The Nature of Vowel Sounds

Abstract

IF you will permit me to refer at this late date to Prof. Scripture's articles in NATURE of January 13 and 20 last on the nature of vowel sounds, I should like to emphasise the great service that the writer has done in pointing out that the ordinary methods of harmonic analysis are not necessarily adequate for the determination of the composition of a given tone, and may, indeed, give quite a false representation of the facts, because the sound may have inharmonic components. At the same time, it is doubtful whether his note in NATURE of March 3 (p. 12) in reply to another correspondent, interpreting some of Prof. Miller's results in this field, are justifiable. Prof. Miller's curves are evidently harmonic, from the fact that they repeat themselves very faithfully at regular intervals and establish without much doubt that vowel sounds (and some others) at least can be so produced that they are susceptible of harmonic analysis, whether they are always of such nature or not. The fact that Prof. Scripture finds the quality of the voice constantly changing in speech is not a matter of surprise, any more than that the human face and form rarely remain exactly the same for two seconds at a time in waking hours; it need not preclude us, however, from seeking to maintain a given quality for a time for purposes of analysis and record, any more than the latter fact prevents us from sitting for portraits.