Sounds and SpellingEdit
Before dealing with the sounds themselves, it will be well to say something of the alphabet to be used in our international language. Most people will at once agree that it would be quite impossible to take the Russian, Greek, Hebrew or Arabic, or any other Oriental alphabet, in a language which must first and foremost be destined for use in the West-European and American world. Nor would it do to try and invent a purely phonetic alphabet independent of the Roman letters, although it would be possible and even fairly simple to devise an alphabet that would be simpler and much more systematic than the traditional Roman alphabet with its many deficiencies (nothing in the shape of the letters to show the parallelism between the pairs p:b, t:d, k:g, f:v; no sign for the simple nasal spelt n in finger, sink, and ng in singer, sing, etc.). There can be not the slightest doubt that we must take the Roman alphabet such as it is, simply because it is already used by the greatest number of people, at any rate of those who are likely to use an international language, and who are thus saved the trouble of learning a new one.
This of course means taking over also a certain number of orthographic habits, such as the differentiation of small and big (capital) letters, though the use of such pairs as a and A,n and N, etc., really means an unnecessary complication, and is, strictly speaking, an infringement of the principle one sign for one sound. There is also some disagreement between various nations as to the extent in which capital letter are used: while all nations agree that they should be used after a full stop and in proper names in the strictest sense, some nations use them also in adjectives and verbs derived from proper names (E French, F français, etc.). Such details, however, are of no importance and need not be discussed here. But all constructors of artificial languages seem to agree in rejecting the specifically German practice of writing all substantives with capitals.
If it is our endeavour to use as many international words as possible in the shape that most readers and speakers would readily recognize, we shall soon discover a great number of difficulties, for even in the most international words the orthographic and phonetic habits of various nations go widely apart. The task of selecting the best forms is therefore to be likened to a dance among eggs: you are sure here and there to jar upon the feelings of someone, and people's feelings in these matter are not to be lightly treated as prejudices. If we compare the written forms of E F nature, D natur, I S natura, it is easy to see that natur with or without a final vowel is perfectly international, but what would have been the result if we had had only the spoken forms to compare? Then we should have discovered that n and perhaps t were the only elements common to all forms. In this case, and in many others, the solution is obvious: we must take the international spelling and give to each letter the most international value: we must therefore pronounce natur as in D I S and leave aside the E vowel in the first and the F vowel in the second syllable. But we cannot in every detail follow the traditional spelling of words, even if it is fairly international, for too great a number of irregularities would make the language too complicated - and we must remember that a great many people have to struggle all their lives to master the orthographic intricacies of their own language.
In the dilemma arising from the tendencies to keep traditional spellings and yet to have easy rules for pronunciation it is perhaps well to keep in mind the fact that a change in spelling does not trouble readers very much, provided it goes in the right direction, i.e. gives well-known values to well-known letters and uses them consistently where the ordinary spelling is inconsistent. English people will therefore have little objection to such a spelling as tsar for czar or skeptic (or even skeptik) for sceptic, but scin for skin would be intolerable. We should therefore avoid irritating the ordinary West-European by telling him that ca and co are to be pronounced as [tsa, tso]:¹ such deviations from his time-honoured habits would only be palatable if they could be proved by unanswerable arguments to be absolutely necessary - which is not the case. Esp caro will be taken by the uninitiated to mean either `dear,' or `meat,' but it stands for `tsar.'
The spelling must be as regular as possible; there should therefore be no mute or silent letters: English and French learners must be specially warned that all written final e's should be clearly sounded with the same vowel as in the middle of a word. For E rhythm, D rhythmus, F rythme we must write ritme (two syllables; cf. the I spelling ritmo). Words beginning with th should be written and pronounced with t: teatre, teologia, teorie, tese; also in the middle of a word: entusiasme, sintese, E synthesis. In all such cases the sound [t] is found everywhere except in E and Greek.
Another orthographic simplification that is self-evident is f for ph: filosofia (as in I Sc), sfere, sofa (F D sometimes sopha).
Where ch represents Greek kh, it should similarly be written and pronounced k: kaos, kimere, Kristo, kronologia, arkeologo, skole, etc. Karaktere = E character, F caractère, I carattere, S caracter, D charakter, Sc karakter.
The Roman alphabet should not be further complicated by the addition of new letters or diacritical marks over or under the letters. It is true that some such signs are used in some of the national languages, such as the F cédille (ç), accents over vowels, č and š in Czech, n con tilde (ñ) in S, etc. These signs are not used universally, nor are they found in all printing-offices, and some of them are used for very different purposes; thus accents over vowels sometimes indicate stress, sometimes length, sometimes different vowel qualities (F é, è), and sometimes different meanings of totally homophonous words (F ou où). Two dots over o (ö) are used in E to denote separate pronunciation (coöperation, also spelt cooperation or co-operation), in D to denote the rounded front vowel, similarly in Hungarian, which even has the variant with two strokes instead of dots to denote the long vowel. All such signs present unnecessary difficulties in writing, in telegraphing and in printing. Printing with extra letters implies greater expense, and it is clear that an international language should be easy to print without extra cost in any, even the smallest, printing-office and in any type whatever (roman, italics, heavy type, small capitals, etc., all the various forms used in advertisements, etc.).
It was therefore the greatest blunder in the history of auxiliary languages when Zamenhof in 1887 created his Esperanto with a number of consonants surmounted by circumflexes: he probably took the idea from Czech, only he made č and š into ĉ and ŝ, and put the same marks over g, h, and j, thereby showing, by the way, that he was no phonetician, for phonetically there is no parallelism between the modifications intended by these signs, which are thus less systematic than the Czech letters. There can be little doubt that the spread of Esperanto might have been many times greater, had all sympathetic newspapers been able without trouble to print specimens and articles in Esperanto. In 1894 Zamenhof saw this clearly, and was even for a moment inclined to abolish these letters, but unfortunately he was later on obstinately opposed to that as well as to other reforms in his own creation. If he and his followers in 1894 had carried out the reforms he then proposed himself, we should have been further than now by some decades in the fight to convince an indolent world of the feasibility of an auxiliary language.
Occ uses an apostrophe after n and l to denote palatalized pronunciation ("mouillé"): this is not so bad as Esperanto's circumflexes, as it can easily be printed; but the sounds are not quite easy and are employed in an inconsistent way. Nothing is gained by writing attin'er for 'attain' - neither sound nor spelling is found in any language. It is much better to do away with these superfluous sounds. Besides, Occ pretty often uses accents over vowels to indicate the place of the stress: this, too, is something which is better left out in an I.A.L. (Cf. below on in- in Occ., p123.)
¹ Phonetic transcriptions are put in brackets [ ]. The only special symbols needed in this book are ʃ for sh in E shall and ʒ for F j, E s in measure.