Notes on the Spelling
The 'transliteration' was spelt in accordance with certain guiding principles that had to be laid down in advance. Though it is claimed that the decisions taken were wise ones, there is nothing binding about the resultant spellings; it is merely proposed that the spellings here shown be looked upon as standard, unless and until others come to be widely preferred, and when good reasons can be found for making a change.
(1) It is desirable that a given word should appear always in a given spelling and not vary from time to time. (This does not preclude individual writers from regularly using some spellings that differ from those in Androcles; it merely recommends consistency.)
(2) It follows from (1) above that a choice of possible spellings has to be made in the case of those very common short words that are differently pronounced at different times by one speaker -- those having what are called 'strong and weak forms'. The decision was taken in principle to spell such words with their fullest pronunciation (since reduced forms can always be derived from fuller ones, whereas the converse is not possible). For the two kinds of exception to this, see (3) and (4) below.
(3) WORD-SIGNS. The design chosen to be the Shaw Alphabet has the characteristic feature incorporated in it of four 'word-signs' for the four most frequently occurring words of the language -- the, of, and, to (it is estimated that one word in six is either the or of or and or to). These word-signs each consist of a single letter -- that for the single sounds of th, e, n, and t respectively. The word-signs save valuable time and space.
(4) THE INDEFINITE ARTICLES. The words a, an are here transliterated not to rhyme with day, Ann (which would be their 'fuller' pronunciation), but with the central, neutral, or shwa vowel actually heard in 'a man', 'another'. This has the advantage that the two words a, an can then be spelt with the same vowel -- which would not otherwise be the case. Moreover, the 'fuller' pronunciation of these two words is hardly ever used. This constitutes the second exception to the principle in (2) above.
(5) Many English words have alternative pronunciations, each speaker generally using one of them consistently, e.g. azure, subsidence, acoustic, controversy, laboratory, and countless others. Clearly, the principle in (1) above required that a choice be made. In general, individuals are of course at liberty to spell in conformity with their own pronunciation. Alternative standard spellings of such words are likely to emerge; but until they do, the spellings in Androcles may be taken as standard.
(6) It is obvious that the spellings in Androcles will fit the speech of some English-speaking people better than others. Nevertheless, it is claimed that none will find it hard to read from the spellings shown, i.e., to get the meaning from the printed page. It is to enable the greatest number of people to read from the spellings easily that words are in general written out in their fullest form (see (1) above), especially since most readers of Shavian are already readers of English in Roman letters, and since this will be their first experience of reading English in the new script.
(7) It is for the reasons given in (1) and (6) above that the letter R is transliterated wherever it now occurs in Roman. The non-pronunciation of R in certain positions, which is characteristic of certain types of English speech, can easily be inferred from the spellings shown here –- as it is now from our traditional orthography; but it would not be possible to deduce the pronunciation of R from a spelling which did not show it. Here again, the fuller form of words is the one shown, thus incidentally making the transliterated spellings more acceptable to, because conforming more closely to the speech of, a much larger number of speakers of English in all parts of the world.
(8) Even so, the spellings in Androcles, while not committing anyone to specific qualities of sound (since each reader will read his own qualities, e.g. of vowel sound, into each different letter), do nevertheless commit to a particular distribution of sounds, and this distribution may be at variance with the usage of different speakers, not only with respect to the alternative pronunciations within a given type of English (see (5) above), but as between the usage in the various areas of the English-speaking world. It is probable that, for example, American writers would favour other spellings in a number of instances, and that therefore further alternative spellings of some words will emerge. These are not likely to interfere greatly with the intelligibility of a text. It is in any case fitting that this first publication in an alphabet constructed in accordance with Shaw's wishes should show spellings in conformity with the kind of pronunciation he thought should be represented.
NOTE: It would be possible to extend the number of word-signs beyond the four provided for in the design. Thus, common words such as the following could regularly be spelt with a single consonant (the corresponding Roman letter is shown in brackets after each word): for (f), be (b), with (w), he (h), are (r), so (s), do (d). Further economies could be made by writing other common words with two letters, omitting the vowel between initial and final consonants, e.g., that (tht), was (wz), have (hv), not (nt), this (ths), but (bt), from (fm), had (hd), has (hz), been (bn), were (wr), and so on. If such spellings became standardized, these invariable written forms would stand equally well for strong and weak forms in pronunciation, each reader supplying whichever he found appropriate in the context (which is what he does now). Naturally, it would always remain possible for a writer to indicate, by spelling out in full, any particular form he wished -- to avoid ambiguity, or for the sake of emphasis, or in order to specify, for example in stage dialogue, some particular reading. For the reasons given in (6) above, Androcles has been transliterated without any abbreviations save those mentioned in (3) above. It is possible, however, that other abbreviations would come into use for private purposes but not for printing; it is also possible that some might come to be adopted in print as well.
PETER MACCARTHY
The University
Leeds
1962
