SPEECH roni.R .ixd nxr.h-cy 



(M 



shows composite curves and Fig. 21) gives a series of cadi kiiul nf iur\es 

 for four speakers. Note the changing mode of stress wliirh is shown 

 in the se(iuence from top to bottom; in tlie first case tiie sylhiblcs of 

 weaker stress greatl\- predominate while in the last ca.se there is a 

 more nearly uniform distribution of the syllal)les with respect to Ui\ 

 degree of stress. It is e\ident from a comparison of the two series 



5—^^= 1 



2 a ~S o 



Powee ^reesi ^Ntea-y Stress 



Fig. 2b — Types of stress curves 



that the speaker's type is much the same whether judged by the power 

 or energy standard. An exceptional case might arise, however, if one 

 should put emphasis on a s>llable by prolonging the time of utterance, 

 for here the increased energy of the syllable would not necessarily 

 mean a greater stress. But from the point of view of phonetics, the 

 energy method should be useful in calibrating emphasis, which can be 

 taken as a function of time of duration as well as of mean power. 



