling matches pronunciation guide spelling less than 40% of the time. The advantage that countries with phonemic alphabets enjoy can be matched simply by adopting a more phonemic writing system for English. Children learning a phonemic code for English progress almost as quickly as children in other countries. The most researched phonemic code for English was Pitman's augmented roman, popularly known as the i.t.a. An even more efficient code, one without digraphs or two letter sound-signs, is shown below: Unifon means "one sound" per symbol. [unifon, 48 phonograms]
Some researchers who have compared reading and spelling skills across languages have concluded that most of the problems identified as dyslexia or associated with low literacy rates in the U.S. are directly attributable to the needless complexity of the traditional English spelling system [TES]. This complexity is largely lacking in the writing systems of most other languages. Cases of dyslexia and illiteracy after four years of schooling are extremely rare or undetected in Italy, Spain, Turkey, Finland, and other countries with highly phonemic or transparent orthographies. These countries have just as many children with the disorder, it just doesn't prevent them from learning a simple code. [more] Simpler codes are easier to master Simpler codes lead to greater mastery of reading and writing, fewer failures, and reduced learning time. If the English spelling system or code is ten times as complex as the spelling systems for Spanish and Italian; it is no wonder that school children in Spain and Italy can achieve in one year what it takes English speaking children four to six years to achieve. When words are spelled the way they sound, it is relatively easy to spell any word you can pronounce. In an alphabetic or highly phonemic writing system, such as Italian or Spanish, the way a word is spelled is a reliable guide to its pronunciations. International literacy expert, Frank Laubach, claimed that English had the worst spelling system in the world. It is certainly true that it is ten times more complex and inconsistent than it needs to be. Dr. A.J. Ellis showed that the 26 letters could have 658 different significations. 40 sounds should be represented with about 40 symbols not 600*. [The number of different spellings one can identify depends on the size of the dictionary. In an abridged 70,000 word dictionary, Dewey found 561 different ways to represent 42 sounds. ] In 1890, the philologist, Alexander J. Ellis, suggested a simpler code for English. This was later promoted as New Spelling and in 1960 became the basis of Pitman's Initial Teaching Alphabet (.i.t.a.). i.t.a. was not a methodology but simply a correspondence table where each symbol was associated with one and only one sound. In English, each of the 42 sounds in the language can be spelled an average of 14.7 different ways. In i.t.a., each sound was spelled just one way. i.t.a. sound signs are quire similar to those indicated in the Truespel alphabet below. Notice that a minimal alphabet for English requires the addition of 12 vowel symbols and 4 consonant symbols. The Truespel alphabet does not have a phonogram for every important speech sound but the remainder of the combinations can be easily figured out [e.g., are=aar, ear=eer, tire=tier, tower=taawer, tour=tuer, toward=tord, terrace=tairis, tauras=tauris] [See converter] What spellings that are changed when English is regularized?: Redundant e
at the end of words without a long vowel: have, give, come, ...All forms
of regularized English drop redundant silent letters. have,
give, done, gone, definite become hav, giv, dun, gon, definit. In Truespel
[twelve thieves] becomes [twelv theevz]. If you think it is hard
to drop illogical letters, imagine the problem that children have in learning
to use them.
Fortunately, 75% of the words that rhyme use only 4 different vowel spellings. In other words, knowing the sound, a student should be able to guess the correct spelling 75% of the time in less than four tries. This experimental finding from a study by Hanna has been often misinterpreted as proving that English is over 75% predictable or over 75% regular. English is nearly 75% predictable at the phoneme level but only 40% predictable at the word level. It is next to impossible to remove all of the guesswork in spelling for two reasons: [1] some words are pronounced differently in different regions and [2] unstressed vowels have no uniform pronunciation or spelling. A phonemic solution is only fully predictable where there is no variation in dialect. Thus, for a broad phonemic solution to English spelling that cuts across many dialects, the goal is about 90% predictability, not 100%. Thus. the 40% phonemic rating in the chart, is effectively almost a 50% rating. About half of
the vowels in English words are unstressed. It is next to impossible
to predict the spelling of an unstressed vowel. About the only solution
to predictable unstressed vowel spelling is to spell them all the same.
Thus [language] is spelled langwaj in Spanglish but it could just
as easily be spelled langwij, langwej, or langwoj, or langwuj. Spanglish
arbitrarily spells all mid lax vowels with an [a]. It works out well
in lanwaj but is less than satisfactory in other words: e.g., pennsal
for pencil or surrcas for circus. It works for
hurry
but not as well for hurrder.
With i.t.a. [Pitman's Initial Teaching Alphabet], the student was supposed to be able to guess the correct spelling 100% of the time after one try. i.t.a. spelling was basically a systematic alternative to invented spelling. i.t.a. was an alternative to guessing or inventing the spelling of familiar pronunciations or guessing the pronunciation of unfamiliar strings of letters. The Truespel alphabet above shows one way to consistently spell 40 important English speech sounds. Any time you hear an "ah" sound as in "ox", it must be spelled aaks in this orthography. An i.t.a. such as Truespel diminishes the need to guess which letter goes with a particular sound from over 4 to less than 2. The result matches the dictionary pronunciation guide. It matches traditional historical spelling less than 10% of the time. http://www.unifon.org/syllabics-semivowels.html
Comparative studies of languages have long revealed that languages with transparent orthographies are much easier to master. The question of how much easier has never been fully answered. It is just as easy to learn to read and write English when a phonemic code is used in place of the polyvalent traditional code. The 29 word study Another approach The initial teaching alphabet or i.t.a. Sir James Pitman believed that a consistent orthography would simplify teaching across any methodology. There was some basis for this optimism. Phonics is often not the approach used to teach children in Spain or Italy where early reading achievement is over twice as high as in English speaking countries. What counts is the orthographic transparency of the writing system. In the late 1960's, the British House of Parliament passed a bill funding a bold experiment with consistent spelling. British schools were given the books and collateral materials they needed to introduce this new medium of instruction: the Initial Teaching alphabet or i.t.a.. Teachers were given a crash course in i.t.a. but were not told specifically how to teach it. i.t.a. was extensively researched in the early 1970's [see Downing]. Children could learn i.t.a. almost as easily as children in other countries could learn their consistent spelling system. The shortfall has been attributed to the inability of parents to help and the shortage of books written in i.t.a. There would be no shortage of books if truespel where used as an i.t.a. because any digital book, [and there are over 10,000 of them on the Web] can be converted to Truespel using the easy cut and paste on-line converter. Transfer of skill
hypothesis
Since the goal of i.t.a. was to find a better way to teach children to read and write, i.t.a. tried to look like TES [Traditional English Spelling]. Unfortunately, there is no way for a systematic spelling system to look much like a chaotic one. i.t.a. spelling matched TES or traditional dictionary spelling only 40% of the time. No phonemic rendering of English speech will match traditional or historical spellings more than 40% of the time. To test this claim, count the number of times the dictionary pronunciation guide matches dictionary spelling. In the chart above, the phonemic or consistent sound spellings [i.t.a., IPA] match the traditional spellings for only one group of words. i.t.a. correctly spells *moon and cartoon but fails to match 17 other [TES] spelling patterns. i.t.a. matches TES when spelling *guru and flu but fails to match 17 other [TES] spelling patterns. [See: How many ways can you spell *day?] With i.t.a.,
student's quickly picked up the idea of how an alphabet is supposed to
work but were left to their own devices when the time came to transition
to TES. No organized attempt was made to help children get from toon as
in *cartoon to *tune. In the 4th grade, children were expected to abandon
i.t.a.
for TES. The reading transition went fairly smoothly and children trained
in i.t.a. were able to retain their advantage. Spelling was another
matter. Many children had problems trying to respell
i.t.a. words
that did not match TES. This is understandable since over 60% of the words
did not match.
Given the fact that i.t.a. worked regardless of the teaching methodology and with serious gaps in the support system [insufficient materials, no assistance from parents, no consistent teaching metrology, lack of linguistic sophistication on the part of the teachers...], why was it abandoned? Basically, it was a fad and all fads soon loose their luster. i.t.a. was not understood by parents. i.t.a. was not supported by the major educational publishers. It was more costly than traditional approaches to the teaching of reading and was deemed administratively inconvenient. Had the i.t.a. been incorporated in a proven teaching methodology it would have been twice as effective but this never happened. Interest in phonics has been revived. Is there any chance that i.t.a. could be revived? It is possible. Except for government support, the conditions are as favorable as any time in history. In 1970, a school had to buy a library of i.t.a. books and materials. Today, an individual teacher with a classroom computer could generate all of the needed materials. There is a vast library of digitized books on the Internet as well as an on-line converter that will change the spelling from TES to New Spelling. [New Spelling and Fonetic are non-ligatured versions of i.t.a.]. Teachers could easily generate their own materials. How the English alphabet was lost In the early 1800's, Noah Webster remarked, "Letters, the most useful invention that ever blessed mankind, lose a part of their value by no longer being representatives of the sounds originally annexed to them." The effect is, "to destroy the benefits of the alphabet." Webster was aware that there was a time in English history when the language had a functional alphabet. Tenth Century clerics devised a Latin based alphabet for English that made it possible to "spell words as they were pronounced and pronounce words as they were spelled." Could the restoration of the benefits of the alphabet be as simple as restoring the Saxon alphabet? Could the usefulness of the alphabet be restored by restoring the sounds originally annexed to the letters? Spelling reformers such as Webster and Franklin desired a closer connection between spelling and pronunciation. Both desired alphabets that allowed people to pronounce words as they are spelled. Benjamin Franklin, a printer by trade, even produced one. The link between spelling and pronunciation was lost in the Great Vowel Shift [around 1400 AD]. Prior to that time there had been some quirky spellings introduced by Norman French scribes but the basic sound system still matched Latin. Now 60% of the words in the dictionary do not match the pronunciation guide. To make matters worse, the vowels in some words did not shift. This created code overlaps where words that are spelled the same have different pronunciations. This is the theme of the poem below and a more famous one called The Chaos. It is important for teachers to be clear about the problem. It is important to be clear about the complexities of the English writing system and their social impact even if no solution is at hand. This paper suggest a possible solution: the restoration of the alphabet. However, the restoration of the last consistent alphabet used for English is probably more difficult today than it was 300 years ago when Noah Webster and Benjamin Franklin made their recommendations. |
| Our Strange
Lingo [code overlaps]
When the English
tongue we speak. Why is
|
Beard
is not the same as heard
Cord is different from word. Cow is cow but low is low Shoe is never rhymed with foe. Think of hose, dose, and lose & think of goose & yet with choose Think of comb, tomb and bomb, Doll and roll or home and some. |
Since pay
is rhymed with say
Why not paid with said I pray? Think of blood, food and good. Mould is not pronounced like could. Wherefore done, but gone and lone- Is there any reason known? To sum up all, it seems to me Sound and letters don't agree |
| transcriptions
of problem words in Spanglish [a=ah, i=ee]
and Unifon
Bold words above are spelled phonetically breik frik [friek]
/ brAk frEk
|
bird h'rd /
bErd
hcrd bierd
h'erd
cord w'rd /kord wcrdcord w'ord cau low / kx lO cau lo' low [go to the COW POEM] shu fow / SC fOshu fo' ho'z do's luz / hOz, dOs, lUzhowz, dows, luuz gus chuz / gUs KUz guus chuuz co'm tum baam / kOm tUm bomcowm tuum bomm dol ro'l ho'm s'm / dol rOl hOm sumdoll rowl howm summ |
peid sed /
pAd
sed
bl'ad-blod fuud gwd [gud] / blud fUd gCd mowld [mo'ld] cwd [c'd] / mOld kCd d'an gon lown / dun gon lOn Izz therr enny rizn nown? Tu s'm upp ol, itt siemz tu mi - Saund 'n lett'rz dont agri Chaos poem |
C
........ |
Can the English
alphabet be restored?
What would be the cost of such a reform? Almost anyone can come up with a more consistent way to spell English words. Two alternative phonemic codes are shown above [and also below]. The new spellings are consistent but appear odd. They can be easily sounded out by referring to a correspondence table. However, they are not what we are used to. The sheer number of alternatives to TS has tended to dilute the support for any one proposal and prevent any rationalization of the more consistent orthography. With 1000's of simpler more efficient codes, agreement on one is difficult. Any phonemic reform of English would require respelling 60% of the words in the dictionary. This is no problem for children or ESL learners, but it is more of an adjustment than most adult English speakers want to handle. Although IPA is a little hard to read, there are a number of phonemic proposals that can be easily deciphered. The objection to them is that they cannot be read as fast as TS. Speed readers read word patterns. They do not try to sound out words. Half way reform proposals which preserve word patterns such as removing all of the silent letters [for example, the silent e in give and have] have not fared any better than full reforms. Those who have completed primary school prefer to keep a familiar code no matter how inefficient and inconsistent. They prefer an inefficient code they know to an efficient one they will have to learn anew. Given the choice, they prefer to spell giv, liv and hav with a redundant and misleading terminal e. Most people are unaware of the fact that English lost its alphabet in 15th century. An alphabet is a consistent set of correspondences between sounds and symbols (letters). Alphabetical writing systems are highly phonemic. Old English (Anglo Saxon) was over 90% phonemic or consistent with its correspondence table. Modern English is only 40% phonemic. It is consistent with its pronunciation guide only 40% of the time. The traditional English spelling system [TS] uses historical spellings but not the historical long vowel sounds or the consistent historical alphabet. As a result, about 60% of the words are not pronounced as they are spelled. Most of the separation between spelling and pronunciation occurred during the 14th century during what was called the "great vowel shift." The vowel shift did not affect the short vowels but shifted the long vowels to a more closed jaw pronunciation: In many words, /a/ [ah] came to be pronounced /ae/ [ash], /ae/ became /e/ [eh], /e/ became /i:/ [amigo, machine, si], and /i:/ became /ai/.[aisle] . Word pronunciation changes over time Pronunciation changes over time. To preserve a consistent alphabet, when the pronunciation of a word changes, its spelling also has to change. Countries that set up academies to revise spelling to keep it aligned with pronunciation have managed to maintain their alphabets. England never set up an academy and the 18th century dictionary writers were reluctant to reestablish a connection between speaking and writing after the great vowel shift of the 14th century. The solution that has been suggested for the past 300 years is to adopt a phonemic notation and spell words as they are spelled in the dictionary pronunciation guide. The chief problem with this solution is that it changes the spelling of 60% of the words in English. Two examples of the phonemic spelling of English are shown above. Literate readers read logographically or in terms of whole word patterns - they rarely sound out a word on the basis of individual letters. Thus changing the look of a word for speed readers will reduce their reading speed. It is not that literate readers cannot read a passage that is spelled phonetically, it is just that they cannot read it as fast. After a phonemic reform, it may take as long as a year for whole word English readers to recover their reading speed. The ones that benefit from alphabetical or consistent spelling are the young not the old. With a better code, the young could acquire a high level of literacy four to ten times as fast as they do at present. Those who have learned the complex code benefit little from its simplification Literacy is largely a decoding and encoding skill. Readers and writers begin by associating written symbols [graphic shapes] with spoken sounds. Since there are only about 40 significant sounds in English speech, an efficient code would associate them with 40 symbols. 40 sounds would be referenced 40 ways. Each sound [phoneme] would be referenced by one and only one letter or letter combination. With the traditional code, however, the 40 sounds are referenced in 615 different ways [Ellis, 1900]. Each letter is associated with an average of not one but 14 different sounds [Dewey, 1971]. [Here is a list of 18 of the 29 different symbol configurations used to represent the /u:/ sound]. Students must associate 26 letters with 40 sounds in over 400 different ways. Instead of learning 40 paired associates students must learn over 400. The complexity of the orthographic code makes the learning task ten times more difficult. It is over 10 times easier to associate a shape with one sound than it is to associate it with 14. The advantages of an alphabetic reform are not quite as great as reformers claim. One reason is because the base pronunciation is not necessarily the same as the one the child uses. After the Saxon alphabet was restored, the spelling pronunciation of [tomato] would be taw-mah-tow - not tow-mey-tow. The child might have to learn two dialects, one for spelling and one for conversing with his or her peers. A phonemic spelling reform would not have an immediate effect on the way that people pronounced words. The traditional (mid 18th century) English spelling system [TS] is based on the notion that the business of spelling is to represent the origin and history of a word instead of its sound and meaning. The playwright George Bernard Shaw (1941)] argued that this reduced the alphabet to absurdity. TS can be called non-alphabetical since the spelling of more than 50% of the words do not match the dictionary pronunciation guide. The disconnect between spelling and pronunciation limit the effectiveness of the phonics approach to the teaching of reading. If pronunciation continually changes, why try to match it? Dr. Samuel Johnson, who wrote the first popular dictionary in 1755, felt that it was folly to imagine that the dictionary could embalm language and preserve its words and phrases from mutability. He saw no reason to standardize English spelling beyond the word level because pronunciation was not standard across dialects and continually changed . As it turns out, English pronunciation is probably more standardized today than in Johnson's time. Compared to the changes that occurred in the 14th century, English pronunciation has hardly changed at all from the way it was spoken in London in 1755. Some words and phrases have dropped out of favor and new words and phrases have been added. Most of Johnson's spellings, have survived intact and seem to have been much more effective at embalming language than he expected. The writing system can either obscure the etymology or historical spelling of the word or obscure the pronunciation of the word. Traditional English Spelling [TS] obscures the pronunciation. The notations
below ignore etymology and morphological consistency. They simply
spell each sound one and only one way. Spanglish, which uses digraphs
or 2-letter combinations instead of special characters, is supposed to
be a little easier to read and write.
Conclusion Go back to the top and see if you can answer the 9 questions.
|
It
is true that the average adult can memorize 40 paired associates in a few
hours particularly if a mnemonic is employed. For example, the 
