FOR half a century after its first appearance in print Hermann Paul's Prinzipien der Sprachgeschichte ( 1880, 5th ed., 1920)1 continued to be the vade-mecum of the philologist. Paul harvested the 'methodological gains' of the previous fifty years, and formulated for investigators a 'Prinzipienlehre' or body of dogmas which gave a particular shape to their collectanea. He opposed the History of Language to 'mere' descriptive grammar, and formally excluded from the category of a science any method that was not historical. His basis of experience was Indo-European, and within that field it was predominantly Germanic. Noting that there is no direct communion of minds Paul insisted upon the physical side of speech, and within this area he concentrated upon speech-sounds. Speech being a continuum of infinitely many sounds, the speaker has a sense of movement but no consciousness of individual elements of sound, and a sound-shift is uniform for all similarly conditioned instances within one and the same dialect. The sound-law, however, though working without exceptions other than those caused by analogy, was considered by him to be an historical summary, not a forward-looking law of cause and effect. Analogical pictures of the speech-process must be rejected. Our knowledge is the more scientific in proportion as it treats the activity of single factors in isolation. The true object of the linguist's study is the general manifestation of the activity of speaking rather than the effect of individuals upon each other; and the primary cause of change is the speech-activity itself.
In the same year ( 1880) there appeared the second edition of Wilhelm von Humboldt über die Verschiedenheit des menschlichen Sprachbaues. For von Humboldt the linguist's task was an openminded inquiry into the nature of Language for which the evidence of the most diverse types was solicited; for Paul it was the application of a technique to a specialized field. Technique is a____________________