[Journal of the Simplified Spelling Society, 1988/2, p20 later designated J8]
[Also on this page: SSS 1984 proposals, 5 questions on Cut Spelling.
The Implications of Spelling Reform for Skilled Readers.
John S Kerr.
Dr Kerr is a cognitive psychologist by training and a researcher in various fields by necessity. At present completing a project on the optimum design of traffic signals at Aston University in Birmingham, he will shortly be taking up the post of research fellow in the Human Psychopharmacology unit at Leeds University. The following contains some of the ideas he presented at the Society's Fifth International Conference in July 1987.
Introduction.Simplified spelling is not of obvious benefit to the reader already skilled at using traditional orthography. In fact it must be expected to be detrimental, at least initially. What might be the effects of reformed spelling on the reading process? This short article consists of the initial reactions of a psycholinguist to the implications of simplified spelling. Some of the ideas presented here are hypotheses rather than facts, and remain to be tested empirically.
Concepts, not letters.Most of the time spent during reading is taken up by the processes involved in understanding the text rather than simply decoding the symbols: cognition rather than perception. This is the case with text which presents both simple and complex ideas. In one common view of the reading process, the readers create a mental model of what the text is about from their own knowledge and experience, and use this model in conjunction with the information contained in the text to build an accurate representation of the discourse. This is rather slow compared with the tasks 'downline' involved in recognising the actual words. Alternatively the 'autonomy' position argues that the reading process consists of discrete operations: recognising patterns, retrieving meanings, parsing, integrating and understanding: again, decoding the symbols and recognising words is only a small, if essential, part of the process.
Word recognition can occur with or without phonological mediation (turning the word into its sound). Skilled readers will tend to by-pass this stage whereas learners and poor readers can be seen to be 'sounding out' the words, even using sub-vocal speech (the reason why Sun readers' lips are said to move). When confronted with new or lengthy words, skilled readers will revert to this strategy and use the grapheme to phoneme conversion rules (although this procedure will not necessarily yield the correct answer - a fact that is the raison d'être of the Simplified Spelling Society). This effect will probably account for most of any detriment that readers have initially with revised spelling; they will be unable to use the faster access mode. Disruption of smooth reading will occur with new forms which look like old ones e.g. the Cut Spelling form add (added). Other CS forms may present problems in that they become very short (e.g. qy for quay) and can therefore be 'missed', although this will be countered to an extent related to their importance in the text.
Changing the spelling will have different effects depending on where the change occurs: certain parts of words, notably the beginnings and ends, are more important than others. In general changes here are more detrimental to reading than alterations of medial letters. An implication of this is that information from peripheral vision will not contribute: readers use information to the right of what they are looking at to 'prime' the upcoming words, so that when seen they are already part processed. The information is mostly based on the shape of the words and the initial and final letters. The changes in word shape themselves may be disruptive. Any differences in reading speed that these effects cause however will be small compared with the process of conceptual understanding, which will not change with spelling: a rose is a roze is a rohz. Conversely readers of a system like CS which economises on letters may not read faster, for the same reasons.
Polysemy & Context.When spelling is simplified, there will probably be an increase in the number of words with two or more meanings: words which sound alike (presumably in a standard pronunciation) would be spelt alike in any phonographic system (with exceptions for special cases perhaps). This will result in an increase in lexical ambiguity, though this will not be a problem, at least for skilled readers, since polysemous words already abound in English e.g. rose has over a hundred distinct meanings.
One reason why lexical ambiguity is not a problem is the way that context influences the interpretation of words at a number of levels, even in unstructured lists such as knitters, seamstresses & sewers versus drainpipes, gutters & sewers. The effect of information contained in the text and in the reader's memory about what to expect in the discourse can be very constraining, and is a major aspect of understanding written language. The skier was buried by the sudden... raspberry or avalanche? In fact readers are rarely aware of the alternative possibilities of what they are reading. The Smiths saw the Rocky Mountains flying to California is straightforward until it is pointed out that the sentence could be part of a science fiction story about aliens rearranging the geography of North America using anti-gravity machines. (Note also the assumption that the Smiths were flying in an aeroplane and not by flapping their arms.) Language is rarely used without some context, and context will rarely fail to disambiguate the language.
Conclusion.Revised spelling (depending on the nature of the revisions) will have little effect on the reader who is already familiar with traditional orthography: a conclusion which is supported empirically by some of the work of Valerie Yule. It is not yet clear whether reformed spelling will confer any advantage on the reader who becomes familiar with it. The advantages of simplified spelling are more clearly in language learning, with certain systems also economising on production and storage.
Clark, H H & Clark E V (1977) Psychology and Language: An Introduction to Psycholinguistics, New York: Harcourt Brace Jovanovich
Garnham A (1985) Psycholinguistics: Central Topics, London: Methuen
Johnson-Laird P N (1983) Mental Models: Towards a Cognitive Science of Language, Inference and Consciousness, Cambridge: CUP
Kennedy R A (1985) The Psychology of Reading, London: Methuen
Sanford A I & Garrod S C (1981) Understanding Written L anguage: Explorations in Comprehensionbeyond the Sentence, London: Wiley
[Journal of the Simplified Spelling Society, 1988/2, p32 later designated J8]
[Stanley Gibbs: see Journals, Newsletters, Tough, Though, Thought leaflet.]
The Society's 1984 Proposals.
Following his note in the 1988/1 Journal(p.33) concerning amendments agreed by the Society in 1971-72 to the 1948 and 1956 versions of Nue Speling, Stanley Gibbs now outlines the 5 reform proposals approved in 1984.The Minutes of the 1984 AGM stated the following:
"Stage 1 shall consist of the five reforms as printed in the November 1983 Newsletter, page 4. Stage 1 shall be the approved house-style for the Society and members are encouraged to use it within the Society and, where possible in their own privat correspondence. That the five proposals listed now above shall be adopted as the Society's policy. This set of proposals shall be named Stage 1."
The Society issued an introductory leaflet entitled Tough Though Thought, listing the Stage 1 proposals as follows:
SR1 developed by Harry Lindgren in Australia, SR1 (Spelling Reform One) calls for the sound 'e' as in bet to be spelt with 'e'. Hence: eny meny frend alredy ses tred jelous hed.
SR:ph This is probably the least controversial of all reforms, the change of 'ph' to 'f' when it is sounded as 'f'. Hence: foto telefone fysical elefant safire.
SR in the:augh Words with 'gh' include some of the most absurd spellings in English. In this reform 'augh' is changed in one of two ways:
1. Delete 'gh' when there is the sound of 'au' as in caught. Hence: cant fraut dauter nauty
2. Replace 'ugh' with 'f' when there is the sound of 'f'. Hence: laf draft.
SR:ough There ar so meny different pronunciations of 'ough' that it is changed in one of fivedifferent ways, depending on the word:
1. Delete 'gh' when there is the sound of 'ou' as in bough. Hence: bou drout plou.
2. Change 'ough' as in bought to 'au'. Hence: baut aut thaut faut saut.
3. Change 'ough' to 'of' or 'uf' (depending on the pronunciation). Hence: cof trof enuf tuf.
4. Cut back 'ough' to 'o' (or 'oe') as in though. Hence: tho altho (but doh for dough and thurra for thorough).
5. Change 'ough' to 'u'.Hence: thru
SR:DUE Meny words end with an 'e' that is not only useless but misleading. This is corrected with SR:DUE (Spelling Reform: Drop Useless E's). The situation arises when the preceding vowel is short and it includes meny common words. Hence ar hav wer serv giv liv opposit negativ massiv activ involv curv (but not the adjectiv live)."
Wer serv curv aut not to be there; as the Minutes clearly state, April 1984: short vowelled syllabls. Obviously we erred and broke our own rules. There have been no changes. Wer, ar, serv, curv, nurs although quite satisfactory, could not be included.
[Journal of the Simplified Spelling Society, 1988/2, p32 later designated J8]
Traugott Rohner asks Chris Upward 5 questions about Cut Spelling.
Traugott Rohner, a new American member of the Society, raises the following queries about the spelling-patterns used in Cut Spelling, to which the editor attempts to reply.TR: Recently the Society has worked on a revision of English spelling called Cut Spelling. I am anxious that it will succeed, but must point out a few of its weaknesses.
1. Its greatest weakness is that it is trying to change too many flags. CS recommends th for the and ar for are. If these turn readers against CS, its advantages will be lost. These two words are among the most used in the language. They should not be changed.
CU: Ther ar two kinds of change: one substitutes letrs, th othr merely omits them; th latr is far less disturbing. Th form 'th' is a gret econmy and removes th misleading paralel with 'he', etc; but it is not vital. 'Ar' for 'are' is mor fundarnentl: it removes th misleading paralel with 'bare' and creates a tru paralel with 'bar', etc.
TR: 2. P. 25-27 of the 1987 No.3 Journal presents 4 pages all in CS, showing both its advantages and disadvantages. It begins "A study of patrns of misprint...". What is patrns? Probably patterns. Could it not also be patrons?
CU: In CS vowl-letrs ar normly short unless othrwise indicated, and a foloing vowl-letr ofn indicates th long valu in CS, as in TO. So th foloing <a> in :fatal' tels us th first <a> is long (contrast 'catl'). 'Patrn' and 'patron' sho th same difrnce. A secnd reasn wy 'patron' keeps its <o> is fonografotactic, and paralels th difrnce between 'barren/ barn', 'modrn/children'. Th avraj user remains unaware of these sutltis, but gets used to th distinction in practice.
TR: 3. Now how does one pronounce sujests? According to Webster ther is a g befor th j.
CU: Webster has th /g/ as optionl; neithr Oxford nor Collins sho it at al. Speakrs cud rIte 'sugjest' if they wishd.
TR. 4. Two words have 3 vowels but only one is written, promnnt for prominent, and difrnt for different.
CU: Th vowl-sounds in 'prominent' can also be spelt as in 'dominant, permanent, consonant' and with sylabic <n> in, 'hadn't'. CS harmnises al these variations by consistntly riting sylabic <n>: 'domnnt, permnnt, consnnt, promnnt'. In norml speech 'different' has only two vowl-sounds; th first <e> is silent. 'Difrnt' is a suficient and unambiguus representation of th sound, maching 'cormrnt, ignrnt'.
TR: 5. A more glaring mistake takes place when all the words with <er> are spelled with just <r>. We are strongly against swallowing vowels even tho they are unaccented. As long as they are spoken, they should be spelled out.
CU: Only sounds can be swalod, but CS dos not afect th sound. TO has many alternativ spelings for that ending, as in 'burglar, teacher, doctor, neighbour, murmur, injure, martyr' etc, wile th <-er> ending can also hav a quite difrnt pronunciation as in 'defer', and confusion results for both riters and readrs. Th sylabic <r> in 'acre, centre' provides th ansr, by shoing that no vowl letr is needd at al; so we get 'burglr, teachr, doctr, neibr, murmr, martr' - omiting th unpredictbl vowl-letr acheves regularity and econmy at a stroke; but th sound remains intact.
Back to the top.