web hosting short URLs photo sharing

 

An Alphabet for English Table of Contents Use the back arrow on your browser to return here

History of Spelling The Goals of Spelling Reform Cut Spelling Multiple spellings of one phoneme /u:/ The problem with traditional spelling Options for orthographers Merging phonemes in Nu Folik Evaluating New Alphabets for English -  Defining the goals of a writing system

The Problem English can't be spæelt. --G.B. Shaw

Traditional English Orthography (TO) is unnecessarily difficult, confusing, and inconsistent. In TO, there is usually only one lexically correct way to spell a word but dozens of ways spell most sounds (way, wave, weigh, wain, ..). In addition, a given spelling [such as -ose] can be pronounced more than one way. eg, DOSE, ROSE, LOSE | (IPA) dous, rouz, lu:z. TO is both polyvalent and irregular. TO was standardized (ca. 1750) before it was regularized.

In the 17th and 18th Century, there were a series of small reforms, some directed toward morphemic regularity, that culminated in standardizing English at the word level. Even during this period of flux, regularizing smaller linguistic units was considered impractical. Now that English spelling has petrified, it is even more difficult to finish the job because using consistent phoneme-grapheme correspondences will change 60% of the standardized TO spellings. Some univalent phonemic spellings may look more like TO than others but, at best, a phonemic reform is a 40% solution. There is no way for a univalent orthography to look likean irregular polyvalent one.

A phonemic reform is at best a 40% solution to the problem of minimizing visual disruption

If the goal is to make the reformed spelling resemble TO beyond a 40% solution, one has to compromise the alphabetic principle. Solutions that will minimize visual disruption must incorporate some polyvalence. This can be done three ways: (1) systematically with positional spelling, (3) unsystematically by accepting a fixed number (usually about 50) very common irregularly spelled words (such as the) or (3) by clipping redundant letters.

Cut Spelling (Upward, 1996) does not address the problem of inventing the spelling of a sound. Cut spelling rules are fine when it comes to eliminating silent and redundant letters. This non-phonemic approach may effectively solve about 75% of the problem with TO. Deleting surplus letters retains the basic word patterns and the "look and feel" of TO. It also retains much of the inherent ambiguity of TO. Cutting redundant letters simplifies the task of spelling but it does not solve the problem matching dictionary spelling.

It is not difficult to come up with a consistent set of 50+ phoneme/grapheme correspondences using only 26 letters. The problem is in coming up with any phonemic notation that resembles TO. One can consistently use one of the many TO spelling patterns for a particular sound but the end result will be a system that is incompatible with TO 60% of the time. 60% of the spellings will appear odd or strange and this will cause some degree of visual disruption and hesitation. The best one can hope for is a system that can be read without a code book or key.

Options for Creative Orthographers

The basic task of the orthographer is to assign shapes to sounds. They come up with graphemes to correspond to the phonemes of the language. Most orthographers

1. use the categories identified by linguists as significant sound segments (or phonemes)
2. use graphemes that are included in the ASCII character set, and
3. try to avoid disrupting established TO reading habits.

This does not leave much room for creativity. When there are no restrictions, it is a relatively easy task to assign graphemes to previously identified sound categories. Children have shown that they are capable of creating such codes.

When all of the current proposals for orthographic reform are compared side by side (See the letter matrix below), most of the disagreements are with the representation of the long vowels and the sounds that are typically represented with digraphs in English. There are three schools of thought on this:

(1) Use a letter or symbol to mark the difference between a long and short vowel

(2) Follow IPA (The notational system first developed in 1889 by D. Jones, P.Passy, and the International Phonetic Assoc.)

(3) Pick the most frequent spelling pattern and use it exclusively

These distinctions can be used to distinguish the New Spelling solutions from the New Follick or IPA-Eurospell solution: New Follick uses two types of symbols: one for an extender and one for the /uh/ and /hook/ sound.

NS and NF do not exhaust all the possibilities. One could also use symbols, such as an upper case letter or the umlaut, and follow all other TO conventions. ANJeL and Monofon use the upper case marker. Valerie Yule has suggested such a notational system where the long vowels [ ei I: ai ou iu ] are represented by ä ë ï ö ü .

New Follick could not move to this kind of (1B) representation for two reasons. (1) New Follick is an IPA-ASCII notation and the umlaut is not part of the available character set and (2) ei is a kind of e, ee is a kind of i, and
ai is a kind of a. So, instead of ä ë ï ö ü, New Follick unigraphic notation would read ë ï ä ö ü. 

The economy of NF notation can be observed by adding R endings. As shown in the following table, when this is done (eir and er) are merged into one short two-letter representation. Two potentially troublesome allophonic mergers are highlighted. Using ar for parachute slightly distorts the pronunciation. The apostrophe ['] is not the best symbol to use to start a word, so when this phoneme begins a word such as *EARN the letter [ u' ] is used. U'r is equivalent to eur in NS. *AGO is written a'go. *er for *AIR works, but there could be some dissonance caused by the TO word *PER [p'r]. Where readers are likely to use TO pronunciation, pe:r could be used instead of per.

In English, the terminal <e> is frequently used to distinguish /ei/ [ate] from /ae/ [at]. New Spelling uses the <e> as the long vowel marker but moves it next to the vowel: [ateà aet] as in [ Hee aet at Joez delee. ]

The first letter of the alphabet in English, pronounced /ei/, is not a pure vowel. It is a diphthong or a blend of two vowels. In fact, almost all of the so-called "long vowels" are diphthongs. Following IPA, the /ei/ sound should be associated with [e] not [a]. [Hi: eit æt Dzhouz deli:. IPA] [Hi e:t at Jo'z deli. NFc ]

Simple allophonic mergers (e.g., eer = ir) are difficult with New Spelling notation. The simplicity and economy of the NF solution is illustrated in the following sentence. 

Four New Alphabets for English

This paper presents four original notational systems for English. To be more accurate, it presents three refinements of earlier notational systems and introduces two versions of an original system [mono'fo'n] .

 

1. New Follick is a refinement of Follick (1945). It is also very close to Lindgren's Phonetic A (1969). New Follick can be considered an IPA-ASCII script because it employs only ASCII characters and because there is a one-to-one correspondence with the IPA symbol set. Since IPA uses a special font, ASCIIbets such as NF are needed to show pronunciation in text documents. Unlike other ASCIIbet proposals, NF looks good in print and can be handwritten with ease. NF uses continental sound values and could be readily understood by those literate in Spanish, Portuguese, or Italian. NF tries to be a bridge to TO but the allegiance is to IPA and Euro-Spelling: /ei/ = e: (Follick ei) not A or AE, /i:/ = i: (Follick ie) not EE. /ai/ = ai or `i (which looks like y), not IE (Follick ai). /ou/=o' not OE (Follick ou). /u:/ = u: not oo or ue (Follick oo) 2. Monofon - is a notational system that has no digraphs. The name means one and only one symbol for each phon or sound. To represent 40 sounds with a single character, at least 18 new letters are required. In this respect, it is similar to John Malone's Unifon (1959). There are two monofons: keyboard and pictographic. Like Unifon, Pictographic Monofon requires a new augmented font. Keyboard Monofon, doesn't. It does away with caps and assigns the 18 new sounds to shifted characters ( /ei/ = A, /I:/=E, /ai/ = I … ). Vowel notation in keyboard Monofon (1994) is amost identical to Beach's ANJeL (1996). Pictographic monofon character shapes resemble their names: [a] avian, [q] ax, [ É ]ox /ahx/, [^]up, ['] acute /uh/, [é ]flag/valley, [­ ] spear tip/ [¯ ]digit. 3. Menu Spel is a refinement of the pronunciation guides found in many cookbooks and exotic menus. It is supposed to be unambiguous and intuitive to those schooled in the traditional English orthography. e.g., toh-may-toh or tuh-mah-toh. Vowel sounds are represented by two or more letters. ay, ee, ah-ee, oh, yoo. This reduces ambiguity but makes the script impractical as anything more than a pronunciation guide. In this respect, Menu Spell is much like True Spell. [pizza pie = peet-zuh pah-ee]. [ Theater = thee-yay-ter ] 4. DBLspel (positional spelling) is an attempt to achieve a closer approximation to TO by allowing more than one sound per spelling and more than one spelling per sound. The result is not phonemic, but it is systematic and fully predictable. The positional spelling can be reproduced by a computer or by anyone who understands the rules. Cut NF is a mild type of positional spelling where redundant markers are left out or clipped. [si: = si *see]. Word patterns generated by Cut NF are very close to those in Cut Spelling. John J. Reilly's Alt.Script is the most serious attempt to mimic English TO by having a different grapheme or digraph for the initial, medial, or terminal position in a word. [ Highly = hyly ] This word shows 2 of the 3 sounds for [y] in Alt.S. 5. Checked Clipped Spelling (CCS) Chekt Spelng takes advantage of the fact that all checked vowels are followed by a consonant and are short by markeing the checked vowels rather than the standard practice of marking the extended vowels. CCS is essentially New Follick with different marking conventions.

 

Consistent Representation

It is not difficult to come up with a consistent representation of the 40 phonemes found in spoken English. In fact, over the years, there have been over 70 proposals to regularize English spelling. The problem is to find a way to make it simple for those who have devoted over ten years learning the difficult irregular traditional system. Once learned, one writing system is as easy to read as the next. Writing systems primarily differ with respect to how long it takes to learn them, how well they show pronunciation, and how easy they are to use as a tool of communication.

No system that has unsystematically evolved over time is as simple as it could be. All cultural systems tend to be more complicated than they need to be. In language, there is often a natural tendency to change in the direction of simplification.. Writing tends to preserve complexity, older conventions, and lag behind the changes in speech. 

---

On this server (relative links which may or may not work) [Other Sections of PV-7]
• The Problem with Traditional Spelling   theproblem.html
• Evaluating Alternative Notational Schemes   evalalfa.html
• Go to a bibliography on spelling reform
• Go to next section    art-alf4.htm (bad link)
 

---

Absolute References to another server
• Go to Spelling Reform at delphi.com   http://www.delphi.com/spelreform
• Go to spelling link page
• Go to spelling link page   http://www.pages.whowhere.com/community/sbett/spel-links.html

Go to Spelling Reform Message Board