[Journal of the Simplified Spelling Society, J32, 2003, pp20-22]
[Steve Bett: see Journals, Newsletters, Personal View, Web link.]

Orthographic Goals.

Steve Bett.

How should alternate schemes and orthographic proposals be evaluated?

There is widespread agreement on what makes for a good orthography but they are expressed in a variety of different ways. Blain suggests consistent, phonemic, compact, familiar, and email friendly. Bett suggested speed and efficiency of reading, writing, and teaching. Hasselquist suggests phonemicity, similarity, learning ease, and usability. While performance indexes have been developed for the listed dimensions, they are not refined to the point where two people could come up with identical relative ratings. It should be possible to first choose the goals and then select the scheme that best achieved those goal. While we can conduct polls and pick winners on subjective grounds, it would be better if we could make a more objective selection.

A fairly recent strategy of the apologists for the current morpholexical writing system, mentioned in the JSSS31 editorial, is to argue that the traditional writing system should not be judged in terms of criteria more appropriate for other types of writing systems. It is not fair, they argue, to demand greater phonemicity and consistency from a logographic or lexical based writing system.

A focus on speed and efficiency of teaching, and learning, however, is not writing system specific. Only those who do not want a system designed for universal literacy, can reject this as a goal.

Consistency without radical change.

Can a writing system that will be phonemic enough to help the slow learners and make a higher level literacy easier to achieve for all avoid annoying the already literate? Is it possible to achieve a high degree of regularity and consistency without drastic changes? This elusive goal has been sought by hundreds of orthographers over the past 200 years.

If it were easy to achieve consistency without drastic change, we would have long since arrived at a spelling system that everyone could support. [see Yule, JSSS32].

At the very least, the new orthography should be readable in context without a key. This is a variation of Bullokar's [1580] conclusion that a successful spelling reform required an "easy conference" between the new spelling and the old.

Almost any writing system can be made systematic or consistent with enough exception rules but when the number goes much beyond 10 they become difficult to teach and use.

Readable but annoying.

It is possible to be readable and yet be annoying to readers. When a 1917 article written in New Spelling was posted to the ssslist, most members considered it to be too tiring to read. In other words. New Spelling is probably too radical for a house style. If it manages to annoy those who favor reform, it will certainly not be embraced by those who see no need for reform.

Rondthaler and Lais [1987?] have made efforts to reduce the annoyance level and I need to rewrite the article in ALC Soundspel to see if they have achieved their goal.

RITE adds a few more rules to achieve a closer approximation of the traditional letter strings. This makes a RITE transcription a little easier to read but more difficult to spell. Both ALC and RITE do not respell unstressed vowels. This increases word familiarity at the expense of consistency. [See links to ALC and RITEspel webs.]

Tradspel adepts often read by word-signs so any change in the shape of the word slows down their reading and may require them to resort to sound spelling.

Similarity to the old code and consistency are often at odds. Attempts to optimize on one dimension will conflict with attempts to optimize on the other.

Consistency and phonemicity are assumed to be the qualities that will help the unlettered. Similarity to the old code is the quality that will make the reform palatable to those who have already learned one code and don't want to change their reading habits.

No matter how easy the New Spelling, most people do not want to go to the trouble of learning it unless there is a pressing need to be conversant in the new code.

I have suggested that we have a new Shaw alphabet competition not because we need more schemes for representing English but because we need to have a clearer set of agreed on goals for evaluating the schemes we have.

One reason that the society has endorsed so few new writing systems is because there is no agreement among members as to how to rank the alternative proposals. As a society, we have not tested our ability to recognize a better orthography.

Selecting the best reform orthography is not the same as selecting a house style. The house style has to be more than just readable. It has to be inviting. It should not discourage readership.

1. Goals.

The goals of the simplifiers are stated in the principles of 1876 and in the six axioms. These are usually repeated on the inside front cover of every issue of the Journal. Nevertheless, it is not uncommon for people to argue that the simplifiers have no clear goals.

The axioms are more like assumptions: In an alphabetic system the letters are supposed to represent speech sounds. When there is a correspondence between phonemes and phonograms. literacy can be achieved in less than 3 months. Writing becomes child's play.

Critics note that all real writing systems are mixes, e.g. contain both logographic and phonographic elements. A system with more logographic elements can be just as good as one with more phonographic elements. [see the arguments of Bradley and more recently by D.S. Taylor from Leeds, JSSS31, p.2]

It is probably true that adepts can read a logographic system faster than a novice can sound out words in sound writing system. However, it is also true that mastering a logographic system may take ten times longer and relegate the less gifted and less-schooled members of society to semi-literacy.

2. Hasselquist's Dimensions, 2001.

[See Web Links.] Similarity to TS [traditional spelling] is the ability of the proposed orthography to resemble the writing native speakers are most familiar. Visual similarity is a key element in backward compatibility which enables readers to access the vast store of knowledge and information from the past. The minimalist reformers also see visual similarity as a key factor in the salability of a new orthography. If the new orthography does not look like English, it would not be acceptable to the powers that be - those who have the power to advance or retard its dissemination and use. The minimum level of similarity could be defined as the point at which 5% or more of the spellings cannot be recognized. The minimum level of similarity is reached messages written in the new orthography can be read in context without recourse to a key. Shavian cannot be read without a key. Most other reform proposals can. However, some of them such as new spelling can be tiring to read until one adjusts to the new configurations.

In 1917, the house style for the spelling society publications was New Spelling. For all its virtues. most find new spelling tiring to read. Too many words have to be sounded out and associated with their spoken form in order to be interpreted. Sound reading requires more effort than logographic or word-sign reading.

Phonemicity is the ability of a proposal to reflect the sounds spoken in a systematic form of writing, such as to enable the decoding of it accurately as to meaning and sound, by all who wish to use the language, with one symbol per sound. and vice versa, considered ideal. The closer this ability approaches that ideal, the better it contributes to easier reading and comprehension, as well as to its writing. especially when compared with the horrendous difficulties encountered using the multiplicity of TS's symbolisms of English's sounds.

Learning ease is the ability of a proposal to enable a potential user of it to learn and remember the complexities of its rules, format and symbolism with the least possible effort, time and memorization required. This is especially necessary for the proper encoding of its writing as well as for knowledgeable reading, and should be in sharp contrast to that encountered with TS, where there essentially are no rules. or a near infinite number if there are, and where spelling depends almost entirely on rote memory to be even generally accurate.

Usability or ease of use is the ability of a proposal to enable easier writing and reading as compared with that considered possible with TS. This could be enhanced by the proposal's simplicity in symbolization and thus reduction in space/time required. This ability would have the additional benefits mentioned just before the first proposal. namely that of saving resources, faster typing speed, as well as being beneficial to reading by increased eve span capabilities, both of the latter saving time and having monetary value.

3. Roy Blain's list of dimensions for his scorecard for notations.

[See Links page.] Consistency. A 'same sound same spelling' relationship is the most important factor in any new spelling scheme. It shortens the learning process and simplifies English writing for the existing and future generations. Maximum points 3.

Familiarity. Since we prefer that older people can read new writings, and young people can read old writings it is desirable that the two styles of spelling are similar enough to be readable by all people. Maximum points 2.

Phonemic Makeup. The selection of 2 single letters to be joined to produce a specific sound (digraph), should be influenced by the sounds of the individual letters. Maximum points 2.

Brevity. Where a spelling is shortened without jeopardising the correct pronunciation or meaning of the word, time and space is saved, writing becomes more efficient and spelling mistakes fewer.

http://groups.yahoo.coin/group/saundspel/files/eval/phonemicity.html

Gus Hasselquist, a retired Lockheed engineer is one of the orthographers active on the Saundspel discussion group.

Archived at http://groups.yahoo.com/group/saundspel.
[See Web Links.]

4. Steve Bett's list of dimensions.

The focus on speed and efficiency can be traced back to the writings of G. B. Shaw. Shaw himself used pitman shorthand. The basic shortcoming of Pitman's phonemic shorthand was that it was not linear enough to typeset.

Speed of reading [decoding speed] All things being equal. brevity and the lack of redundant characters will increase reading speed. All alphabetic writing systems are read as "word sign" or logographic systems by the most adept. For familiar high frequency words. there is no particular advantage for alphabetic or phonemic spelling. Where phonemic spelling helps is on unfamiliar low frequency, multi-syllable words. [high frequency words tend to be short and monosyllabic].

Speed of writing [encoding and typing speed] Writing speed can be increased by eliminating redundant letters. The fewer the number of letters. the faster the speed. The brevity of the code makes a difference. Cut Spelling had 10% fewer letters than standard written English. It used up 10% less space on the page and could be written 10% faster.

Shorthand is an example of a writing system that sacrifices reading speed for writing speed. Many shorthands are not phonemic. Most shorthands cannot be typed because they are not linear. What Pitman and Twain wanted was a phonemic shorthand [brachytypy] that could be typeset. The Shaw alphabet achieved this objective. Shaw also stipulated that the proposed british alphabet [BPA] be different from the traditional Roman alphabet to insulate it from ridicule. It was spared ridicule but was also ignored because few wanted to take the time to learn it.

ENgliS is an example of a phonemic unigraphic [36 letters for 36 pure sounds] script that has been optimized for speedy typing without a drastic reduction in readability. When limited to the ASCII character set, there are only two diacritics available: double letters and capital letters. ENgliS rejects the digraph option. "ENgliS iz cptimYzd for rqpid tuC tYping." [c=ox horns, q=ax, S=prow of a Viking ship, E=eel, Y=eye, C=chin]. ENgliS can be ridiculed but there is little chance that it will be mistaken for the traditional script. It would not be called ignorant spelling but Mark Twain could still call it "ugly"

To reduce the uglyness of caps in unfamiliar places, the caps could be converted into diacritics. Downsizing without marking would mean that the sound associated with vowel letters would be ambivalent. But, two sounds per letter is certainly better than 14. [Dewey, 1971 showed that the sounds associated with each letter ranged from 1 to 30 with an average of 14.7 sound associates per letter].

Speed of teaching/learning. There are two distinct groups that have to be learn the code. Illiterates who are unfamiliar with any code and those who are literate in at least one code. It is easier to teach someone a code for a language they are familiar with than a code for one they cannot use conversationally. In all cases, it is easier to teach a consistent code than an inconsistent one.

It should be up to ten times easier to teach a writing system that spelled 40 sounds in 40 ways than one that spelled the same 40 sounds 400 ways. The traditional writing system spells the 41 sounds of spoken English 561 different ways in a 72,000 word dictionary. Over 85% of the spellings in English are spelled one of five different ways.

Theoretically, the traditional English writing system should take 5 times as long to learn as the ideal writing system. There are no ideal writing systems, but 95% of them, according to Laubach, are highly phonemic. If the ideal is 100% phonemic, then Finnish and Korean [Hangul] are in the high 90's. Italian and Spanish in the mid 80's. Using the same measure of phonemicity: phonemes/phonograms, English is 7% phonemic. Making adjustments for irregularities that exist in other writing systems such as Spanish and Italian, English is hypothetically over 4 times more difficult. This is about what is observed in several cross cultural comparisons where late 1st year students in Italy, for instance, perform at the level of 4th year students in England and America.

Resiliancy: The ability to withstand the outrage of those who have learned the word-sign code. No matter how good the new code may be, it has to minimize the annoyance to tradspel adepts in order to soften the backlash. People become very attached to their codes. Hofstadter said most would rather change religions than orthographies. The attachment transcends the work involved in learning a new code. Those with an excessive attachment bordering on worship and reverence will rise up in organized opposition when they feel threatened. The first line of opposition is ridicule. The new code will be called "dumbed down" English, ignorant hillbilly spelling, or worse.

When the traditionalist mounted an editorial campaign against T.R.'s respellings, there was no effective response from the simplifiers. There was a government publication on simplified spelling that explained in detail the rationale for the 300 spellings that were to be the preferred spellings in government publications. The reaction was not so much to these respellings since they were already among the variant spellings in a dictionary as to what they might portend.

5. Summary.

Selecting the best compromise notation requires a focus on the scales to be employed. The dimensions of evaluation include:

Phonemic regularity: more alphabetic and consistent than the current system and a more reliable guide to pronunciation.

Economical: Greater space efficiency - no redundancies.

User Friendly - Easier to learn and use, keyboard friendly.

Convergent sign design: backward compatible to provide access to our heritage of print and to be readable by TO adepts without having to use a key or a code book.


Back to the top.