Spelling Reform Now!

The Latin alphabet was introduced to Britain when the Romans took over the island (more on the history of English); ever since then, we have used the Latin alphabet for writing English without a major orthographic change in a thousand years. Printers in England could not use letters such as edh and thorn when types were imported from the Netherlands, because those letters were not in the Dutch alphabet. Until I upgraded to Mac OS X, I could not view Icelandic characters properly in my browsers. Read what Michael Everson has to say about where thorn belongs in the English alphabet.

Pronunciation has changed, but spelling has not kept up since the time of Samuel Johnson (circa 1755). The earliest reform attempt I have found dates from 1580, in William Bullokar’s The Amendment of Orthographie for English Speech. Although Noah Webster achieved some success with his simplifications of some British spelling patterns (partial rectification: e.g., color instead of colour and dropping the last two letters from words like programme), the increasing international use of English makes spelling reform all the more sensible.

One thing I hope to see is an English edition of the Bible using a reformed orthography such as NuSpel, Michael Avinor’s Latin-Hebrew alphabet or the Simpler Spelling Association phonemic alphabet, refined by type designer Frederick W. Goudy. Any reform should be used first as a pronunciation guide for foreign words and names; the current orthography should not be replaced in everyday writing until the public is ready. At present I favor NuSpel, although the letters a, e, i, o, u should ideally be limited to their sounds in pat, pet, pit, pot, putt. I have created two tables for my preferred additions to the Latin alphabet. I have also transliterated a comic strip from traditional English spelling to NuSpel.

Which Orthography Is Best?

• Some systems are restricted to the present 26-letter alphabet, either by eliminating extra letters (e.g., Cut Spelling) or by using groups of letters to represent some phonemes (e.g., World English Spelling).
• Augmented alphabets, such as NuSpel, Unifon and Isaac Pitman’s Phonotypy, add letters to the current 26. I discourage the use of figures (like 2 or 5) or currency symbols to represent phonemes. I also doubt the validity of using the two forms of lowercase a for different vowel sounds, as is done in NuSpel and the SSA alphabet. The Latin-1 character set already provides enough characters to write English phonemically, but most browsers for Mac OS Classic do not display all of the characters properly — especially the Icelandic letters edh and thorn. Use of Latin-1 as a general-purpose English alphabet would not require digraphs, but it would use accented letters for several vowels.
• Non-Latin alphabets, such as Deseret, Shavian and Ewellic, have entirely new letters and often no distinction between uppercase and lowercase. These alphabets look interesting at first, but fail to catch on. One significant reason is the lack of an established sorting order for the letters, although the Deseret and Shavian scripts are now part of the Unicode Standard.

Creating a Font

I have made a NuSpel font, using Ares FontMonger, based on the Futurex typeface by Fred Nader (a/k/a Apostrophe) and Graham Meade. I would also like to make NuSpel or Goudy/SSA versions of the Chalkboard, Chicago and Sand fonts. Now that I am using Mac OS X, I plan to try to create fonts using an open-source TrueType font editor.

I should like to see NuSpel supported by Unicode, but I recognize some significant issues:

• The current NuSpel keyboard layout assigns new letters to the top row, moving the numbers to Option-number positions. The 15 new letters, in upper and lower case, should instead be assigned to expendable positions elsewhere on the keyboard (e.g., Option-letter positions on Mac). I would also change the NuSpel letter for the consonant in thigh from Greek theta to Icelandic thorn.
• I do not know the extent of the NuSpel user base, or whether it would justify adding NuSpel to Unicode.
• Where to put the eight uppercase/lowercase pairs of NuSpel letters not already in Unicode (seaspeak names Uniform, Awesome, Amazing, Indigo, Cello, Shadow, Thinker, Zsa Zsa)? The most logical place would be codepoints U+E830 to U+E83F in the ConScript Unicode Registry.

My conclusion is that spelling reform for the English language is best achieved in the long term with an augmented Latin alphabet. My recommendations for regularizing English spelling:

• Introduce five “new” consonants: Icelandic edh and thorn, and IPA esh, eng and ezh (pending availability of a TrueType or OpenType font that contains those characters; picture available). Confine the letter c to its sound in cello; the letters q and x may be retained for use in scientific names of plants and animals, as well as in un-anglicized Latin expressions.
• The alphabet should be extensible, to accommodate sounds in foreign names and words.
• If this alphabet is adopted for everyday writing beyond pronunciation guides, letters should be distinct in various styles: printing, script, computer displays, etc.
• The letters a, e, i, o, u should be limited to their sounds in pat, pet, pit, pot, putt. The usual turned e is recommended for schwa. Other vowel sounds should be written as in World English Spelling until new vowel symbols can be developed.
• While lowercase letters should be the norm for text, uppercase letters do have their uses: e.g., titling and marking proper nouns.
• Eliminate letters not relevant to pronunciation — e.g., enuf instead of enough.
• Write f instead of ph and j instead of soft g.
• An unstressed vowel before syllabic l, m, n or r need not be written — e.g., lisn instead of listen. Exceptions: korral (stress on second syllable), baron, forum, vilun and others.