A Survey of English Spelling

Edward Carney

• Publisher: Taylor & Francis
• Year: 2012
• Print ISBN: 9780415092708
• eBook ISBN: 9781134872237
• Edition: 1
• Topic: Languages & Linguistics
• Subtopic: Linguistics
• Index: Languages & Linguistics

Published at a time when literacy and spelling are issues of topical concern, A Survey of English Spelling offers an authoritative, comprehensive, and up-to-date overview of this important but hitherto neglected area of the English language.
The text brings together a vast body of knowledge, both synthesised from diverse sources and original, unpublished research. The emphasis is on a functional exploration of the spelling regularities and markers that underpin literacy in English.
An extensive database has been used throughout to provide a wealth of examples, statistics and analyses. The carefully signposted text and detailed contents listing allow students, professionals, teachers and academics in all areas of English Language, Linguistics and Speech Pathology to access specific information with ease.

Chapter 1

Alternative approaches to describing English spelling

1.1 A PHILOLOGICAL APPROACH

How did the present system come about?

English spelling, like most of our institutions, has a history. So, in matters of spelling, the past can help us to explain the present. We can see that loan-words acquired from particular languages at particular times have brought with them their own spelling conventions. The words village and entourage are both borrowed into English from French. Village was borrowed in the Middle Ages, but entourage seems to have been first used in written English in the 1830s. The final French stress has shifted to the front in the early borrowings, making them more like other English words and the <-age> of village has become an unstressed /-Iʤ/. In a late borrowing such as entourage the <-age> is usually unchanged as /-’ɒ:ʒ/. Studying the changing relationships between written and spoken English through time is the business of philology.

Philology studies language as part of cultural history. Its viewpoint is diachronic, looking at language evolution, showing how the present-day state of the language has developed over the centuries. A study of English spelling that is primarily philological, such as Cummings (1988), will consequently focus on the writing system as a transmission of culture. This will include material which, though interesting in itself, need have no relevance at all to the purely synchronic problems of achieving literacy in the language as it is today.

The philological approach can dig out an interesting antiquarian diversity often unsuspected by the common reader. Cummings (1988: 332), for instance, points out that the double <bb> spellings of chubby and shabby come before a word boundary: ‘apparently the adjective came from chub, a fattish fish’; ‘shab reflects an obsolete word used to refer to a skin disease of sheep and to a low fellow’. But for all practical spelling purposes, chubby and shabby are to be treated as simple forms on a par with the nouns hobby, lobby, tabby or the adjectives silly, merry, happy. Similarly,it has no bearing on the problems of literacy to know that: ‘The only known Romance instance of /ú/ = oo is the monosyllabic rook “chess piece’”. The word did indeed come into English via French. Before that, along with the game of chess itself, it came into French by way of Arabic and beyond that from Persian. Yet, in spite of this chequered history, it is spelt by §Basic spelling conventions and is identical to rook ‘type of crow’, a purely Germanic word.

On the other hand, it would be useful to keep apart (unlike Cummings 1988: 356) instances of <cc>≡/k/ with §Latinate prefixes in words such as accumulate, occasion, occupy, from the <cc> of §ltalianate words such as peccadillo, staccato, toccata. The <cc> of accumulate has to be linked to the double letters found with other prefixes, as in affiliate, alleviate, annihilate, appreciate. The staccato-type words have the common feature of penultimate stress and a final vowel. These two groups of words with the spelling <cc>≡/k/ belong to different subsystems for practical spelling.

In a philological approach, the concept of ‘rule’ may be rather different from the kind of rules (or generalizations) that might underlie competence in spelling (see §2.8.1). For instance, Cummings has a ‘Stress Frontshift Rule’ (ibid.: 127), which simply reflects the fact that the final stress of French loan-words, as we saw in village, tended to shift to the first syllable. This has happened in cover, honest, gravel, lemon, model, refuge, river, scholar. The importance for present-day spellers in this group of words is that a short vowel followed by a single consonant then became stressed in a spelling context where you would expect a long vowel – cf. over, navel, demon, modal, polar. This stress shift was a ‘rule’ in that the historical process affected most of the early French loan-words. It is not a rule in any synchronic sense, since there are no criteria for identifying these early loanwords by their present-day pronunciation or any other marker of ‘Frenchness’. They are simply a group of exceptions. The consequences, too, are different for reader and writer: the writer must avoid putting a <CC> spelling after the short vowel (*<lemmon>); the reader must avoid ‘saying’ a long vowel */li:mәn/.

Framing such rules as active processes may be slightly misleading in a philological framework. This is so of the rules which account for the short vowel before the suffix <-ity> (ibid.: 112) and before the suffix <-ic> (ibid.: 115). The underlying form is taken to be the vowel of the base form. So the long vowels of sane and mime are ‘shortened’ in the derived forms sanity and mimic. The rule is explicitly framed as a process: ‘instances of /i/ = i in VCV strings whose head vowels have been shortened (sic) by the Suffix -ic Rule’, ‘head vowels have been shortened by the Suffix -ity Rule’ (ibid.: 224). Many of the words given as examples of the ‘shortening’ process did not have a long vowel to shorten, either in their etymology or in related forms in English: critic, monolithic, prolific, acidity, humidify, ability, logic, symbolic, topic. The process of ‘shortening’ means, of course, that the vowel ‘ends up as’ short. So, here the historical difference between the stressed vowels of mimic and topic is lost sight of by the descriptive device.

1.2 A FUNCTIONAL APPROACH

How does the present system work?

A philological approach is very much concerned with asking: How did these spellings come about? The present study is primarily concerned with the question: How does our spelling system work? The two approaches are not at odds; they are complementary.

The difference between the two approaches can be shown by comparing what they would say about one particular spelling problem: the difference between final <-o> in words such as bingo, fresco, limbo, and words with final <-ow> such as billow, farrow, follow. The philological approach would tell you that halo is one of some old adoptions from French and Latin, that photo is one of a number of back-formations or ‘clippings’, that motto is one of several similar loans from Italian, that mango is from Portuguese, and that there are several words, such as jumbo, of obscure origin. The <-ow> words, such as follow, got their <-ow> spelling in Middle English as a regularization of a number of Old English word endings: <we>, <rg>, <lg>, <rh> and <u>.

On the other hand, the present study is concerned with how spellers can distinguish between the two groups of words as they are today, so that they do not make errors such as *<sallo>, *<mangow> for sallow and mango. If we look at the structure of the two groups of words, we find a pattern which could be exploited to provide teaching material. None of the disyllabic <-ow> words has stress on the second syllable as does hello, or a long vowel in the first syllable as does photo, or medial consonants outside the range /d, n(d), l, r/, as does mango (see §3.3.2.5, pp.173f.). This approach to the description of spelling conventions is best described as functional.

Since exceptions do invite comment, it may be of occasional interest to mention the historical reason for an exception even in a strictly functional approach. For instance, pickerel has an exceptional ‘doubled’ <ck> in the third syllable from the end (see #D6 p.123). The word means ‘a young pike’, so there is a hidden boundary after the first syllable, which makes the word rather like shrubbery in structure, where the <CC> is regular. Frippery, gallery and scullery are similar exceptions to the doubling rules, compared with celery, misery, but to add that frippery comes from Old French fripe, meaning ‘rag’, would not help. The <frip(p)> has no claim to be an English morpheme. The reverse is the case in shabby and chubby (see p. 1), where there is a hidden boundary that need not be known to explain the doubling.

If we are to focus on the problems of literacy in English, we are not primarily concerned with how the writing system came to be in its present state. We shall try to look at the system as it functions today for a normal literate adult and uncover those regularities that appear to be exploitable by a competent speller. These will not usually be in the form of recallable and explicit ‘spelling rules’ to be chanted in moments of insecurity. They will often be letter patterns which correlate well with speech patterns, with types of word-formation or with the various subsystems of foreign loanwords. In trying to describe the system by exploring the awareness of an average literate adult, we shall not assume knowledge outside the adult’s competence as a speaker and reader of English.

Chapter 2

Literacy and English spelling: methods and problems

2.1 WRITING WITHOUT SPEECH

Can the written forms of English be described systematically without linking them to spoken forms?

It would be quite feasible, though far from easy, to describe the English writing system without referring to speech. Egyptian hieroglyphs and Chinese characters are examples of writing systems that were originally designed to encode meanings directly, without indicating what the spoken forms were. Even when a writing system is based on an alphabet, as is English, the written form of the language can be studied as an independent system of purely graphic signs. This is obviously true of those signs in the English writing system which are not made up of letters: written signs such as <$>, <%>, <+>. The ampersand <&> was originally a manuscript ligature of Latin <et> ‘and’, but for present English users, it is just a single squiggle. For ordinary silent reading, we do not need to know a way of saying these signs as /’dɒlә/, /pә’sent/, /plΛs/ or /ænd/. Indeed, in the case of <@> we probably do not know a way of ‘saying’ it. These signs, like numbers and scientific symbols, are international and independent of speech.

The writing system also includes punctuation signs such as <.,:; – ! ?> and indeed the space between words. These relate to speech, but only indirectly. Their main function is to show grammatical structure and its boundaries, and to some extent the attitude of the writer. We know that ‘Doesn’t she enjoy a game of bridge!’ is not expressing doubt, as it would if <?> were used. We know that ‘… bridge!’ will probably have a falling tone as an expression of certainty. But <!> can associate directly with the certainty; there is no need to imagine the speech.

Usually, of course, the signs of the English writing system are made up of letters strung together and we can make more or less regular connections between the letters of the spelling and the phonemes of the pronunciation. But we do not have to make this connection. Quite early in their schooling, competent readers acquire the skill of skimming over the written text without referring to the corresponding spoken forms. Indeed it comes as a surprise to learn later on, if they ever take a course in phonetics, that the single written form of a morpheme such as {-ed} or {photo} has several different contextual pronunciations in their speech.

In describing such complex written signs we are quite free to ignore speech altogether and simply say which letters go with which and, conversely, which fail to go with which. The letter <q> is always followed by <u>, except in foreign names such as <Iraq>. That a similar-sounding <w> never follows <q> is, however, no more interesting in such a description than the fact that <t> or <p> never follows <q>. We can relate a final <-ed> to {past tense} without being bothered by its different phonetic correlates: /Id/ (waited), /d/ (begged) and /t/ (watched). We would note that <h> occurs as the second member of some common clusters which seem to have the status of units: <ch>, <gh>, <ph>, <rh>, <sh>, <th>, <wh> as in church, ghost, physics, rhetoric, show, thin, when. Other clusters with <h> might only be found at morpheme boundaries: <nh> in inhuman, <yh> in boyhood, <lh> in girlhood, etc. It would not be part of such a description to say that in some accents of English the graphic contrast <wh> – <w>, as in whine – wine, correlates with a phonemic contrast /hw/ – /w/, but not in all accents. Nor would it be necessary to point out that <th> has the phonetic correlate /ð/ initially in demonstratives (this, there, then, those, … ) and a different phonetic correlate, /θ/, initially in lexical words (thin, theorem, thesis, …). These would be j...

Table of contents

• Cover Page
• Half Title Page
• Title Page
• Copyright Page
• Contents
• Acknowledgements
• Preface
• Conventions, symbols and Technical Terms
• 1 Alternative approaches to describing English Spelling
• 2 Literacy and English spelling: methods And Problems
• 3 Speech-to-text correspondences: encoding
• 4 Text-to-speech correspondences: decoding
• 5 Identical forms – homographs And Homophones
• 6 Conventions used in the spelling Of Names
• 7 Standardization and Spelling Reform
• References
• General Index
• Selective Word Index
• Index of Spelling Correspondences
• Index of initial and final letter strings In Words
• Index of spelling errors and re-spellings