The greatest mischief-maker in European languages is the letter c. This is owing to the historic fact that the sound for which this letter stood when the Latin alphabet was framed, namely [k], in course of time has changed very considerably, and has changed in different ways in the various countries into which the Latin language and the Latin alphabet penetrated - but the spelling has generally been kept unchanged. Where the old Romans pronounced two [k]s in such a word as circulus, only the second c has kept its sound, while the first, like any c before i and e, was drawn forward by these front vowels, the result being that the initial sound is now [tS] in I circolo, [phonetic symbol] (pronounced like E th) in S circulo, [s] in E circle, F cercle, Sc cirkel and P circulo. Before becoming [s] the sound in F was [ts], and this pronunciation was taken over into D (where the spelling is now zirkel) as well as into Slavic languages: in these latter the same value of the letter c was extended so that c is even used before a, o, u and consonants: in Polish the Russian tsar is spelt car, the D loan-word zug (`draught, train') is written cug, and sugar (D zucker) cukier; the name Potocki is pronounced with [tski]. Similarly in Czech.
Now, what are we to do in our constructed language with this rebellious letter? So far as I can see, there are the following possibilities:
(1) We keep c everywhere and give to it its ancient value of [k] in all positions. This seems to be absolutely impossible in spite of the fact that Latin is now taught in many schools in various countries with this pronunciation, Cicero being pronounced as Kikero, etc. But who could seriously think of pronouncing kivil, kirkle, kentigram, etc., in ordinary life ?
(2) C is kept before a, o, u and consonants, where it has the value [k], but is changed in spelling to s before e and i. Thus we could write corespond, canon, sircle, sentre, etc. This is not at all a bad solution, as it would not lead any West-European or American into the temptation of mispronunciation; still I think this way out of the difficulty far inferior to the one proposed below. It has the inconvenience of having no less than three letters for one and the same sound: c,k, and q, and of causing difficulties with certain derivative endings: from critic, msic, fisic (physic) we must be able to form words in -ere, -iste: but how is the c to be pronounced before these endings or in scepticisme ?
(3) C is kept everywhere, but is given two different pronunciations according to the letter following it. This would preserve the spelling known from E F S in most cases, and that of course is an advantage. Nor would there be any difficulty with the [k]-pronunciation, but should c before e and i be pronounced [s] or [ts], or how ? In Occ this system is followed with the pronunciation [ts]; and not only has the letter c thus two pronunciations, but the same root takes automatically two forms according to the derivative ending used; thus we find electric electricità, public publicist publication publicmen, etc. The reason alleged is that it is in this way possible to arrive at a great many forms found in existing languages by means of "regular" derivation - this, however, only means that the irregularity is hidden away behind the duplicity of the spelling or is shifted onto the rules for pronunciation. In order to have the regular formation simplicità Occ even is obliged to have a form simplic with a [k] for the adjective (and simplicmen for the adverb), which is found with that pronunciation in no existing language - and not only this, but Occ also has another form for the adjective: simpl, which is demanded on account of the word simplifiction, which also has to be formed regularly. Such is the simplicity arrived at when one follows consistently the inconsistencies of national languages.
Another objection to Occ in this respect is that the pronunciation [ts] given to c is not the one used for this letter in the most widely known languages; a form like simplicità with its ending and accent-mark would even downright tempt many people to give c its Italian sound.
(4) K is written for c when so pronounced; otherwise the letter c is used. The pronunciation of c is thus one and the same in all positions, but might be either [s] or [ts]; in the former case (as in the corresponding case sub 3) we should have two ways of writing the same sound, which might cause difficulties for those who are not familiar with traditional European spelling and who learn the I.A.L. orally - a class of people which is negligible nowadays, but must be counted with in the future, if our I.A.L. is really to spread according to its idea. The second pronunciation, [ts], is the one given everywhere to c in Esperanto and Ido, in the formerly undoubtedly on account of Zamenhof's Polish extraction; and in both languages c is used extensively with this value not only before e and i, but also before other vowels. This is in many cases a direct consequence of the structure of the language with its frequent grammatical endings o, a, as, os, us; as all substantives end in -o, we must have not only princino princess, but also princo prince, paco peace, further verbs like intencas, intencos, intencus (various tense of the verb for `intend'). In Ido we have the demonstratives ca, co, taken from F ce, but with a pronunciation and endings not found in F, and further a great many verbs like formacar, importacar, where the sound [ts] is taken, curiously enough, from the Latin (F E etc.) ending -ation, which has no place as such in the system. This use or abuse of the letter c is one of the most striking features of this group of constructed languages, but, as already remarked, is somewhat repellant to most people outside the Slavic world.
If c is pronounced [ts], and this pronunciation is insisted on everywhere, some international words have to be pronounced with harsh consonant groups, e.g. science (Z has scii for `to know,' with sts), except, etc. Ido alleviates these groups and has such forms as cienco, ecepter, ceno scene, etc. But these forms are not quite natural, as they are found with that pronunciation in none of the national languages, and it would be in better conformity with what exists already if we simplified them into sientie, exept (eksept), sene, etc., as actually done in the received pronunciation in E F Sc and other languages. This means one sound less than Ido has, in the sc-words, and the group [ks] instead of [ts] in xc. But this leads us to the last possibility.
So far as I know, it has not been proposed to use ç instead of c and otherwise keep the Esp spelling; nor would much, if anything, be gained by this clumsy expedient.
(5) Instead of c we write k and s according to circumstances, thus kanon, kultur, obstakle, sirkle, sivil, serebral, sentre, aksept, oksidente, etc. - further sene, septre for scene, sceptre, etc. (and arkeologe, skole). This spelling has the advantage of expressing unmistakably and according to one simple rule that pronunciation of all these words which is best known to the vast majority of speakers (E F P Sc besides Andalusian and American Spanish), and it will offer no difficulty to those accustomed to other pronunciations of the letter c, as k and s are everywhere known with the values here given to them.
Those who object to such spellings, which at first blush may seem barbaric, may perhaps be pacified by the following considerations. Here and there we already find c and s alternating both within one and the same language and when two or more languages are compared. In E we thus have side by side licence license, practice practise, prophecy prophesy, pace pass. How many English people know exactly the difference between council and counsel ? E has c, where F has s, in many words: dance, defence (cf. also E defensive, defensible), offence, ace F as, juice F jus, vice F vis the instrument. Cf. also F E race, D rasse (I razza, S raza). Italian has sigaro and Dutch sigaar for cigar. That c is not sacrosanct in words of Latin origin is shown by the official D spelling zivil, zirkel, etc. Compare further E eccentric, I eccentrico, F excentrique, and E ecstacy, F extase.
In favour of the spelling here advocated I may also say that if -o and -a are to be marks of the two sexes, as will be proposed below, we could not keep c in such a word as prince, for then we should have princo prince, princa princess, which are impossible unless we pronounce in the Zamenhofian way. Prinso and prinsa are much better. In the same way Franso a Frenchman, Fransa a Frenchwoman; and with other suffixes, for instance pasal, pasosi, adjectives from pase peace, etc. Everything hangs together in such a language as the one we are here constructing.
Very often, where Ido has c, it is best to reintroduce the L ti, e.g. tendentie, silentio, natione, sientie, pretie - with the ordinary pronunciation of t, not with [ts] or [S]. in some of these ti is found in derivatives in some languages, e.g. D pretiosen, Dan pretiosa.