Phonozoic Text Archive, Document 103

The Phonograph

Nature, Vol. XVII, March 28, 1878, p. 423

WE shall be much obliged if you will allow us to draw the attention of your readers to a curious fact which the phonograph has allowed us to prove, and which we announced last Monday at a meeting of the Royal Society of Edinburgh.  We have seen no mention of the fact elsewhere.

            Not only are vowels unaltered by being spoken backwards, but the same fact is true of consonants.  Whether the pulsations of air be made in a given order or in the reverse order the ear accepts the sound as indicating the same letter.  This is true of all the simple vowel sounds and of all the simple consonant sounds, including of course several combinations which in English are spelt with two letters, as th or ng, but which are really simple consonants.

            We tried the experiment on single pairs of syllables separated by a single consonant, as ada, aba, aja, ete. [sic]  A person coming from outside and ignorant of what consonant had been spoken was able to identify the consonants quite as well backwards as forwards.  The chief difficulty was found in distinguishing affa from assa

            We find that this peculiarity is not limited to consonants between vowels, but that ab said backwards becomes ba.  We have here a standard as to what does really constitute a single letter or element of articulate speech; it is any one reversible part.  Your readers who possess a phonograph may most easily verify this observation by saying a word backwards, and hearing the phonograph say it intelligibly forwards; for instance, noshäeesossa produces association beautifully.

            We shall be glad to learn whether this fact has been already published, and also whether it was foreseen as a possibility by any writer.



Edinburgh, March 25