50 THE STUDY OF SPEECH CURVES. 



extension of the principle. The next line contains a portion from the 

 middle of [o] in "do," comprising the top of the ampUtude-convexity. The 

 steady change in form is apparent. The eye tells at once whether the 

 curves in the other lines are from beginning, middle, or end of a vowel 

 simply by noting how the amplitude is changing. 



A comparison of the curves for the different vowels is interesting. The 

 vowel of "Come" is the one described by phoneticians as the indefinite 

 vowel and indicated by [a]. The waves show that — like all short vowels — 

 it changes its character greatly within a short time; from such a changing 

 sound the ear obtains only a general impression with no indication of the 

 variations. If we select the waves at the right of the line as typical of 

 this vowel, we find that somewhat similar waves occur for the vowel of 

 "do"; as far as the ear can judge the two vowels sound ahke. The ear is 

 absolutely incapable of deciding what the vowel is in "what" in this record; 

 I have indicated it by [a], but its curve bears no resemblance to those of [a] 

 in the " Come" and "do." To the ear the vowel of "you" in the fifth line 

 is also the indefinite one, but its wave shows that it is quite different 

 from the vowels in " Come" and " do. " The ear hears that these "indefi- 

 nite vowels" differ from one another, but we can not state the distinctions 

 because phonetics has not provided the means of specification. As the 

 curves in plate x show, these vowels have sharply defined, characteristic 

 curves; the indefiniteness does not lie in the vowels, but in our knowledge 

 concerning them. 



The vowel of " to" is distinctly heard to be a ver}^ short [u] and not 

 [d]; its waves are similar to those of "your" and "good." This "your" 

 is spoken with [u] and not like the "you" in the fifth line. The writer 

 is utterly unable to decide by the ear what vowel is used in "your" of 

 the nineteenth line; it sometimes seems to sound rather like [o]; its curve 

 resembles no other in the figure. The indefinite article " a" is pronounced 

 with the vowel [a]; the curv^e is somewhat Hke that of [a] in "you." The 

 waves for [si] in "glass," "that," "schnapps," "and," and "family's" show 

 Uttle differences from one another, except in their amplitudes. The vowel 

 in "pros-" shows waves that are not similar to those for the [a] in the last 

 line; the vowel in this case is certainly not a "short a," although that would 

 probably be the English pronunciation. I have indicated it by [o], which 

 seems to be the most common American pronunciation. Its latter part 

 shows a shght similarity to the first portion of the vowel in " Come." The 

 line for "per" shows only a few irregular vibrations; Jefferson was accus- 

 tomed to speak this syllable practically as a surd. 



The cui-ve of laughter on plate vii presents some interesting problems. 

 The vowel in these laughs is not the vowel [a] or [o], although it seems 



