[Spelling Progress Bulletin, Winter 1979, pp19,20]
Also on this page: Book revew.
[Donald Scragg: see Journals.]

Analogy in English Spelling, by D. G. Scragg.*

*Instructor, Univ. of Manchester, Manchester, Eng.
Presented at 2nd International Conference of S.S.S. at Nene College, Northampton, July 29, 1979.

It is my contention that for any effective improvement to be made in the efficiency of our spelling system, there must be very full information not only on the way in which we use the orthography today (which is what most of the papers given at this conference are concerned with) but on the way in which our system has developed over the past ten centuries. The major problem with English spelling, after all, is that it has a longer history than the spelling of any other language using the alphabet - there is a longer continuous history of writing in English than there is in any other European language. We cannot - and should not - ignore this history, but we may learn from it.

English spelling has grown organically - not haphazardly, as some have supposed - over the centuries, and clearly some of its developments have been advantageous (for example, an alphabet of 26 characters is obviously better geared to a reflection of speech patterns (if that is the desired aim) than is the alphabet of 24 characters which was all that was available to Shakespeare). I believe that the more we understand about 'how our spelling system evolved, the better our chances of seeing why - not just how - it went wrong, and the more successful we are likely to be in putting right those anomalies which we all recognize that it contains. Four years ago at the first conference I talked about the way in which English spelling stabilized into the present system whereby we have a single fixed form for each word (with a very few exceptions). Today I intend to examine with you some of the effects of one particular linguistic factor - analogy - on spelling of different periods, and I hope to be able to show how we could capitalise on it in our efforts to improve the efficiency of our written language.

First let me explain my specialised use of the term analogy. Linguistic analogy is the influence exerted by the use of language in one context on another in which it is historically inappropriate. To explain by means of an example: the historical plural of staff (a stick or long narrow piece of wood in its original meaning) is staves, but this plural involves a considerable disruption of the sequence of the sounds of the singular the replacement of a simple vowel by a diphthong // and the alteration of /f/ to/v/. Most English words form their plural by adding s, so by analogy with the majority pattern in the language, a new plural staffs was formed. Staves as a plural was reserved for situations in which the word in its original sense was regularly used in the plural, e.g., a fence is made of a series of upright sticks or staves. When you pick out one of these, it is called a stave - by analogy of the fact that to form a singular in English you normally simply take away the s. So here we have two new words created in the language: staffs and stave - both created by analogy.

Analogy is a powerful factor in many areas of language, not least in spelling. We all know how children make spelling errors by applying to one word criteria for spelling which are strictly applicable to another. Hence because the sound /ɔ/ is spelt or in /or/ itself, why not cort or bort? This is spelling by analogy. Equally, on the analogy of bought, caught, daughter, fought, naught, sought, and taught, all of which have the sequence a or o followed by ugh, why not pought for port or waughter for water? A child acquiring the tortuous (or taughtuous) technique of modern British spelling has no inhibitions about such forms, but he is normally weaned away from them to the conventional spelling of the printed word. If in doubt, he turns to the dictionary for the generally agreed and acceptable form. Now dictionaries are a relatively new invention. There was no attempt at an all-inclusive lexicon of English before the 18th century, and consequently before such a complete catalogue of all words with their fixed spelling was achieved, it was much easier for an individual word or a group of words to be influenced by a prevailing fashion or popular analogy. Writers had no dictionaries to stifle them. English borrowed the word delight from French. The French spell it (or did when they used it) -ite. But in English, words with the sound sequence /aɪt/ are more rarely spelt ite than ight (c.f. might is commoner than mite, sight commoner than cite or site, and there is also bright, fight, fright, flight, light, height, night, right, tight, against bite, kite, and quite), so by analogy we have come to spell delight with igh. Frequency of use of a symbol (however cumbersome that symbol may be) is an important factor here: it is the basis of Prof. Wijk's proposed reform.

Analogy, however, need not be confined, in the case of borrowed words like delight, to the recipient language. The first man to write delight as we do did so because he was thinking of the sound sequence which is also found in light, but many of his contempories in the 16th century were less interested in the echoes of the written word in the spoken language than in the reverberations from other written contexts. Lots of words we use are what might be called literary words, in that they are more often encountered in written language than in speech. I suppose analogy is a good example. In the 16th century English prose, especially the prose of the technical language such as that of science, medicine, learning generally, was still in its infancy. Most learned works were written in Latin. When a writer used English, he was aware that all his more literate audience was familiar with Latin, and hence when he created a literary echo in his reader, he was as likely to do so in a passage of Latin as in a passage of English writing.

Today it is sometimes said by opponents of spelling reform that it is useful to have visual links between related words. The linguistic philosopher Noam Chomsky has used the example of the pair doctrine, doctrinal. 400 years ago, a similar visual link was created by inserting an excrescent and entirely unpronounced (perhaps unpronounceable) b into debt and doubt to show their association in meaning with such Latin words as debitum and dubitare. Let us take note of the fact that writing is not a simple reflection of speech - writing has no way of symbolising patterns of intonation and has very inferior devices for denoting the variations of stress and pitch practised in speech. But, in a form of compensation if you like, writing has this ability for creating visual echoes which speech lacks, and spelling reformers will ignore this aspect of written language at their cost. What I would say, however, is that visual echoes of Latin are useful only when it can be guaranteed that the readers of English are equally fluent readers of Latin. Such is not now the case and I would say that the b in debt and doubt has outlived its purpose.

However, there are instances of the practical operation of analogy. Take the verbs could, should and would. All three have a silent l. In two, should and would, the l is etymologically acceptable in that these are historically the past tense forms of shall and will, in which d is added much as it is in walk, walked. As in walk, pronunciation of the l has been dropped since the word acquired its fixed written form. But could is a different kettle of fish entirely. Historically, it is the past tense of can. It acquired its very different pronunciation through a complicated series of changes which began some centuries before the birth of Christ, but at no time did anyone ever pronounce could with any of the sounds we normally represent by l (except perhaps in error). In this word, l is an analogical spelling, introduced by association with should and would. But is it not a useful association? After all, these three words are unique in the sense that while they carry no lexical meaning (they have no referential meaning which a dictionary can define, as it can say with horse or ride), they do have a distinct grammatical meaning (they are modal auxiliaries - not indicating an action as a lexical verb like ride may do, but a subject's attitude to the action: could ride, should ride, would ride). In this sense it is perhaps useful to have them marked off, by their visual form from the general run of written forms.

Let me return finally to Prof. Wijk's Regularized Spelling which I briefly referred to earlier. He has applied the principle of analogy systematically and I take this opportunity of applauding his system publicly. I offer only two thoughts on it.

The first - and this is very obvious - is that the success of any system based on analogy depends wholly on the point of the analogy. In other words, if the statistical survey on which the analogy is based is insufficiently broad, then the disturbance of the conventional system will be out of proportion. (A simple example - I showed at the beginning that there are more words in English with the sequence ight representing the sounds /aɪt/ than there are words in ite. But this does not take into account the frequency of occurence of such words, and in order to arrive at a proper estimate of the frequency of occurence, you have to choose your sample very carefully. The word light will occur more often in War and Peace than the noun kite, but this alone is not enough to prove that it is of more frequent occurence in the language as a whole. Conversely, readers of Enid Blyton may be more familiar with kites than with fighting.)

Secondly, I should like proposals for spelling reform to take account of what technically I call morpho-phonemic rules. My example of could, should, would falls into this category but perhaps it is simpler to see it in the plural marker in nouns: in speech, we have three regular plural markers depending on context. /s, z, -ɪz/ in cats, dogs and horses. Would it be wise to have three differing markers in writing, or is it simpler to have just one?

The moral of my tale (let me stress it again): there is more to writing than a simple reflection of speech.


Editor's comment:

Something simpler is not necessarily better. In this case, simpler to him (the author) means keeping the status quo, even tho it fails to suggest the proper pronunciation. If a pupil is going to be taught to sound out words according to the pronunciation indicated by the spellings, then failure to indicate this, as in T.O., must cause failure on the part of the pupil to understand the proper pronunciation. And this will cause confusion. How can a pupil be taught pronunciation when many words only partly indicate pronunciation? Which is more important - that a beginner learn the derivation of words thru morphophonemic indicators, or that he learn how to pronounce words properly?

Morphophonemic indicators, such as the silent l in could, etc., are not helpful to learners. They are not helpful to teachers either. They are harmful because they interfere with teaching by phonics, and they are not useful to the teacher in teaching morphophonemic relationships because they are unreliable in this extent. Also a reform along morphophonemic lines would mean very little reform. And this kind of a reform would still be so unphonetic that it would be no help at all to the beginner. Actually it would still offer all the confusing anomalies now handicapping learners. Only in the fourth or fifth grade could such rules be of the slightest help - too late for learning reading - if indeed it was any help at all.


[Spelling Progress Bulletin, Winter 1979, p20]
[Kenneth Ives: see Journals, Newsletters, Anthology, Bulletins, book Written Dialects.]
[Helen Bonnema Bisgard: see Journals, Anthology, Bulletins.]

Book Review, by Helen Bonnema Bisgard, Ed.D.

Kenneth H. Ives, Written Dialects n Spelling Reforms, History n Alternatives, Progressive Publisher, Chicago, Il, 112 pp., $5.00 (to readers of SPB, $4.00)

This paperback book has a neat, pleasing format with an inviting preface and table of contents. Any student of education or linguistics who is researching the history and future of the spelling reform movement will find the book informative and provocative. The reader becomes involved immediately as a participant in an experiment. He sees that two short, common words the and and are represented by [n] and [ħ] thruout, [1] and is directed to note his first impressions upon noticing these wordsigns, and later his final reaction after finishing the last page. From his own feelings he can deduce the readiness with which the public might shift gradually from traditional orthography to a streamlined system.

The opening chapter reviews the history of English spelling thru six phases which are recapped in a convenient table, and delineated more fully in four pages of the appendix "Chronology of periods, authors, events." The "Resources for Further Reading" beckons to a score of works published within the present decade, as well as to an equal number of books of long-standing reputation.

Having approved the above mentioned excellent features, it is with reluctance that this reviewer makes the following criticism. In the title and thruout the book the way he uses the term dialect leads to confusion. Thru the ages, since the Greek use of dia (between) legein (to speak), the word has meant language transmitted only orally, and has never referred to an alphabet written to show the sounds of such speech. A dialect possesses not only its own sound system but also its distinctive vocabulary and features of sentence formation. However, the author applies the term dialect to World English Spelling, Gregg shorthand, Pitman's Initial Teaching Alphabet, Sensubul Spelling, and many other systems which do not possess their own phonology and grammar. He overlooks the fact that a sentence written in World English Spelling (WES) is pronounced exactly the same as one written in Gregg Shorthand. Even tho author Kenneth Ives designates these systems as written or eye dialects, the fact remains that they are merely alphabetic scripts used to graphically represent human speech. The distinction between dialect and writing system is important, for some people believe that their language (or dialect) will be changed if the spelling is reformed. They need to be assured that only the graphic symbols representing the sounds will be altered.

After the author describes a number of what he calls "written dialects," he recommends "Economy Spelling" because it changes common words by removing redundant letters, thereby saving space, and because it is to be put into use by 30 gradual steps. Altho this reviewer does not agree that the plan would be superior to a complete one-step reform, she does acknowledge that the pro's and con's of Ives' recommendations should be considered by the serious student who is planning a strategy for eliminating resistance to spelling reform.

[1] A new reader is unable to understand the contraction n. Is it an, in, on, un, or and?

Back to the top.