As for stress (the place of the accent of intensity) it is evident that it is quite impossible to give one single rule that would satisfy the individual habits of all nations, for even some of the most international words are in this respect pronounced differently in different countries. E character has stress on the first, D charakter on the second, and F caractère on the third syllable; cf. also Icel. fílosofi, E philósophy, R filosófya, D F philosophíe, etc. Among national languages there are some which have a fixed rule for all words, this Icelandic, Czech, Finnic, Magyar on the first, Polish on the penultimate, F on the last syllable, while many languages have no fixed rule, so that the place of the stress must be learnt separately for each word, and even may vary from one inflexional form of the same word to another, as it does in Russia. If we study the history of languages and the reasons why changes have taken place in this respect, we find that there are three powerful principles, which sometimes work in the same direction, and at other times act counter to one another, namely the principle of value, the principle of unity, and rhythm (see my Lehrbuch der Phonetik, ch. xiv; Mod. Engl. Grammar, vol. i, ch. v.)
According to the first principle that syllable tends to receive the strongest stress which is felt as the most important for the meaning of the whole word: this is specially visible in the case of contrast and emphasis. The principle of unity is shown when a word or a group of words is held together through one strong accent, generally at the end of the group. Rhythm finally is seen to be most powerful in determining the place of a secondary stress, separated from the main stress by one weak syllable, or sometimes by two.
In a constructed language the rule or rules for the place of the accent should be as simple as possible, but we should also to some extent take into consideration the three principles just mentioned, especially where they lead to an accentuation which agrees with what seems natural according to the stress found in that or those languages from which the words themselves are taken.
It would obviously be a mistake, in a language constructed like ours, to apply strictly and solely the principle of unity, stressing everywhere the last syllable, for this very often contains subordinate elements, chiefly of a purely grammatical character, and we should thus in most cases neglect the principle of value. On the other hand, it would be equally impossible everywhere to apply the principle of value, for which syllable is the real valuable bearer of the meaning in such words as character or charlatan? In many words, however, there can be no doubt as to the importance of some syllable, for instance the first in senda send, blindi blind, sembla seem, turna, etc., the second of atrakto attraction, reforme, etc.
The fact that the majority of our words are taken from Romanic language makes it natural that the stress should lie near the end of the word, and by giving as our main rule that the penultimate (the last syllable but one) is stressed, we get what would seem to most people the natural pronunciation of a very great many words, e.g. exemple, kusino, parente, krokodilo, elefante, simbole, honore, exakti, triumfa, etc. etc., though it must be confessed that English has changed the stress of not a few such words.
There are, however, a certain number of cases in which the rule if formulated in this way would lead us into difficulties, because it is not always easy to say what constitutes a syllable. Take such words as filio son, folio, kordie heart, statue, kontribuo. Here it is possible to pronounce the combinations of i and u with a following vowel either as one syllable (in which case the stress would fall on the preceding syllable) or else as two, with i or u stressed: the former is the system of Ido, the latter that of Esperanto. To avoid such doubts it seems best to state the rule as in Idiom Neutral: stress the vowel preceding the last consonant.
But this rule will have to be supplemented by the rule that the addition of the ordinary consonantal endings does not change the stress of the main word, thus we have stress on the first syllable of patros, patron, nulum, nulim, amad and on the second of families, privatim, amusad - all in accordance with the principle of value. (Cf. below on adverbs.)
There are certain words in which the stress demanded by our rule may seem unnatural to those familiar with the language from which they have been taken, but it should be remembered that formúle, konsúle, gondóle, fasíli, kritíko, tekníke, Afríka, Ameríka, splendídi, etc., stressed in the way here indicated are not more strange that the corresponding accentuations in French, and that amfíbie, inséndie, júnie, fúrie, komédie, Itália, Fránsia, etc., would be less natural if i before the final vowel had been stressed. Then we must take galérie, enérgie, etc., with the same stress. It should also be remarked that many derived words have a secondary rhythmic stress on the very syllable which some people would like to have seen stressed: fàsiléso, fòrmulísa, kònsulàl, amèrikáno, krìtikál, etc. Filosófia, filológia, seremónia, litúrgia is the Latin stress, which was changed in France, whence the stress on i has spread to other countries.
In words like duo, boa the first vowel naturally gets the stress. But in ídee, héroo, and a few other words the stress on the vowel before the last consonant seems a little strange; idea, however, may be heard pretty frequently in English, at any rate in America.
The rules given here are simpler than those of Occ; and then we have nothing corresponding to those exceptions in Occ, which must be marked with an accent-mark over the vowel - which is always inconvenient to write, to telegraph and to print, not to mention the difficulty of remembering these special cases, for we cannot, of course, take it for granted that all learners and users of an I.A.L. know Italian and Latin well enough to know beforehand the stress of these exceptional words.