[Journal of the Simplified Spelling Society, 1988/3 p.11-13 later designated J9]

[See Journal and Newsletter articles by Edgar Gregersen.]

The Strategy of Spelling Reform in Stages: Pros and Cons.

Edgar Gregersen.

Edgar Gregersen is Professor of Anthropology at Queens College and Graduate Center of the City University of New York. He has a special knowledge of accents of English, Egyptian hieroglyphics, the alphabetization of West African languages, and the spelling of Norwegian. This article is based on a paper given at the Society's 1987 Conference 'Spelling for Efficiency'.

Advocates of reform by stages - and objections.
The Simplified Spelling Society has recently proposed a series of modest reforms rather than a radical one-step overhaul of the present orthography of English. These proposals have been set forth in the Society's leaflet Tough Though Thought. As an alternative, some members of the Society have urged that Cut Spelling be promoted as a first step.

In a similar vein, Harry Lindgren, in his Spelling Reform: A New Approach (1969), specifically called for a 50-stage reform, to take at least 50 years. Actually, the time period for the full reform would probably - even under optimal conditions - be considerably longer because the very first step may take several years all by itself.

On the other hand, Edward Rondthaler, the proponent of 'Simplified American Spelling', has changed his mind and is now against stages. Originally he proposed about three, but now he feels that anything less than a total overhaul would cause a great deal of confusion, if only because many words would have multiple representations.

My own view is that an overnight total reform would be the most efficient and desirable approach in the long run. But barring enlightened despotism, a Kemal-Atatürk-style revolution, or persuading Oliver North to divert funds from the Contras to stage an orthographic coup in the USA, this is unlikely.

I am not against stages, however, if used as a tactic to arouse public interest in reform. Certainly the use of stages in private publications and in propaganda is quite justified. But getting governments in English-speaking countries to go along with a piecemeal approach is something else

Let us consider two practical situations.


The first of these is the spelling reform of Russian that occurred shortly after the Revolution. Altho initially planned by a commission under the last tsar, Nicholas II, the reform was carried out under the Communists, many of whom saw it as a first step towards their international-minded goal of romanization. In fact, the Soviets created decent roman orthographies for many non-Russian-speaking native peoples in Siberia and elsewhere. Ultimately, Russian nationalism triumphed over Communist internationalism: plans for romanization were abandoned and modified cyrillic alphabets replaced the roman ones set up just a few years before. Since the major spelling reform for Russian (in which several letters believed to be superfluous were dropped, including <i> and <θ>), only occasional and trivial changes have taken place, e.g. и д т и (idti) for и т т и (itti), to go. Improvements of a more thorogoing kind, such as the marking of stress, have apparently not even been considered.

The basic moral from the Russian situation is that if stages of reform are indeed accepted, each stage had better be selfcontained because it may be the last one carried out.


The second practical situation I shall consider is that of Norwegian riksmål/bokmål. The Norwegian situation has considerably more complications than most other languages, in large part because of a powerplay involving social classes and geographic regions. A large part of the controversy that has inflamed the Norwegian reforms does not involve spelling as such, but rather what is to be regarded as the standard spoken Norwegian, which the spelling would reflect.

In the early 19th century, most Norwegians wrote following Danish conventions even tho they did not use Danish pronunciation. Let us consider the changes that occurred in the three major reforms of the 20th century, those of 1907, 1917, and 1938, by examining the following five words, given first in their older Danish spellings (a spelling reform in Denmark occurred after World War II, two of the major features of which were abandoning the use of initial capital letters for nouns and the introduction of the letter <å> from Swedish and Norwegian for older <aa>): Blæk (ink), bleg (pale), Kagen (the cake), Gaden (the street), Gaade (riddle).







The reform of 1907 tried to introduce as the standard it reflected the language of educated speakers from Oslo using a relatively formal style. The spelling therefore abandoned for the most part <d, g> between vowels, to reflect the current unaffected pronunciation with <t, k>. The reform also very greatly restricted the use of the letter <æ> (unless it occurred before <r>) and generally substituted <e>. Further, the reform did not deal with vowel length consistently; hence Blæk with a short vowel and bleg with a long one both wound up as blek. This confusion of long and short vowels before final consonants was systematic, so that except for the earlier capitalization distinction, men (Men) (damage) and men (but) have traditionally been written the same til this day. (But note menn [men], former Mænd, with a short vowel, pronounced the same as men [but].)

This systematic confusion was a major defect of the writing system. In 1917 it was resolved by usually doubling a final consonant after a short vowel (as had been done within a word, e.g., 1907 blek [ink] but blekken (the ink], 1917 blekk [ink], blekken [the ink]). The 1907 orthographic peculiarity was memorialized in the phrase:
"Hvad trenger du med pen [penn] og blek [blekk], du som er saa pen og blek?" (Why do you need pen and ink, you who are so beautiful and pale?)
The 1938 reform introduced few new spelling rules, but tried to change the standard used from upper middle class dialects to urban working class dialects (e.g. gata). Altho such forms are official they have met considerable resistance.

The result of all these changes is that people in different age groups may continue to write using spellings that are no longer official. Until quite recently (and possibly still), some older people even used the 19th century Danish conventions. For the most part, people take all this in their stride. Dictionary makers usually just ignore older variants. Even if they didn't, the result would be only slightly fatter dictionaries and considerably more cross-referencing than commonly found.

The moral for us is that reform in stages is not an impossible option, altho it is a messy one. The Norwegian situation got more complicated than need be by juggling nationalism with linguistic requirements (e.g. dropping <w> to become more Norwegian-looking - or at least less Danish-looking - vs. marking length of vowels).

The Simplified Spelling Society.

Let us now consider various reforms in stages as proposed for English.

The Simplified Spelling Society in its Tough Though, Thought leaflet suggested a reform it labeled 'SR:ough'. In line with this reform, the following changes were to be made:
drought→ droutplough→ plou
thorough→ thurrathough→ tho
dough→ dohcough→ cof
bought→ baut 

Everyone agrees that traditional spellings with <-ough> are horrible, but an enormous number of problems confront us in solving them.

The least objectionable change is from drought to drout, since all that is involved is dropping the totally superfluous letters <gh>. With plough, one might argue that the same thing applies. But note that another spelling already exists: plow. British and Commonwealth speakers of English may brand this as an Americanism - and certainly cultural feelings of this sort must be taken into account. But are we to reject all reasonable forms because they are American? Furthermore, plou has a decidedly unenglish look to it because <-ou> normally doesn't occur finally, only <-ow> (how, now, cow, etc).

The form thurra for thorough is not Nue Speling (where it is written as thurro). And the suggested spelling of the final vowel opens a whole new kettle of fish that I'm afraid is more unfortunate than one might think. Altho the RP pronunciation of the word ends on the same vowel as China, the normal US and Canadian pronunciation of the word has the final syllable the same as in follow, sorrow. To spell this word (and also borough) with final <-a> rather than a compromise <-o> would tend to split the English-speaking world. I think it unwise to introduce such complications into the earliest stages of reform and probably into any stage of spelling, which it seems to me should be as internationally acceptable as possible.

The spelling tho is fine (tho I myself prefer dho). But doh for dough is just out of the blue and certainly goes nowhere. The Nue Speling form doe would have made sense, but no spelling system for English I know advocates <oh> for the vowel sound in go.

The spellings cof and baut for cough and bought again introduce dialect differences. In the speech of older RP speakers the vowels of both words are often pronounced the same, and this is also true probably for most Americans. To use different vowels in the spelling without any further clarification is therefore quite unfortunate. (By the way, words such as bought, ought, fought, frought, wrought could perhaps better be spelt as boht, oht, foht, froht, roht in a first stage: dropping the <ug> should appeal to proponents of Cut Spelling and it more nearly approaches an international value for vowel representations.)

To sum up this rather tortuous discussion: 'SR:ough' is simply too complicated to be a desirable early stage of reform. A considerable number of decisions of unequal value have to be made at once. Sometimes the changes made do not suggest the general direction reform is going (as most obviously with doh for dough). In short, 'SR:ough' should be abandoned.

Harry Lindgren.

The suggestions made by Harry Lindgren have much that is admirable to my mind. For example, he maintains that each and every step must be unambiguous and complete. He seems also to suggest that there must be a concern for the sequence of stages. I for one think it would be very unfortunate to change hence to hens before present hens had become henz; or to change off to of before present of had become ov. Lindgren in effect acknowledges the same sorts of problems, but without actually giving the details about specific stages, except for SR:1, which always writes stressed short /e/ as <e>.

However, his scheme does present serious problems, most notably in his insistence that the 'obscure' unstressed vowel shwa [ə] as found in about, China, and so on, be consistently shown (as <'>). This decision immediately confronts us with a very fundamental question: preserving the unity of the English language community. Enormous variation exists with regard to how unstressed vowels are pronounced. Consider simply the following few examples contrasting usage in RP and one variety of American English.




o'mit (moribund?)
ko'kejn (moribund?)
General American




Another drawback to showing shwa is that doing so obscures the relationship between related forms as in phətógrəpher vs phótəgraph, or históricəl vs híst(ə)ry.

Whatever the final judgment as to how such unstressed vowels should be shown, any decisions that would tend to break up the unity of the English-language community should be weighed very seriously and delayed until the very last stage of reform - if ever adopted at all.

In other words, it strikes me that the approach used at present in Russian of not showing vowel reduction could be adopted in English - or rather retained, since traditional English spelling does precisely that for the most part. (In more technical terms, I'm advocating the orthographic inviolability of the morpheme, the smallest unit with meaning - e.g. the photo part of photograph[er].)

The danger of having to reverse reforms.

A third and last situation to be considered comes from a Cut-Spelling-like approach to reform. (Chris Upward assures that this particular solution is not advocated by proponents of Cut Spelling, however*). Consider the traditional English spelling breathe. One possible reduction within a Cut Spelling approach would certainly seem to be brethe. Since this is so, forms such as sleepy, sleeping, and sleeper might be cut to the shorter forms slepy, sleping, and sleper, even tho the underlying form sleep would have to be retained unchanged. Here we have, first of all, a problem of unnecessarily breaking up related forms (i.e. we would violate the integrity of the morpheme). But what about the final stage of reform? What if we decide that the vowel sound of sleep is always to be written <ee>? We would go back to the traditional spelling.

An even worse situation could occur with the two words who and hoe: [1]
TOwho  hoe
 \ / 
Intervening stage (CS)  ho 
 / \ 
Final stage (NS)huu hoe

Here the intervening reform stage lumps together two words pronounced differently only to have them re-differentiated in the final stage.

To avoid such awkward situations, which could only invite scorn from people opposed to spelling reform, stages must be planned with an eye to the final comprehensive system. It has been said that the proposals of the Simplified Spelling Board of the United States (now defunct) failed precisely because it proposed no final comprehensive scheme and gave the impression of wandering in the dark with some very ad hoc solutions. (See their Handbook of simplified spelling, 1920). For example, for the sound of the vowel in sleep no clear direction was given. Words ending in <-ceed> (proceed, succeed, exceed) were to coalesce with the <-cede> suffix (like precede); hence, procede, succede, excede. Words with <ei> were to be rewritten with <ie> (reciev for receive); words with <ae> and <oe> were to be cut to <e> at the beginning and middle of a word but not finally: fenix for phoenix, enciclopedia for encyclopaedia, but alumnae unchanged. This approach is the way of madness.

Let me restate my position: any kind of piecemeal changes, even if single words, may be a justifiable strategy for jarring the public into an awareness and eventual acceptance of rational spelling. But these changes should be self-contained and most of all should not have to be undone in later stages. My own preference for a stage one reform would simply introduce new symbols that are necessary for a decent spelling of English but have no tradition behind them such as accent marks to indicate stress, or the Klasik Nue Speling forms <dh, zh, ngg, aa, uu>. To do so would in effect get the most difficult job done relatively then have the embarrassing situation where the end result painlessly.

However, once a momentum for change is achieved, spelling reformers should abandon a strategy of stages and push for a comprehensive, one-time reform.

[1] Pt 2 of 'Conflicting Eficiency Criteria in Cut Speling' in Journal J10 1989/1, will discuss how CS could treat these long vowels and potential ambiguities such as who:hoe. - Ed.

Back to the top.