[Journal of the Simplified Spelling Society J21 1997/1 pp13-20]
[Chris Upward: see Journals, Newsletters, Leaflets, Media, Pamflet 15, Book, Papers.]

Th Potential of Stylgides as Vehicls for Spelng Reform.

with a case-study of
The Times English Style and Usage Guide.
Christopher Upward.

This articl is ritn in Cut Spelng and was presentd as a lectur to th Societys AJM on satrday 11 may 1996.


Section 1 examns th relationship between jenrl spelng usaj in english and its reflection in dictionris and stylgides. Section 2 anlyzs th spelngs recmendd by th The Times English Style and Usage Guide from th point of vew of consistncy, econmy and aplication of th alfabetic principl. Section 3 extends this analysis to othr kinds of stylgides, noting som difrnt principls by wich they may oprate, and concluding that th presnt state of english spelng creates enormus, unecesry dificltis for al print-producers. And Section 4 considrs th incentiv that producers of print may hav to initiate a co-ordnated atak at least on th most trublsm anomlis of th english riting systm.

1. Stylgides and spelng developmnt.

1.1. Introduction.
In his classic history of English spelling [1], SSS Presidnt Profesr Donald Scragg made th foloing point (p86): "Printers' style-sheets may be seen to be of fundamental importance in the establishment of current spelling trends." This paper considrs th role of stylgides and styl-sheets in determnng spelng trends, wat ther implications ar for th evlution of english spelng, th kinds of things stylgides (and one in particulr) say about spelng today, and wethr, by aproachng ther task in a principld and concertd manr, stylgides cud help improve th spelng of english.

1.2. Dictionris and usaj.
Ask th avraj persn ho is responsbl for deciding how words ar spelt in english, and a likely reply wil be: "th dictionry." But ask th avraj dictionry ho is responsbl, and th usul reply wil be: "usaj." We here hav a chikn-and-eg situation. Wen we ar unsure how to spel a word, we look it up in th dictionry, hos verdict we then normly abide by. Conversly, modrn dictionris claim to base ther recmendations on 'usaj', wich they can nowadays establish by computer analysis of corpra of millions of words of authentic text wich presumebly also reflect how dictionris told riters to spel.

Howevr, th way in wich dictionris in practis reflect usaj may be less straitforwrd. For instnce, th orijnl 17-milion-word COBUILD text-corpus givs th form gaol as ocurng 34 times, as against 502 for jail, but th Collins Cobuild English Language Dictionary (1987), hos evidnce derives from th COBUILD corpus, merely tels us that gaol is "used mainly in British English". It dos not tel us that gaol is a relativly uncomn variant of jail. In anothr respect too dictionry claims to reflect usaj apear questionbl: such widely used forms as *accomodate with singl M or *recieve with IE ar not listd as variants on accommodate, receive. So wy do dictionris lend respectbility to a ludicrus, confusing (being esily confused with goal) and unpopulr form such as gaol, but deny it to th quite sensbl and notebly populr spelngs *accomodate, *recieve?

1.3. Closed loop of dictionry and print.
Th difrnce between goal and *accomodate/*recieve is that th formr ocurs in th mor professionly composed environmnt of print, wile th latr do not - exept as 'misprints'. Th spelngs that dictionris rejistr as acceptbl ar those that lexicografrs observ in print, wich means those that printrs, typ-setrs and publishrs, rathr than users jenrly, prefer. At th same time, dictionris ar th sorce of orthografic authority to wich those print-producers ar likely ultmatly to apeal wen in dout.

It begins to apear that this two-way feedbak loop myt be completely closed: print-producers use spelngs givn in dictionris, and dictionris list spelngs found in print. If this wer th hole story, english spelng myt be lokd forevr into its presnt forms, with no posbility of furthr evlution, let alone reform. It is ofn asertd that pland spelng reform is superfluus because "chanjes in languaj ... take place natrly and graduly" [2]. At least as far as th ritn languaj is concernd, th two-way loop described abov sujests that ther ar intractbl, intrinsic forces obstructng such alejd "natrl" evlution. If that is so, it is hard to argu that natrl evlution rendrs pland reform superfluus.

1.4. Fluidity of english spelng.
Yet th systm is not in fact rijidly fixd at evry point, because larj numbrs of english words do not hav a singl agreed spelng. Typs of spelng variation include th foloing: difrnces between british and americn usaj (harbour or harbor?); words with fundmently diverjnt forms (gaol or jail?); vacilation over word-breks (on to or onto?); hesitation over final silent E (caviar or caviare?); douts about consnnt dublng (T or TT in benefiting/benefitting?); morflojicl uncertntis (slyer or slier?); derivationl alternativs ('french' Renaissance or 'latn' Renascence?); difrnt degrees of anglicization (english plural S in bureaus or french plural X in bureaux?); variant translitrations from othr alfabets (how to represent russian щ in borsch/ borscht/ borshch/ borsht/ bortsch?); chanjed spelng conventions in a sorce language (befor 1901 in jermn Neanderthal, today Neandertal); chanjed translitration conventions in a sorce languaj (Wade-Giles conventions superseded by Pinyin in chinese, turnng Peking into Beijing); chanjes to base-words wen sufixs ar add (OU cut to O in british vigour/vigorous); and, particulrly pervasiv, th fluidity of newly coind or recently borod words that hav yet to acheve a singl agreed ritn form (yoghourt or yoghurt or yogurt?) - this latr typ of variation includes many hundreds of modrn tecnicl terms. In al such cases dictionris typicly ofr a choice, and/or disagree with each othr in ther prefrnces and recmendations. This catlog, of corse, ignors th endless variation that ocurs in (especialy privat) daily use, for instnce with 'greengrocers' spelngs and uncertnty over hetrografs, as between to bail out or bale out. [3]

1.5. Stylgides resolv uncertntis.
Print-producers ar ofn typograficl perfectionists anxius not to iritate ther readrs with inconsistnt or controversial spelngs. Because ther ar so many uncertntis not curently resolvd by dictionris, they comnly compile ther own 'in-house' lists of preferd forms for use in ther publications. These ar typicly incorprated in styl-sheets or stylgides, wich thus mediate between dictionris and th printd text, telng riters, proofreadrs, subeditrs, typ-setrs, etc, wich spelngs to adopt. Admitdly, th strictness with wich publishrs insist upon ther preferd forms varis considrbly: in som cases authrs individul prefrnces ar respectd, but in othr cases th stylgides recmendations ar prescribed as 'mandatry'.

Thanks to stylgides, th orthografy used by a particulr publishr may be mor consistnt than dictionris, wich atest numerus alternativ forms, can sujest for 'english spelng' as a hole. Insofar as a majority of stylgides myt opt for one spelng in prefrnce to its alternativ(s), jenrl usaj myt be graduly afectd, and if dictionris keep abrest of such chanjes in usaj, then ther recmendd forms myt, over time, com to reflect th domnnt prefrnces of th stylgides. Thus todays british dictionris tend to recmend encyclopedia with E rathr than encyclopaedia with AE, wich was mor comnly used erlir in th 20th century. Thanks to stylgides, therfor, som potential for flexbility and evlution myt aftr al be bilt into th two-way loop between dictionris and print. A ke question then must be: do th prefrnces of todays stylgides tend to improve th quality of english spelng, or not? This articl wil atemt at least a provisionl ansr to that question.

2. The Times English Style Guide.

2.1. Purpos of th case-study.
Th foloing analysis wil examn th efect of one particulr stylgides recmendations on th english riting systm (as wel as providing a survey of how th problms of english spelng ar perceved to impact on one of th most prestijus print-producers in th english languaj). Our findngs wil then be briefly compared (§3 belo) with equivlnt findngs from a ranje of othr such publications, to try and provide a mor representativ pictur. In §4 of this articl we shal presume that th results of our analysis ar typicl, and discuss prospects for an improved spelng of english arising from th colectiv efect of stylgides.

2.2. Choice of stylgide.
The Times English Style and Usage Guide, ritn by Simon Jenkins, came to hand fortuitusly. Its authrs concern with spelng first atractd th SSSs atention sevrl years ago thru reports that, as editr of The Times, Simon Jenkins had chanjed that newspapers traditionl use of th greco-americn -IZE endng (as in organize) to th alternativ anglo-french -ISE (se §2.10 belo for discussion of this chanje). Mor recently, Simon Jenkins voiced his furthr intrest in spelng in a Times articl [4], wher he rote:
"The world must ... admit that it has a universal language at last. ... The one quid pro quo it should demand is that English lexicographers bow before the great Webster, and accept American orthography. English spelling is still awful. If we English can alter denie, interiour and musick, we can surely end the absurdity of thorough, centre and enough, if the world is to learn them by heart."
From this we inferd that Simon Jenkins acceptd at least som of th SSSs basic premisses, namely th superiority of most americn spelngs over british, th importnce of taking th requiremnts of non-nativ-speakng lernrs into acount, and abov al th need for an improvemnt to th way th english languaj is spelt. We therfor contactd him to se how much farthr these shared vews myt strech: myt he indeed prove an activ and valubl suportr of reform in th public arena? In reply, he kindly donated a copy of The Times English Style and Usage Guide (henceforth referd to just as th Style Guide), a hansm litl blak-bound hard-bak of 159 pajes ritn, as he tels us, "som years ago" Altho th book laks th norml bibliograficl data (even th authrs name is misng), a ruf date of publication can be deduced from a refrnce to febry 1992 on th last paje.

Th book is enormusly informativ about many of th dilemas facing riters of english, and is ritn with a lyt tuch, its recmendations being intrspersd with witicisms such as th advice "Rarely start a sentence (except this one) ... with an adverb". At th same time, it exudes authority, as wen it says that most of its precepts ar "mandatory" for Times writers, or opines that "Sentences starting with adverbs are normally built on sand."

2.3. Spelng principls implyd.
Th books Introduction (pp5-9) sets out its aims and principls with exemplry clarity undr th hedngs 'Brevity', 'Verbs', 'Abstractions', 'Punctuation', 'Conjunctions', 'Paragraphing', 'Quotations', 'Headines', 'Style'. Ther is no hedng for 'Spelling', nor dos th Introduction mention it explicitly. Nevrthless, th foloing quotations from th Introduction can be interpretd as berng strongly upon spelng:
English is not a language fixed for all time. Speech changes and its written forms should change too. The Times must use the language of its readers, but that language at its best, clearest and most concise. The writing in a newspaper should also be consistent: hence the need for a style guide.

Most of the guidance is mandatory for staff writing in The Times, though some is permissive. ...

The introductory remarks are expanded under the relevant heading in the alphabetical list. ... Where the guide conflicts with The Oxford Dictionary for Writers and Editors, the guide should take precedence, but the Dictionary should be consulted where the guide is silent.

Brevity. A writer's most precious commodity is space. Space is time for writer and reader alike. ... So use short rather than long words.
Two criteria for good styl ar thus sujestd wich ar aplicbl to spelng: conciseness and consistency (these ar also criteria undrlyng th desyn of Cut Spelng). We shal want to se how far they ar observd in th spelng advice givn in th alfabetic section of th Style Guide. It is noticebl that adherence to th alfabetic principl of predictbl sound-symbl corespondnce is not one of th criteria, but 'consistncy', if consistntly aplyd, can take us a long way down that road. In adition, th alfabetic section (undr hyphens) states that "sense and appearance are more important than rules", but dos not specify how to aply these encurajng words in practis.

2.4. Wat proportion concerns spelng?
Th books alfabetic list is 150 pajes long and contains an estmated 600-700 items. Som ar very short (ofn just 3 words: 'forum plural forums'), wile othrs covr sevrl pajes (eg, 'titles', 3pp). Of slytly over 60 hedwords beginng with A, nearly half hav nothing to do with spelng (a few ar just cross-refrnces: 'and see punctuation'), nearly a quartr ar partialy concernd with spelng ('amid not amidst'), and nearly a third ar purely or mainly spelng recmendations ('amok not amock or amuck'). We myt conclude that these figrs (ie, around half an english stylgide taken up by spelng advice) sho just how serius a problm th presnt spelng of english is for print-producers. That fact, one myt think, shud hav induced spelng reform long ago.

2.5. Typs of spelng advice.
Th Style Guide is in jenrl aimd at professionl riters ho require few warnngs about comnplace spelng traps. Th entry accommodation thus carris a stylistic warnng ('a long dull word best avoided'), but no remindr of th need to dubl th M. Nevrthless, Times riters ar warnd about th R/RR variation between harass/embarrass, th EI/IE reversl in seize/siege, and th G/J altrnation between sergeant/Serjeant. Rarer words ar mor likely to atract coment: we must remembr to rite PH, not P, in diphtheria, diphthong, ophthalmic; and Y not I, in dysentery, aneurysm. Varius distinctions of meanng corespondng to difrnces of spelng ar specifyd, such as dependant/ dependent, discomfit/ discomfort, discreet/ discrete, enquiry/ inquiry, falangist/ phalangist, forbear/ forebear, forgo/ forego, gaff/ gaffe, gibe/ jibe, grisly/ grizzly. Orthografrs wil natrly ask wethr th advantajs of such distinctions outwei ther hazrds, it being clear at least that, if those distinctions wer oblitrated, th Style Guide wud not need to discuss them. We may certnly dout th valu of distinguishng bogey (golf), bogie (weels), and bogy (gost), wich dictionris alow variusly to overlap; wud it not be mor helpful to use bogey for al meanngs? Riters ar urjd always to chek a few names hos spelng is notoriusly unpredictbl: Alasdair, Alastair, Alaster, Alistair; Catharine, Catherine, Katharine, Katherine; and th surname prefix Mac-.

A numbr of items deal with th spelng of foren words: nearly 3 pajes discuss arab names and ther transcription problms; th most comn chinese names ar to folo th old british Wade-Giles spelng systm, not th new pinyin standrd (Peking, not Beijing); and ther is advice on wen to use diacritics (eg, th acute accent on final É to be retained, as in communiqué). Th perenial uncertntis over capitlization (President Bush, but George Bush, the president) and hyfnation (co-operation, but uncooperative) ar also recurent themes.

Many of th items 'partialy concernd' with spelng involv prefrnce for a word-form wher difrnt pronunciation natrly means difrnt spelng (eg 'amid not amidst'). We here note th prefrnce for mor concise past tense forms endng in -T over alternativs with -ED. Thus: 'burnt not burned', and simlrly dreamt, leant, leapt, learnt; learned then necesrly has two sylabls. For our purposes, such variations can be ignord, since spelng difrnces corespondng unambiguusly to pronunciation difrnces shud be th norm anyway.

2.6. Americn spelngs.
In adition to th abov categris, ther ar a few jenrlized spelng recmendations. One relates to americn spelngs:
the titles of American institutions etc that contain words of which the American spelling is different from the English (eg, Secretary of Defense, American Federation of Labor) should be written in The Times with English spellings: Secretary of Defence, American Federation of Labour
Americn readrs wil hopefuly be amused rathr than irkd that 'americn' spelngs ar here contrastd with 'english' rathr than 'british' variants. We may also remark that th form defence is historicly and systemicly indefensbl.

This rejection of americn spelngs contradicts both th pro-americn vew Simon Jenkins advocated in his febry 1995 articl (se §2.2 abov) and th criterion of conciseness propoundd in th Style Guides introduction (se §2.3 abov). Howevr, in practis th Style Gide is not consistntly anti-americn - and therby contradicts th Introductions othr main criterion, wich is consistncy. So we hav AE rathr than americn E in archaeology, leukaemia, palaeo-, praesidium, but dieresis, Greco-, medieval without A. We note the longer british forms jewellery, storey, but americn artifact, jail preferd to british artefact, gaol. Th recmendd forms install, instalment, instil mix british L and americn LL, tho this is a rare instnce wher british conventions ar mor concise than th americn. Inconsistncy is also th consequence of preferng british pedlar to americn peddler, since th verb is to peddle. Likewise, adherence to th principld british inconsistncy of practice/practise (to be consistnt, british spelng shud also distinguish a *promice/to promise) prevails over th esir americn blurng of th difrnce.

2.7. Treatmnt of mute E.
Anothr jenrlized recmendation is givn undr th hedword e: "delete this often superfluous letter when the dictionary offers you a choice, as in judgment, acknowledgment. See also mute e." This mute e entry deservs quoting in ful.
if there is to be any consistency in keeping or dropping the final mute e in words taking the suffixes -able, -age, -ing, -ish, a rule is needed. Fowler has one that is simple and sensible: drop the e unless it is needed to emphasise the soft sound of a preceding c or g. Thus changeable, knowledgeable, singeing (from singe - soft g must be emphasised to distinguish from singing), swingeing, traceable. Observance of the rule leads to a few unusual but defensible forms: aging, debatable, milage, ratable. When in doubt consult a standard dictionary: in all cases where an alternative spelling without the e is given, that is the one to be used in The Times. Remember, the English language should be allowed to evolve, under pressure of common usage, in the direction of simplicity.
Th spirit and intention behind these recmendations ar admrbl (they also undrlie Cut Spelng). Howevr, som of them unfortunatly create systemic problms wich th orthografr needs to be aware of. (A practicl iritation is furthrmor that my spelchekr queris many of th abov forms!) Th recmendd forms ar not so much markd by th claimd simplicity (if that means predictbl sound-symbl corespondnce) as by brevity. It is one of th major defects of th presnt spelng of english that adng sufixs ofn entails an unpredictbl altration to th base-word (such as insertng or deleting letrs), altho ther is no equivlnt chanje to th pronunciation to indicate wat th spelng chanje shud be. Endless errs and uncertntis (of th kind th Style Guide atemts to resolv) ar th inevitbl consequence.

Varius such problms arise from dropng mute E. One problm confronts readrs ho do not no how a word is pronounced (typicly, th very constituency of non-nativ-speakrs hos needs Simon Jenkins elswher recognized). For them th paralel spelng of, say, milage and cartilage increses th dificlty of decoding. Wat is mor, lernrs in jenrl hav trubl mastrng th function of 'majic E' in words like mile, rate, and th deletion of that E in milage, ratable only agravates matrs. Th lojic of th sujestd omissions is anyway obscure: if E must be kept in traceable to mark th soft C in trace (contrast implacable), wy is it not also needd in ratable to mark th long A of rate (contrast palatable)? Nor dos th Style Guide explain wy E is not needd in aging to sho th soft G. Th confusion is most blatant wen E is dropd from acknowledgment but retaind in knowledgeable: if th digraf DG sufices to sho soft G in th formr, wy dos it not sufice in th latr?

2.8. Plural -OS or -OES?
A famus inconsistncy of english spelng arises in nouns endng in -O: shud ther plurals ad -S, as in pianos, or -ES, as in potatoes? (Readrs may recal th previus US Vice-Presidnt Dan Quayles public humiliation wen he reduced potatoes to th singulr *potatoe.) Times riters get no help here, being required to memrize frescoes, grottoes, mosquitoes, tomatoes with -OES, but crescendos, ghettos, manifestos, tornados with -OS. Admitdly, as Cut Spelng discovrd, neithr endng is ideal: -OES is less concise and conflicts variusly with does, shoes, Averroes, wile -OS conflicts with th regulr 'greek' endng of words like ethos, pathos; but th -OES/-OS distinctions proposed by th Style Guide ad inconsistncy and unpredictbility to th snags of both.

2.9 Dublng consnnts.
Anothr inconsistncy in th Style Guide concerns that ubiquitus problm of wethr verbs endng in an unstresd vowl plus singl consnnt shud dubl th consnnt befor a sufix. Th Style Guide ilustrates, but dos not explain, th rule by contrastng preferred/proffered. Most exampls listd ar givn with a singl consnnt: balloted not ballotted, and simlrly benefited, bigoted, riveting, biased, focused. But for som reasn hiccup has -PPED (no othr -P verbs ar givn, so we dont no wethr Times riters ar to rite gallop, gossip, kidnap, worship with -PED or -PPED). Th problm of -L dublng is not mentiond, tho presumebly th anti-americn principl wil ensure that travel and its numerus brethren (exept, presumebly, parallel, wich is an exeptionl exeption that gos against al th othr norml exeptions) wil keep ther perverse british -LLED.

2.10. -ISE or -IZE?
We noted in §2.2 abov how Simon Jenkins reversd The Times traditionl -IZE usaj, perhaps to avoid an odius americnism. Th Style Guide says th foloing:
avoid the z construction in almost all cases. This is volcanic ground, with common usage straining the crust of classical etymology. This guidance is a revision of the Greek zeta root ending in the direction of a Latin ending and common usage: apologise, organise, emphasise, televise, circumcise. The only truly awkward result is capsize which should be left in its Grecian peace.
Striking tho th volcanic imajry is, th Style Guides etmlojicl explnations ar at best garbld: latn translitrated greek zeta as Z (th latn verb baptizare is th erliest nown exampl), not as S, and english has taken th -ISE sufix from french, not from latn (french baptiser). Som words endng in -ISE do hav it from latn, but it is then part of a longr root, not a sufix: in televise th root is vis, as in vision, visible, and in circumcise th root is cis, as in incision, (s)cissors. Th implication that capsize has anything to do with greek is fantasy - it is a modrn nauticl term. Th basis of this hole argumnt in favor of -ISE is curius: it uses etmolojy to overturn etmolojy.

In fact, th sole advantaj of -ISE is that it faces riters with fewr exeptions than -IZE dos. Th exeptions with etmlojicl -ISE ocur in sevrl comn words hos franco-latn endng is not th greek zeta root. Th 10 most comn ar exercise, surprise, enterprise, compromise (acordng to COBUILD these first 4 ar al mor comn than realise, recognise, th comnst -IZE posbilitis), advise, disguise, devise, despise, demise, advertise. Two jermanic roots ar rise, wise (tho th latr has IZ in wizard).

Th objections to -ISE ar systemic, syclojicl and sociolinguistic. Stylgides ar ritn for riters, not for readrs, but orthografrs hav to remembr that th systm needs to serv riters and readrs equaly. Only th alfabetic principl, not etmolojy, not dictionris, and not usaj, can provide th requisit gidance to both. Letrs shud be used, as far as posbl, to represent speech-sounds regulrly and predictbly. Then riters can spel words corectly without needng to chek in dictionris or stylgides, and readrs can pronounce words corectly wethr they hav herd them befor or not. That is th ideal, from wich th traditionl orthografy of english is so far removed. Spelng reformrs disagree about how closely ritn english can be made to aproxmate to pronunciation in th short term, but as regards isolated, marjnl featurs of english spelng as listd in stylgides, then evry recmendation in favor of a mor predictbl sound-symbl corespondnce is a step in th ryt direction.

Between -ISE/-IZE th ryt choice, by this criterion, is clear. Th final consnnt is voiced, and in english Z, not S, represents th voiced valu unambiguusly. Tho -ISE may, in traditionl orthografy, entail fewr exeptions for riters, it entails many ambiguitis for readrs, as seen in th alternativ sound-valus in promise, paradise, expertise, organise. And if -IZE leves riters with th problm of a dozn or so exeptions, th remedy is clear: stylgides shud recmend exercize, surprize, enterprize, compromize, advize, disguize, devize, despize, demize, advertize, etc - and rize, wize.

Then th -ISE/-IZE dilema wud be finaly resolvd - or nearly so. For ther remains a paradoxicl litl bunch of words wich ar jenuinly of greek orijn, but hos greek root was ritn with sigma, not zeta. Th main exampls ar analyse, paralyse (wich th Style Gide lists undr -LYSE), hos spelng with YS derives from th nouns analysis, paralysis. Americn spelng has long been recnciled to analyze, paralyze, but th alfabetic principl strictly speakng requires these words to be fuly alynd with capsize, givng analize, paralize. If this seems too bold a step for stylgide authrs, let them considr spanish (analizar, paralizar) and italian (analizzare, paralizzare), wich hav no truk with 'greek' Y - nor with many othr spelng problms that beset english.

A last point has to be made in favor of -IZE. Wile th british tend to be alerjic to most americn spelngs, on th hole America is mor tolrnt of british forms (canadian usaj tends to fal between th two). But in th case of -ISE/-IZE th reverse is th case: 'americn' -IZE is quite comn in british usaj (Oxford University Press and Collins, two publishrs of renownd dictionris, both prefer it), wheras 'british' -ISE is strongly disliked (quite rytly) by americns. For th sake of a world standrd of english spelng, therfor, evry stylgide shud recmend -IZE.

A final obsrvation here arises from th CS forms anlyz, paralyz: if, as New Spelling 90 recmends (and CS tends to implmnt too), th long valu of I is best always rendrd by Y, then al th abov words wud reduce ther endngs to -YZ, thus baptyz, surpryz, c/kapsyz.

2.11. Mislaneus inconsistncis.
Sundry othr inconsistncis wer noted in th Style Guide. Ther wer th discrepnt translitrations Chekhov, Tchaikovsky, hos initial consnnt is spelt with th same letr ч in russian; english has endd up with one proprly anglicized form with CH, and one form borod from french, wich regulrly spels that consnnt-sound as TCH, and lojicly also rites Tchekhov. Th principl of conciseness is breachd by th prefrnce for loath over loth; tho th paralel of oath implys that th alfabetic principl is perhaps being tacitly observd (but shud we then distinguish both/cloth as boath/cloth?). Tho not exactly an inconsistncy, we may regret (Cut Spelng difrs here) th prefrnce for gypsy, pygmy over th less ambiguus gipsy, pigmy (contrast th long valu of Y in gyrate, pyrites). Conversly th Style Guide prefers siphon to syphon - admitdly in traditionl orthografy initial SY- mostly has th valu as in syllable, sympathy, syndrome, tho PSY- dos not).

3. Comparisns, sorces and authoritis.

3.1. Th need for comparisn.
The Times Style Guide has been anlyzd in detail because its authr had expresd concern about th state of english spelng ("still awful"), and th hypothesis therfor seemd worth pursuing that its recmendations myt represent a systemic improvemnt to the orthografy. In that respect our findings, tho othrwise of intrest, ar disapointng. Som brief comparisns with othr stylgides and simlr publications may at least indicate how typicl th Times Style Guides recmendations ar of th jenrality of publishng practis in Britn. Dos The Times ofr anything aproachng a british standrd, or ar its recmendations based on foundations as fluid as th dictionris wich it urjs its riters to consult wher th Style Guide is 'silent'?

3.2 The Economist Pocket Style Book.
Th stylgide produced by th british weekly The Economist (1986) resembls that of The Times in size and structur. Most of its 107 (slytly larjr) pajes constitute an alfabeticl list. But only about a dozn of its hedwords concern spelng, tho thre of them ar fairly substantial. Minor items treat diacritics (hedword Accents), program/programme (hedword Americanisms), autarchy/autarky, th precednce of pinyin over Wade-Giles (hedword Chinese names), forgo/forego, Hyphens, and meting out meet punishmnt. Th thre mor substantial items com undr th hedwords -able/-eable/-ible, Plurals, and Spelling. The Economist difrs on one jenrl point from The Times: wheras th latr claims to aply principls of consistncy and conciseness, th Economists Preface says: "The aim of this style book is to ... set some arbitrary rules. The arbitrary choices made are those of the paper's editors over many years." (my italics).

Th item -able,-eable,-ible merely lists 12 words endng in -ABLE, 4 words endng in -EABLE, and 7 words endng in -IBLE. Th -UBLE/ -UABLE dilema, as between voluble/valuable is not mentiond. No gidance is givn for deciding wich endng is apropriat, non of th words ocurs among th 20 most comn with eithr th -ABLE or th -IBLE endng, and a wel-nown spelng trap (*responsable) is absnt. Th base-forms of 10 out of th 12 -ABLE words end in -E (eg, debatable from debate), and rase th question wethr -ABLE or -EABLE is mor apropriat. (Som dictionris list debateable with -E, wile othrs do not; wat dos this imply about dictionris reflectng usaj?) On th hole The Economist recmends th same forms as The Times, with no E exept aftr soft C or G (serviceable, manageable); but sizeable keeps E wher The Times wud presumebly cut it. (Dictionris too somtimes list sizeable wen they dont list debateable; why?) In vew of al these variations, we may point out that, by adng just -BL to th base-form in nearly al cases, Cut Spelng avoids these uncertntis, with debatebl, manajbl, responsbl, servicebl, sizebl, volubl, valubl. Th -ABLE, -EABLE, -IBLE problm is thus not insolubl.

Th item Plurals also just consists of lists, for th plurals of words endng in -O, -EAU, -US, -UM, -F and -A. For -O, 21 words ar givn with plural -OES and 15 with plural -OS; only 5 of these wer also givn by The Times, two of them with difrnt recmendations (manifestoes, tornadoes, for wich The Times gave -OS). For -EAU, plural -S is recmendd for bureau, plateau, but, anomlusly, th french plural -X for chateau, altho if this wer a tru french form, a circmflex accent wud be required (châteaux). The Economist agrees with The Times that th plural of -US endngs shud not dubl th S, ie, focuses, but it introduces a complication of its own, distinguishng th V plural of hooves, scarves, wharves from th F plural of dwarfs, roofs, turfs, and th -S plural of agendas from th -E plural of formulae. A stylgide that gave priority to consistncy cud harmnize many of these variations without any daring orthograficl inovation at al, but The Economist evidntly regards spelng as no less arbitry than othr aspects of styl - and therby helps to make it so.

The Economists item on Spelling consists of a two-paje list of "common difficulties", by wich is ment dificlt words, not dificlt patrns or rules; again no gidance on jenrl principls is givn. As with The Times, americnisms ar tabu (write -ISE, not -IZE), exept that current account may be ritn checking account "when explaining to Americans", and inquire is preferd to enquire. Many words also ocur in th Times list, but here sho th foloing difrnces: bogey can serv for al meanngs, exept "on a locomotive"; hiccup has -OUGH, not -UP; enquiry/inquiry ar not distinguishd in meanng, th latr servng for both senses; loth is preferd to loath (despite loathe); and mileage is preferd to milage. Th Economist aditionly recmends riting adviser with E beside advisory with O; Filipino with F, P beside Philippine with PH, PP; fullness with LL beside fulfil, fulsome with L; and gypsy with Y beside pigmy with I.

One is inclined to conclude that, if The Times is inconsistnt, The Economist is confused. If we take them as togethr representng th preferd orthografy of th british press, we hav an amalgm of inconsistncy and confusion.

3.3. Othr sorces.
Without atemtng any comprehensiv or systmatic listng, we may mention som furthr sorces of gidance for british riters on english spelng. Both The Times and The Economist ar exampls from th field of jurnlism, but of corse book publishrs hav no less a need for gidance. Ther is, howevr, th difrnce that, wile newspapers and jurnls ar keen to ensure consistnt spelng of individul words for ther readrs from issu to issu, book publishrs ar mor concernd to ensure consistncy within each publication, rathr than between seprat publications.

Cambridge University Press (CUP) publishs a stylgide (Butcher, 1992), with 24 refrnces to spelng in its index. Th quality book publishrs orthografic task is mor complex than that of th newspaper publishr. Regardng americn spelngs, CUP recmends respectng th prefrnces of authrs, wile ensuring that these ar internly consistnt (CUP has to cater for americn and non-americn authrs, and for readrs worldwide). This is a far mor trublsm procedur than simply banng americn spelngs outryt; for instnce, CUP editrs and proofreadrs ar liabl to hav to implmnt any recognized variant of th -ISE/-IZE rules, wile corectng non-recognized variants; but quotations within books may folo difrnt spelng conventions from th main text. Th CUP stylgide, like th othrs, lists spelngs wher mor than one form is comn, but on th hole it expresses no prefrnce (exept that th -YZE endng [analyze, etc] is only acceptbl in americn contexts), merely ruling that th same form be used thruout th book (exept in quotations). One unmistakebl anomly is unshakebly insistd on: unmistakable must end in -ABLE, but unshakeable has -EABLE.

Two dictionris of english spelng may also be mentiond here, one publishd by Collins (1993), th othr by Harraps (Wileman, 1990). Unlike th stylgides previusly discusd, these ar refrnce works specializing in spelng, tho each has a difrnt target readrship. A main function of th Collins is to sho text-producers wher words may be split with a hyfn at th end of a line. Th foloing entry shos th format used: un+shak+able (but no un+shake+able alowd as an alternativ). Collins givs som alternativ forms, as with hic+cup+ing or (on th foloing line) hic+cup+ping (hiccoughing as recmendd by The Economist dos not apear). British cen+tre, cen+tring apear alfabeticly a good colum furthr on than cen+ter US, cen+ter+ing. Jail apears som 73 pajes later than gaol, with no cross-refrnce, and wheras gaols ar gardd by gaolers, jails ar mor jenrusly stafd, with both jailers and jailors.

The Harrap dictionry is by contrast desynd to cope with th foloing conundrm: "Please, how do you spell 'inconceivable'?" - "Why don't you look it up in the dictionary?" - "I did, but it's not there!". To resolv this problm, Harrap contains larj numbrs of likely mispelngs (printd in blu), wich can be lookd up alfabeticly to discovr th conventionl form. If u cant remembr how to spel once, u look up a likely candidat such as wunse, or wunce (but not wonce), and ther u find once printd in blak beside th mispelt blu form. If u no th spelng of gaol, u can look it up in blak print, tho u find it imediatly abov gaol printd in blu, wich is a mispelng of goal; conversly, black gaol is also givn against blu goal. A gaol is gardd by a gaoler, but jails apear ungardd, eithr by jailers or jailors. Th forword trys to reasure th user that "far from being a haphazard arrangement, English spelling is based on a complex system of rules". These rules ar presumebly exemplifyd by th books "spelling hints", wich contain such jems as: "In 'ee' sound words, c is always followed by ei... However, there are many exceptions"; and: "Most words ending in the 'ize' sound are usually spelt ise, e.g. analyse, paralyse." It wud be intrestng to no wethr such advice has evr helpd anyone improve ther spelng.

3.4. Authoritis.
Wile CUP and Collins present themselvs as spelng authoritis in ther own ryt, both The Times and The Economist defer to othr authoritis. We quoted The Times abov (§2.3) referng to the Oxford Dictionary for Writers and Editors (1984), to wich it deferd by defalt (tho defyng Oxford wenevr it thot fit), and it furthr refers to Collins English Dictionary (is it coincidnce that Collins and The Times ar both Murdoch compnis?) and Fowler (1983). Such defrnce may, howevr, be a blunt instrumnt: th Style Guides advice "when in doubt consult a standard dictionary" leves open th questions of wich dictionris ar standrd and wich ar not, and wat to do wen dictionris ar found to disagree. The Economist refers to th British Standards Institution for Proofreader's marks, and undr Dashes quotes Gowers (1979) precept "Do not use it as a punctuation maid-of-all-work". Undr Additional reading it also lists Hart's Rules (1983). We hav in these works a body of authority on english (especialy british) spelng wich we canot examn mor closely here, but wich cry out for ful analysis to establish wethr th gidance they ofr is any mor coherent than that of th works investigated abov.

4. Stylgides and spelng reform.

4.1. Verdict on th presnt styl gides.
Th abov analyses sujest that stylgides at least ar not based on any coherent undrstandng of english spelng as it now is, nor on any coherent principls for determnng wich spelngs shud be recmendd wher a choice is availbl. They seek merely to help riters shape ther text to meet mor or less arbitry editorial requiremnts. They do not considr th riting systm as a hole, eithr in its role in world comunication nor in th myriad internl intractions of its orthografic featurs. Nor do they considr th brodr needs of al categris of user, wich include non-nativ as wel as nativ speakrs, readrs as wel as riters, th barely litrat as wel as skild practitionrs, lernrs as wel as teachrs. Th two principls enunciated in th 'Times' Style Guide, conciseness and consistency, ar fine in themselvs, tho they ar not noticebly reflectd in th spelngs recmendd. But th alfabetic principl of predictbl sound-symbl corespondnce, wich shud be th ultmat criterion for al orthografic decisions, is rarely hintd at in any of th abov publications (th use of E to indicate that a preceding C or G is soft is such a rare instnce).

4.2. Stylgides as a developmnt mecnism?
In §1.5 abov, we askd wethr stylgides myt ofr an escape from th trap of th two-way loop between dictionris and usaj, and therby provide a mecnism for th evlutionry improvemnt of english spelng. Alas, our analyses hav shown no evidnce for that hapnng at presnt. Th stylgides disagree with each othr, or in th case of CUP ar wilng to submit to most of ther authrs persnl prefrnces, wich ar likely to sho even gretr variety than do th stylgides used by th press. Th global inconsistncy of anglo-americn difrnces by defnition imposes a dilema on al british stylgides, wich ar torn between rejection of th stranjeness of americn forms and atraction to ther self-evidnt advantajs. On th one hand they may recognize that som americn forms ar incresingly acceptd evrywher (eg, encyclopedia); but on th othr they may wish to emfasize th difrnce by insistng on -ISE in prefrnce to -IZE.

One of th obstacls to expectng stylgides to produce progress in english spelng is epitmized by th Style Guide recmendation that incomunicado be spelt with a singl M. Non of 4 british dictionris and 1 americn consultd lists that spelng. Two argumnts suport singl M: th word is borod from spanish, wich nevr normly rites MM (eg, acomodación), so singl M in english wud reduce discrepncis between languajs; and if english nevr rote MM, a considrbl simplification wud result. On th othr hand, since at presnt th root commun- is always ritn with MM in english, creating an isolated exeption for incomunicado wud ad a furthr hazrd to an alredy extremely hazrdus riting systm. In othr words, tinkrng with th spelng of a singl word is bad. Th only beneficial reform is one that aplys across a definebl ranje of forms, as wen Cut Spelng simplifys almost al dubld consnnts, predictbly and by rule. If a stylgide confines itself to ruling on individul words here and ther, it is most unlikely to confer any benefits on english spelng at al, in fact it is mor likely to make an apalng situation worse; and with incomunicado th Style Guides even ofends against its own admrbl criterion of consistncy.

This hole situation testifys to th practicl caos produced by th systemic caos of english spelng. Th endless nuisnce sufrd by riters, proof-readrs, editrs, publishrs, etc, etc, that results from th lak of a rule to determn, for instnce, wethr nouns endng in -O shud ad -S or ES to form ther plurals, shos them to be as much victms of th presnt spelng as ar lernrs worldwide. That ot to motivate al those involvd to seek a solution, but a solution requires first, an undrstandng of th problm (ie, neglect of th alfabetic principl), and secnd, co-ordnation of responses.

In his 'Times' Style Guide Simon Jenkins shos som insyt into th potential for chanje wen he says: "Remember, the English language should be allowed to evolve, under pressure of common usage, in the direction of simplicity." Wat our analysis has demnstrated is that ther is no syn today that th "pressure of common usage" has any significnt efect. Th lesn from othr languajs, indeed, strongly sujests that a good riting systm can only be created by co-ordnated planng.

4.3. Futur steps.
Wat steps can be envisajd to take matrs forwrd? Regardng prelimnry reserch, a survey of al english stylgides curently in use that contain spelng recmendations shud be undrtaken (eg, se 'Chicago' for the USA, Pam Peters, and 'Australian Style Council' in th Refrnces belo), for these ar an expression of practicl need of professionl riters and print-producers and hence perhaps a pointr to prioritis for reform wich they themselvs myt be most inclined to embrace. This reserch shud be suportd by a survey of spelng-variants listd in dictionris. A useful methodolojy for this was developd by Donald Emery (1976) for a survey limitd to americn dictionris, and it shud be extendd now to australian, british, canadian and othr dictionris, including bilingual dictionris wich help disemnate english spelng patrns to th non-english-speakng world. Th infrmation these surveys cud provide wud then valubly complmnt al th evidnce from education and sycolojy to bild up a ful pictur of th efects of todays spelng of th premir languaj of world comunication.

Beyond reserch to establish th ful details of th presnt situation, steps ar needd to bring togethr al th intrestd partis, th professionl users, th teachrs and lernrs, and th forml recordrs of english spelng, to unite them in ther (as yet larjly unaknolejd) comn cause: th improvemnt of th English spelng systm. It is beyond th scope of this articl to speculate on how such an aliance myt be bilt, but since Simon Jenkins Style Guide provides 159 pajes of suport for his forthryt pronouncemnt that "English spelling is still awful", we myt look to him for a first ansr. A minml step myt be for his stylgide to be revised to ensure gretr consistncy and improved implmntation of th alfabetic principl. A maxml step myt be for th Murdoch orgnization to espouse th cause of th worldwide modrnization of english spelng.


[1] Scragg, Donald G (1974) A history of English spelling, Manchester University Press.

[2] as asertd by Baroness Hooper in the House of Lords on 4 march 1987 - se JSSS J5 1987/2, p4.

[3] Hutton, Will (1995) The State We're In, London: Jonathan Cape, has 'bale out' on p134 and 'bail out' on p135.

[4] Jenkins, Simon (1995) 'The triumph of English' in The Times, 25 febry.


Australian Style Council. Australian English in a Pluralist Australia: proceedings of Style Council 1995, Sydney: Dictionary Research Centre, Macquarie University.

Butcher, Judith (1992) Copy-Editing: The Cambridge Handbook for Editors, Authors, Publishers (3rd edition, revised and updated; 1st edition 1975).

The Chicago Manual of Style, University of Chicago Press, 14th edition (1993).

Collins Cobuild English Language Dictionary (1987), London: Collins, 1st edition.

Collins English Spelling Dictionary, Glasgow: HarperCollins (1993).

Emery, Donald W (1976, revised edition) Variant Spelling in Modern American Dictionaries, USA: National Council for Teachers of English (first edition 1973).

Fennelly, Laurie (1991) New Spelling 90, Simplified Spelling Society.

Fowler, H W (1983) Dictionary of Modern English Usage, Oxford University Press, 1st edition 1926, 2nd edition revised by Sir Ernest Gowers and corectd 1983.

Gowers, Ernest (1954) The Complete Plain Words, Penguin, 1979.

Grimond, John (1986) The Economist Pocket Style Book, London: The Economist Publications Ltd.

Hart's Rules for Compositors and Readers at the University Press Oxford (1983, 39th edition), Oxford University Press (1st edition 1893).

The Oxford Dictionary for Writers and Editors (1984) Oxford: Clarendon Press, reprinted with corrections from 1981 edition.

Peters, Pam 'In search of Australian style' in English Today 45, Vol.12 No.1, pp3-10.

Wileman, Bud & Wileman, Robin (1988) Wordspell: a spelling dictionary, London: Harrap.

Back to the top.