MODERN HISTORY OF LITERATURE 505 



undoubtedly is an anaphora. It is a difference, to be sure, not not- 

 iceable to the eye, but only to the ear. The style of a sentence 

 containing a few paltry, absolutely unaccented interjections is still 

 regarded as "excited" because of statistical summarizing, whereas 

 a single strongly accented interjection throws a dozen weaker ones 

 in the shade. Prosody has been far more intensively cultivated, to be 

 sure, but only by a small group of men whose results are not accepted 

 by the large majority of literary historians. In the long run, however, 

 it will not be possible to construct statistical tables about accents 

 before having a clear understanding of what accents mean; to explain 

 questions of quantity from the nature of sounds and syllables, 

 whereas the same word of course is employed by one and the same 

 poet, and in the very same kind of verse, now this way, now that; 

 or to infer directly, because of the various kinds of unstressed syllables, 

 that it must be another poet, or at least another period of the poet, 

 even though it can be seen from the dated manuscript of the Wal- 

 purgisnacht that Goethe used on one day theses of one syllable, but 

 on the next theses of several syllables, simply because the thought 

 required a more lively movement of the verse ; or finally, to shake one's 

 head at the placing of a stressed syllable in an unstressed position, 

 because we not only sing but also read " Heil dir im $iegeskranz." 

 In addition to force we have also learned now to take into account 

 pitch in a verse, and here in America investigations have been 

 begun to determine the melody of language by means of physical 

 instruments. Careful observations of the rhythm and melody of the 

 spoken not the written language of the sentence, and not the separate 

 word, will enable the study of meter in the future to base its ob- 

 servations less upon a lifeless theory of statistics than upon sound, 

 just as the poet composed his verses according to hearing. In this 

 way, the theory and practice of prosody, so long at variance with 

 each other, will become reconciled. Even in the study of poetry 

 attempts have been repeatedly made in the last decades to make 

 use of the inductive instead of the deductive method. Some have 

 wished to reconstruct it upon a scientific, others upon an ethno- 

 graphic-anthropological, and a third upon a psychological basis. 

 But here also the leap over intermediate terms of hypotheses will 

 hardly lead us back to the origin; only that path will that starts 

 out from the safe mean, by tracing step by step and ordering the 

 rich detailed observations, of which even here there is no lack. 



Latterly, the need has been set forth energetically and from differ- 

 ent quarters of advancing beyond the limits sot to purely philosoph- 

 ical treatment of the history of literature and of coming in contact 

 with other branches of knowledge. Accordingly, much has been said 

 of the comparative study of the history of literature, without, how- 

 ever, the same idea being everywhere associated with the expression, 



