[Journal of the Simplified Spelling Society, 1989-2 pp.4-8 later designated J11]
[Donald Scragg: see Journals, Bulletin.]

English spelling and its Reform:
Some Observations from a Historical Perspective

Donald G. Scragg.

We here present the inaugural address of the Society's new President, given at a meeting of the Society on 24 September 1988. A profile of Donald Scragg can be found on p.4 of Journal 1988/2 J8.

1. Role of President, scope of address.

Some fifteen years ago, in A history of English spelling [1], I said that the Simplified Spelling Society and indeed the spelling reform movement generally have had periodic bursts of energy. I think in future decades we will look back to the 1980s for one of these. Some excellent new work is being produced today, much of it recorded in the Society's Journal, and I look forward to reading more of it during my term of office as President.

I will not pretend to you, however, that this paper offers new insights or new information comparable with the best of the research being undertaken at present, nor does it offer an exciting new synthesis of the progress in recent years. I have never had an active role in the promoting of reform, and I do not see it as my function as President of the Society either to propose or to direct a reform campaign, since constitutionally that is the job of the Chairman and his committee; nor do I wish to enter into controversy about the relative value of one reform scheme over another, although I shall always be happy to comment in detail on any proposals that come before the Society. I speak today not as a practical reformer, nor on the subject of a practical reform; rather, I would like to put reform into a slightly wider context and look at some aspects of the nature of the written language, and to suggest ways in which these might influence the thinking of the practical reformer. In particular, I shall be looking at them from a historical perspective, since that has always been my special interest.

2. Speech and writing.

I shall begin by making the less than profound but very necessary observation that speech and writing are two independent forms of communication. We were all once able to communicate in speech without having any knowledge of writing, and although some people never learn to read, their ability to speak is in no way limited because of this. Similarly it is possible to learn a foreign written language without having any knowledge of its spoken form. In practice, however, since most literate people have command of the spoken form of their language as well, it is convenient and economic to have links between the two. How close the links are depends on the language and the history of its written manifestation.

The development of a written language is always secondary to that of a spoken language, both in the general sense that writing is a relatively late development in human society, and in the particular sense that most people learn to speak long before they learn to read. The universal link between speech and writing is on the level of the word, in that written languages generally have a representation either of a whole word or of a segment of a word, either a syllable or a sound. In languages using the common European alphabet, historically the link is at the level of sounds, since the alphabet, as its name implies, provides a symbol for each sound in the language system. English, using the roman variety of the common European alphabet, has the sound-symbol match as the underlying principle of its writing system, as may be seen from the fact that literate speakers of English when faced with having to pronounce a written word which is new to them, most usually a name, resort to "spelling it out", or assuming that its spelling is broadly phonetic, however much experience they may have had of what may be called the Leicester-Arkansas category.

3. Evolution of the alphabet.

Most authorities now agree that the common European alphabet stems from the North Semitic alphabet in use in the second millennium B.C., and that this in turn drew on the pictograms and ideograms of Egyptian hieroglyphics. For example, the pictogram of the head of an ox, stylized as an arc with two prominent horns, and that of a house, stylized in ground-plan as two conjoined squares, were adapted as the first two letters of the alphabet, and given names signifying their origin: aleph is Hebrew for an ox and beth for a house. That great trading nation the Phoenicians seems to have been responsible for spreading the alphabet around the eastern Mediterranean, where it was picked up by the Greeks, who developed it by the introduction of regular representation of vowel sounds. (It is sometimes maintained that only after its modification by the Greeks did the Semitic syllabary truly become an alphabet.) At some point in the tradition, writing shifted from vertical lines to horizontal ones, and the letters were tipped onto their sides, so the ox's horns moved from the top of the character to the right side, and the conjoined rooms of the ground-plan of the house were set one below the other rather than side by side. With the move from the stylus to the pen and from clay tablets to skins, we find a more cursive script in which ox-head and houseplan become the Greek letters alpha and beta. Of all the many and varied writing systems developed by Semitic tribes and those who imitated them a thousand and more years before Christ, the Greek alphabet proved to be the most flexible and efficient for Indo-European speakers. Its use spread to Italy, where it was later adopted and adapted by the Romans, and it was carried by them - and especially by Christianity - from Italy to the world. It was the Romans, incidentally, who coined the term alphabet.

4. Adapting the alphabet to English.

Christianity introduced the roman alphabet to the Anglo-Saxons at the end of the sixth century A.D., and there has been a continuous tradition of written English in that alphabet from then until now. The principal reason for the complexity of English spelling today is the very full and largely unbroken tradition of recording legal, historical and literary texts in the language over this long period. [2] English of the sixth century was represented in letters roughly matched with sounds, in much the same way that missionaries of the nineteenth and twentieth centuries recorded the languages they came into contact with according to broadly phonetic principles. Since sixth-century English contained some sounds not found in Latin, a few new characters were needed. A <d> with a cross-stroke through the ascender, for example, was the earliest attempt at recording the dental sounds of this and bath. Almost all the alterations proved marginal or temporary, however, and failed to make any serious impact on the history of writing. Crossed <d> for example survives only in the International Phonetic Alphabet and in Icelandic. But another sound which gave difficulty was the bilabial semi-vowel of was and which, for which Latin provided only the ambiguous <u/v>. Early Anglo-Saxon scribes doubled the Latin symbol, and when these were ligatured some centuries later, as part of a series of modifications to the script which took effect around the twelfth century, the letter <w> was created, the one alphabetic innovation which English has given the world.

Many more signs and symbols than are generally realised originated in some manipulation of alphabetic symbols. For instance, the practice of grading student essays with Greek alphabetical symbols, <α> for first class, <β> for second class, and <γ> for third class, which still operates sporadically in universities, began in the Renaissance when schoolmasters at all levels adopted the practice as part of the emphasis laid on the classics in education generally. The first and last of the symbols were soon equated with good and bad, right and wrong, and written with an ever greater degree of freedom, so that <α>, which was written from the point of the bottom horn, round the arc and out through the upper horn, gradually had the lower horn attenuated and the upper one extended, until it became a tick (<√>), and <γ>, with the bottom loop widened and then split, became a cross (<x>). Children today are introduced to Greek before they - or their teachers - know it.

The development of the English alphabet is a fascinating topic in its own right, and one of considerable moment to the reform movement. One of the projects that the Centre for Anglo-Saxon Studies at Manchester is working on at the moment is the mounting of an exhibition of varieties of script used up to the Norman Conquest in all forms of writing, on parchment, wood, bone, stone and metal. This would include examples of the runic alphabet, a variety of the common European alphabet used by all the Germanic tribes while they were still pagan, and introduced into England first by the Anglo-Saxons and later by Danish and Norse Vikings, as well as use of the roman alphabet in both Latin and English writings produced in England. The aim would be to show the development of distinctively English styles of writing by comparison with continental use, and the hope is that this would be the first of a series of exhibitions illustrating changes in script down to the present day. Some such investigation is a necessary prolegomenon to the compilation of the complete history of English spelling which seems to me to be one of the most important tasks facing the historian of the English language today.

5. Different qualities of speech and writing.

To talk of varieties of English script, even to introduce the question of upper and lower case letters, is to acknowledge that English has moved far from the phonetic principles at work in the sixth century. But without developing this theme fully, I would like to add a general word on the necessity of a link between speech and writing at all. Writing, as a system of communication independent of speech, was invented to serve different purposes, and it still does. Speech, leaving aside as relatively peripheral such modern developments as mechanical recording, storage and transmission of sound, is a means of immediate communication between physically contiguous individuals. The segmental sounds of speech, whether they be the three successive sounds which the layman easily perceives in cat or the less easily perceived three in thought, are placed within a framework of suprasegmental pauses (between groups of sounds - which may or may not equate with words - and between sense-groups of words) and within a similarly suprasegmental framework of stress and pitch patterns. And speech is usually accompanied by facial expression and gesture - body language without which radio and telephone communication is that much more inefficient.

Writing serves purposes very different from those of speech. It communicates across time and space, and however close a match it has with speech in terms of sound - symbol relationship -however phonetic a writing system is - there can be no equivalent of the body language which accompanies speech, and writing is therefore limited in communication value in the same way that radio and the telephone are. There are other important differences: writing has only a very imperfect representation of the suprasegmental phonemes, but has evolved other means of representing syntax. Word-division is marked more clearly in writing than in speech, and the division of words into syntactic units (phrases, clauses and sentences in conventional terminology), which is indicated in speech by a combination of stress, pause and the rise and fall of pitch, is represented much more uniformly in the written language by means of marks of punctuation. In other words, although there is some very vague parallel between punctuation on the one hand and variation in pitch and sentence-stress on the other (because each serves the same function ultimately, that of communicating syntactic information), there is no real attempt at a suprasegmental sound-symbol match.

A spelling reform which in any sense aims at a closer match of spelling and pronunciation is usually concerned only with segmental sounds. But we need to be clear about the different functions and modes of operation of these two systems of communication, speech and writing. Speech in conjunction with body language is a more flexible system than writing, but is by its nature it is ephemeral. Writing has the advantage of permanence and of open-ended recapitulation, but lacks the accompanying body language, and has, as a built-in compensation for its failure to reproduce the variations of stress and pitch of speech, a different set of syntactic markers. Hence, for example, it is easy to define a sentence in writing - generally, it begins with a capital letter and ends with a full stop; but the definition of a sentence in speech is much more difficult, and has led in the past to an attempt by those instructing the young to impose the conventions of the written language onto the spoken language, to try to encourage an improvement in oral communication by imitation of the patterns of written communication. In some cases such improvement is possible, but not in all, because, as my message so far has proclaimed, speech and writing are independent forms of communication, and their linking is a convenience, not a necessary condition of comprehension.

6. Abbreviations.

Another feature of written language which can be overlooked is its frequent use not of words in the conventional sense but of cyphers. In the Middle Ages, writers of Latin had a very wide vocabulary of abbreviations designed to facilitate the copying of books when this had to be done by hand. Today we use abbreviations as much to increase reading speed as to facilitate writing, but their variety is no less great, from the more obvious ampersand, Mr, Mrs, Co. Ltd, £, $ etc, to the numerals and mathematical symbols. We notice such forms only when they depart radically from what we expect, as, e.g. when we attempt to drive on the opposite side of the Atlantic from the one we are used to and find that a road sign abbreviation for south is sth or simply s in Britain but so in North America. But all these abbreviations have their counterpart in whole words, and the literate reader copes with them as easily as he does with variations in letter shape or with the two alphabets in upper and lower case.

7. Increased rigidity of written English.

Finally, in considering the characteristics of written English, one must note the increasing rigidity of the system during the last two centuries. Undoubtedly individual words have continued to develop in spelling in this period, probably many more than most people realise, as I indicated in A history of English spelling some years ago. But the general trend of the nineteenth and twentieth centuries has been towards ever greater stability within the system. This is not a movement which coincides with the spread of literacy, for the literacy rate of seventeenth-century England was higher than that of the nineteenth century, but it is linked, I believe, with the increase in availability of printed material, particularly books.

Nowadays, we all accept a relatively wide variation in the pronunciation of individual words, not only because of regional or social dialects but in contextual variation caused by such factors as differences of stress patterns. It is for reasons such as these that any attempt to make spelling strictly phonetic has now largely been abandoned. Why, when variety is so fashionable in the spoken language, is such store laid on lack of variation in the written language? The simple immediate answer is speed of comprehension. If a given word always has exactly the same written form, recognition, it is argued, is faster, and so is reading speed.

8. Reading psychology and spelling reform.

However the human mind is a very complex organ, as the operation of language itself amply illustrates. In spoken communication, an audience is constantly faced with sentences that it has never before heard, and understanding is achieved by recognition of the potential meaning of individual words coupled with recognition of the grammatical structure in which they occur. In written language too, the mind focuses not on individual isolated symbols (letters) but on sequences of letters (words) and sees them in the context of other words. The process, for the competent reader, is a fast one, involving recognition of the whole shape of a word rather than its constituent parts. That is why proof-reading a book is difficult - it is hard to force yourself to read letter by letter for compositors' errors - and indeed a great many printing errors (e.g. assimlation, asssasination) may never be noticed because few people read words segmentally, letter by letter. The eye takes in the outline, and writing has become hieroglyphic again, with words not now being pictograms like the ox - head but ideograms, stylized pictures representing ideas.

The fact that literate people read in this way must be taken into account in a spelling reform proposal. It has relevance, I think, in the instance of highly irregular and unpredictable words like those involving <gh>. I assume that by any statistical count, the word sight is much less frequently misspelt than the word separate, simply because the word picture is both easily recognized and easily reproduced. The difficulty is that spelling reform takes as its starting point the convenience - felt particularly by learners or those experiencing learning difficulty - of the link between speech and writing. Whatever differences of surface-structure there may be in the two systems, they have the same deep-structure, and since virtually everyone learning to read and write has already acquired a spoken language, it would appear that the closer writing is made to match that spoken language, the sooner fluency in it will be achieved.

As far as English is concerned, this has been the position, explicit or implicit, of spelling-reformers for over 400 years, ever since the first practical suggestions for reforming spelling were published in 1569. It is a perfectly logical premise, as long as the function of the reform is recognized. A reform that aims at assisting learners very reasonably attempts to build on the known of the spoken language. But what I have been saying must call into question the usefulness of developing a general principle of reform on the practicalities either of basic teaching of reading or of remedial teaching in writing. Should not a general reform take account of the wider issues of writing as an independent means of communication?

9. New Spelling.

Let me take as a basis for discussion the system of reform officially proposed and supported by the Society, New Spelling, which was created almost fifty years ago and which has been modified in only minor respects since. Without wishing to engage in any debate about its appropriateness to conditions prevailing today, I would argue as a historian that New Spelling represents the most significant advance in the philosophy and practice of reform during the twentieth century, and arguably of the whole 400-year history of the movement. It is a scheme with a sound academic basis which has been worked out most carefully over many years by a series of dedicated and learned scholars. Also, because it was adopted by Sir James Pitman for i.t.a., it is the most successful reformed system there has been in that more people have been exposed to its recommendations than to those of any other reformed spelling, even if they used them only in conjunction with a transition script. Its guiding principles are those which have been reiterated most frequently by reformers over the centuries, and summarising them will help to identify the traditional concerns of those advocating reform.

Firstly, the roman alphabet as traditionally used is retained (minus 'unnecessary' letters such as <q> and <x>) and no new diacritics are introduced. Secondly, by the principle of least disturbance, current usage is retained wherever possible, and new combinations of letters are excluded as far as possible (although exceptions are made in the difficult case of vowel representation). Thirdly, the most fundamental principle is that of regularity: each letter or combination of letters is self-contained (in other words, no double consonants or final unpronounced <e> to indicate the quality or quantity of a preceding vowel) and each has a match in the sound system, so that, for example, spelling may be deduced from pronunciation and pronunciation from spelling. The scheme has all the advantages sought by generations of reformers, in regularity of representation of, and closer match with, the spoken language (although, in conformity with twentieth-century linguistic thinking it avoids close phonetic representation as neither feasible nor desirable), and it meets many of the familiar objections to reform in that it retains links with traditional orthography [= TO] as widely as is consistent with its principles.

10. Imperceptible change.

To take those principles in turn, there can be little doubt that a reform proposal that seeks to move too far from TO will have little chance of success. Four centuries of effort have proved that. But it is worth taking a moment to consider how far the rigidity of the written language extends. For those of us who are already practised readers, the written language appears so conventional that we are no longer aware of, for example, the sequential arrangement of words, lines and pages. No doubt we would all be greatly disturbed by printing which reverted to the practice of the earliest uses of the alphabet and arranged words in vertical columns rather than horizontal, but how much alteration can be effected before the average reader becomes upset?

From the historical perspective, printed English has changed more than might be supposed over the last few centuries. Up to the eighteenth century it was usual to print at the foot of a page a word or two anticipating the opening words of the next page (catch words) but these have now totally disappeared. It may be hard to see what relevance this might have to spelling reform, but in fact it is a significant change, in relatively recent times, in our expectations of the presentation of written material. Others of a comparable nature have occurred much more recently. In the twentieth century, particularly since the paperback revolution, many books have been printed without running titles, the page headings which remind the reader of the title of the book or chapter. And in the last twenty years right-hand justification, or straight right-hand margin, has often been abandoned, while the placing of the page number, traditionally top left in left-hand pages and top right in right-hand ones, has in some presses become uniformly top left.

Changes in letter shapes are often even more difficult to detect. People concerned with commercial book and newspaper production have been creating new typefaces and printing styles since Gutenberg and Caxton. Significant and wholesale changes, such as the abandoning of long <s> (<ʃ>), are rare, but alterations less easily perceived do nevertheless take place frequently. The most widespread change in recent years is the decline of ligatured letters. Ligatures (which initially imitated those of medieval manuscripts) have always been used in printing, sometimes quite widely, but today they are almost wholly confined to <f>, still sometimes physically joined to a following <i>, an <l>, or perhaps another <f>.

And finally there is the use of upper case letters, which in the eighteenth century heralded the majority of nouns (as they still do extensively in German) but today are confined to proper nouns or names. Even here practice varies, but the tendency is for capitals to be used ever more sparingly. Most noticeable are nouns and pronouns referring to the deity, which until recently would invariably be capitalised where now the pronouns in particular are not. This is not so much a feature of the secularisation of our age, I suggest; rather it relates to the prevailing view that initial capitals are a nuisance which we can well do without.

You may here glimpse at last the trend of my argument. The written language is not as static as is sometimes assumed, and the reading public has accepted without any noticeable outcry a series of changes, particularly those which publishers have deemed to be commercially desirable. There is hope here for the reformer. The concern shown by the creators of New Spelling for keeping the closest possible links with TO might be modified: the reformer should be ready to capitalize on those features of TO which show signs of alteration anyway.

11. Computers and spelling reform.

Some of the changes in the presentation of written material that I have just been describing are the result of the introduction of computers to production processes, and it is worth digressing for a moment to look at the impact of computers on the written language more widely. At first it appeared that computers might work to the advantage of the spelling reformer, since at the press of a button a global search and replace procedure might convert the written form of any given word throughout a large body of material. But the advent of the spelling checker, a computerized word-list which will highlight any aberrant spellings for those anxious not to have any departure from TO, seems likely to fix even more firmly total adherence to the artificial norm created by the lexicographers over the last three centuries. On the other hand, because computer software is often international, a spelling checker produced in America or in the Far East will take account of the greater variation which exists in spelling in U.S. or World English than in British English. Spellings such as program and disk, already widespread because of the computer industry, may spread beyond the reach of the ultra-conservative advocate of TO.

12. Avoidance of diacritics.

This brings me to a second principle adopted in New Spelling: the avoidance of any new diacritics. English is uniquely blessed among major Western languages with a relative freedom from extra-literal characters above and below the line. The one we have, the apostrophe, gives nothing but trouble. In historical terms it is quite a newcomer, not being fully established as we use it today until well into the nineteenth century, and though it may well now be in its death throes, these are proving unnecessarily long drawn out. Euthanasia is called for.

13. Regularity and flexibility.

But the major principle of New Spelling is that of regularity. Regularity is seen as the key to easier learning, and to the resolution of difficulties in either spelling or pronunciation in those who are generally competent in written English. I have shown already that there are difficulties attendant on the basing of a reform on the closer match of sound and symbol. There is, I believe, a particular problem with New Spelling in its retention of the desire to use spelling as a guide to pronunciation, a concern of reformers which goes back as far as the sixteenth century. In the later twentieth, this seems to me quite outmoded.

But New Spelling, like all its predecessors, has failed in its main purpose of reforming spelling principally because of public resistance to change. A first step in the breaking down of that resistance might come through the dissemination of the concept of greater flexibility in spelling, and through the promoting of an awareness of how much variety already exists in the written language, for instance of letter forms and in the use of abbreviations. It is ironic that the single most tangible outcome of four hundred years of pressure for reformation of the English spelling system has been the creation of a highly inflexible public attitude to spelling. One of the concerns of the earliest reformers was to stabilize spelling, to reduce the variety still available in private, and to a less extent in public, spelling. By the mid eighteenth century they had succeeded. I think they were wrong.

14. Conclusion: progress through tolerance.

I am not, let me stress, arguing for anarchy in either public or private spelling. But I am arguing for a significant change of public attitude, so that minor variation - in double consonants, in unstressed suffixes such as <-ance, -ence>, in all those peripheral details which do not hinder communication even momentarily - should no longer be seen as in any way important. I can talk from the moral high ground here, because my spelling is relatively good, but I freely admit to occasional uncertainty. Very few of us, I suspect, can claim never to have recourse to a dictionary merely to check a spelling. Is this an appropriate use of time? Can we not convince the general public that obsession with the goal of what is seen as orthodox spelling at the expense of other aspects of language is misconceived? Personally I would rather see too many <e>s in separate than a failure to distinguish the meaning of disinterested from that of uninterested.

Ought we not to be opening a public debate on the true nature of written language, not as a system carved in stone and impervious to development and change, nor as a pale reflection of its older sister, speech, but as an independent vehicle of communication, just as effective when allowed scope for variation as speech has proved to be? Let me offer you the proposition that the worst public speller is the greengrocer, whose colourful permutations on broccoli and lichee are matched only by less comprehensible ones on asparagus and lettuce. Yet we continue to buy his produce. I would like to see this tolerance extended much more widely. Over the last twenty years, we have seen in Britain the encouragement of regional and class dialects, and a general rejection of the proposition that only those who can speak with a Received Standard voice are worth listening to. Is it too much for the historian to hope that by the twenty-first century, strict adherence to the fixed spelling of the printing houses is not seen as a necessary condition of literacy?


 [1] Donald G Scragg A history of English spelling. Manchester University Press, Mont Follick series Volume 3, 1974.

 [2] Many problems. of course, go even further back, and relate to the creation of the alphabet itself. Cf. the judgement of Ernst Pulgram, in 'The typologies of writing-systems', Writing without letters, ed W Haas, Manchester University Press, Mont Follick series Volume 4, 1976, pp. 1-28, esp. p. 24: "It is worth noting that there did not occur at any point in the history of writing a revolutionary step, a scrapping of the old and an entirely new beginning, a break that had to do with the invention of an altogether new system. It is all a matter of gradual evolution. If man is an inventive animal, he certainly has not proven himself a scintillating innovator in conceiving ways of writing: nothing new has in fact been added since the adoption of the alphabetic system about three thousand years ago."

Back to the top.