[Journal of the Simplified Spelling Society, 12, 1992/1 pp3-5]

On Harmnizing Cut Spelng and New Spelling.

Som ideas from Chris Upward for discussion at th meeting of th
Simplifyd Spelng Society on 27 april 1991.

1. Bakground.

At its meeting on 23 September 1988, th Society revewd its stratejis, in particulr discussng wethr it shud in futur comit itself to just one spelng reform proposal as its 'oficial 'policy. Minut 8 from that meetng states: "... it was agreed that we needd a pluralistic aproach to reform", in othr words, th Society wud se its role rathr as a forum for discussion of al ideas on spelng reform and a sorce of suport for watevr scemes it decided to aprove rathr than exclusivly as th vehicl for a singl sceme. Since then, two main scemes, Revised New Spelling (RNS), subsequently renamed New Spelling 90 (NS90), and Cut Spelling (CS), hav receved th Societys main atention.

2. Difrnt scemes mutuly exclusiv?

In acordnce with th Societys new pluralistic aproach, ther is no necessity for th difrnt reform scemes it may aprove to relate to each othr in any explicit or lojicl way. Th Society can perfectly wel say, we hav a number of scemes, and each must be jujd on its own merits without regard to th othrs. Such a view howevr at once rases questions such as: ar th difrnt scemes mutuly exclusiv? dos th public hav to choose between them, if it is intrestd in implmntng a spelng reform at al? how can th public be persuaded to implmnt any reform if it is ofrd a ragbag of possbilitis?

3. NS90 and CS complementary?

Wen these questions ar askd of th relation between NS90 and CS, th ansr is that they hav difrnt purposes, but since each purpose is valid in its own ryt, they ar not necessrily mutuly exclusive. NS90 is desynd (as wer its ancestrs, th successiv versions of NS from erly in th 20th century onwrds) as a radicl reform of english speling, mor or less giving each foneme a singl standrd symbl or digraf, and not atemtng to preserv many TO forms for ther own sake. CS on th othr hand is conceved as a first-staje reform, being larjly confined to th omission of redundnt letrs, and therfor retainng a recognizeabl similarity to TO spelings for most words. Th public myt then react by asuming that CS leads on to NS90, and that th two systms ar desynd to be mutuly compatbl, NS simply taking CS furthr.

4. NS90 as an extension of CS?

To a significnt degree that is th relation between NS90 and CS. Being desynd as an 'ideal' spelng systm for english, NS90 has shed th most obvius redundnt letrs from TO, and to that extent bilds on th omissions of CS. As ocurs by CS Rule 1, NS90 rites head without <a>, debt without <b>, scissors without <c>, handkerchief without <d> (?), heart without <e>, foreign without <g>, honest without <h>, and so on with exampls of redundncy for most letrs of th alfabet. And as ocurs by CS Rule 3, dubld consnnts ar simplifyd in NS90. Th major advance by NS90 beyond CS is that it respels sounds representd by aberant letrs in TO (e.g. meni for TO many), and it regularizes th long vowls with <ae> for long <a>, <ee> for long <e>, etc. Thus far we can brodly say that NS90 and CS ar mutuly compatbl, indeed complmentry, and th Society cud to that extent launch them in tandm in a concertd campain.

5. Wher NS90 and CS Conflict.

Howevr ther is one major area in wich th two scemes conflict. CS Rule 2 cuts out certn letrs wich ar retained (i.e. restord, if NS90 is seen as a subsequent staje to CS) in NS90. These ar th vowl-letrs asociated with post-accentul schwa befor <l, m, n, r>. CS treats them as redundnt insofar as th <l, m, n, r> can be anlyzed as having sylabic valu. It takes words such as TO apple, rhythm, mightn't, centre, wich respectivly contain sylabic <l, m, n, r), as modls for cutng TO chapel, fathom, important, enter bak to th same endngs, givng CS apl/chapl, rythm/fathm, mytnt/importnt, centr/entr. NS90 on th othr hand adopts <e> as a standrd symbl for shwa, and so presumably spels these words as chapel, fathem, importent, enter. (Th sampl of NS90 provided by th minuts to th Societys meeting on 14 july 1990 howevr demnstrates sevrl patrns wich retain TO vowl-letrs for post-accentul shwa: sekretari simbol, leson, japaneez, prymari, inishal, eksepshonal; such exceptions to th regulr spelng of post-accentul schwa <e> in NS90 leve a major problm of TO unreformd - riters wil stil hav to lern a difrnt spelng for th midl sylabl of separate, desperate, corporate.)

6. Th problm of post-accentul shwa.

Post-accentul shwa represents one of th gretst spelng dificltis in TO, since th pronunciation of th unstressd vowl is identicl, but many difrnt letrs ar used to spel it, ofn arbitrly. Th larjst numbr of alternativ spelngs ocurs befor final <r>, as in TO burglar, teacher, amateur, Yorkshire, actor, neighbour, centre, murmur, injure, martyr, but many alternativs also ocur befor < l, m, n>. Th magnitude of th problm in TO can be jujd from th fact that out of 290 pajes in th Penguin Rhyming Dictionary as many as 14 ar needd to list al th shwa + <l> endngs, 9 to list al th shwa + <m> endngs, 19 to list al th shwa + <n> endngs, and 37 to list al th shwa + <r> endngs, a total of virtuly 80 pajes, or wel over a quartr of th hole dictionry, and that excludes al those words in wich othr letrs folo th <l, m, n, r> (like difficult, important, history). One myt argu that post-accentul shwa represents perhaps th bigst singl problm in english spelng, certnly it causes an enormus numbr of errs.

7. Post-accentul shwa in CS.

By usuly cutng out th vowl letrs concernd, CS removes th major part of th problm, riting for instnce just <-r> in burglr, teachr, amatr, Yorkshr, actr, neibr, centr, murmr, injr, martr. Thr ar howevr som patrns wher CS canot make this cut, because th vowl letr also indicates th valu of a preceding letr. CS thus fails to solv th post-accentul shwa problm aftr soft <c>, aftr a palatized consonnt, and aftr long vowls that ar not spelt with a digraf:

- CS has to keep th <e> in grocer in ordr to distinguish it from th patrn of TO soccer, CS socr, and th <i> in pencil in ordr to distinguish it from th patrn of TO uncle, CS uncl.

- CS has to keep th <io> in ration (altho it cuts TO fashion to CS fashn) in ordr to distinguish it from th patrn of TO fatten, CS fatn. Simlrly, (io) pronounced shwa has to be kept in al th 1,000+ words ending in <-cion, -sion, -ssion, -tion, -xion>.

- CS has to keep th <e> in later, diner, super in ordr to distinguish them from th patrns of TO latter, dinner, supper, wich ar CS latr, dinr, supr.

8. NS90 cud finaly solv th problm of post-accentul shwa.

Despite th enormus advantajs in regularity and economy acheved by CS Rule 2, wich normly cuts out th letrs stranding for post-accentul shwa, th patrns of exception listd in §7 abov represent quite serius limitations on CS's achevemnt in this area. Howevr, a mor radicl reform, such as NS90, cud resolv these dificitis entirely, by riting groesr/ socr, pensl/ uncl, rashn/ fashn/ fatn/ suspishn/ invaezhn/ pashn/ naeshn, laetr/ dynr/ soopr. As it now stands, NS90 not merely dos not remove these residul dificultis of CS, but it actuly abandns th substantial gains that CS dos acheve and in so doing preservs som serius problms of TO. To that extent, NS90 is clearly incompatbl with CS, and th public wil need to be givn a convincing reasn wy this shud be so, if NS90 is to enjoy th credbility that its desyn myt othrwise deserv.

9. A practicl problm of inconsistncy in NS90.

So far this problm has been anlyzed simply in terms of inconsistncy between NS90 and CS, and th failur of NS90 to retain th regularitis acheved by CS. Howevr, even on its own terms, NS90 is inconsistnt: for exampl, defer retains <-er>, concur retains <-ur> and astir chanjes to astur. Such inconsistncis represent a major trap for both readrs and riters.

9.1 Th problm for riters.

If, as apears to be th case, TO apple/chapel ar not alynd in NS90 (as they ar in CS apl/ chapl, but left as apl/ chapel, then th riter canot tel from th pronunciation how to spel eithr word and that NS90 dos not rite apel, chapl. Simlrly with final <m>, th riter canot tel from th pronunciation wethr to spel TO rhythm/fathom as rithm/fathem or as rithem/fathm; with final <n>, th riter canot tel wethr to rite hadnt/importent or hadent/importnt, and with final <r>, th riter canot tel wethr to rite centr/ enter or center/ entr. Th contrasting NS90 spelng of th rymng endings of TO radical/article (NS90 radical/articl, CS radicl/articl) is anothr exampl of th problm. If over a quartr of dictionry words in english contain post-accentul shwa befor <l, m, n, r>, NS90 leves th riter with a vast area of orthografic ambiguity.

9.2 Th problm for readrs.

Th problm for riters cud of corse be resolvd by insertng <e> for shwa in al cases, giving apel, rithem, hadent, center. That howevr merely increses th problm for readrs, since in NS90 th letr <e> also has th traditionl valu of TO short <e> as in bet and, befor <r>, in her. This means that th readr can no longr tel how to pronounce <e> in th environmnts concernd, and th foren lernr wil stil be unable to tel from th spelng that TO to rebel/a rebel ar difrntly pronounced (cf CS to rebel/ a rebl). Worse stil NS90 creates new hetrofones, wich is th most glaring anomly of TO (as in th two pronounciations of tear in tear drop/wear & tear). If we take th TO pair deafer/defer, we find that CS distinguishs them quite satisfactrly as defr/defer; but NS90, it apears, wud spel them both as defer. Total ambiguity of this kind may not be very comn in NS90 (and it is to avoid ambiguity with TO conquer, aster that NS90 rites concur, astur), but thr wil be latent ambiguity in th NS90 forms of al th 25%+ words in english wich contain post-accentul shwa. For foren lernrs, ho canot be expectd to no th pronunciation of words as nativ speakrs do, this is an extremely serius drawbak to NS90.

10. Compatbility of NS90, CS th solution.

It is symtmatic of this previusly intractbl problm that NS itself has thru its histry fluctuated in its proposed spelng for post-accentul shwa. Now that a solution is availbl, it shud be adoptd. Riting post-accentul shwa as in CS has th foloing advantajs: it wud -

- remove a major ambiguity from NS90 for both readrs and ritrs;
- create mor economicl spelngs (e.g. rashn insted of rashen);
- enable CS to point ahed to NS90;
- enable NS90 to refer bak to CS;
- giv th Society a mor coherent overal program.

I therfor propose that NS90 shud be amendd, to incorprate CS patrns as the standrd spelings for post-accentul before < l, m, n, r >.

Back to the top.