44 TUNES

between the closest variations of the best-preserved melodies one always
finds differences, however slight, of inflexion or rhythm. There are
some also whose relationship can only be divined through their general
form, with similar rhythms and melodic cadences falling always on the
same places, though the outline of the song differs completely.

The paragraph expresses our difficulties concisely; it also serves
to illustrate the rather generalized language which musical critics
of this subject too often employ. Critical language should be
specific; it should be possible to state exactly what are the resem-
blances and differences between tunes, and to do so by means of
symbols which are intelligible to all. For this purpose it is doubtful
whether staff notation is convenient. Staff notation provides the
best standard of reference, but on points of detail it is costly to
insert fragments of staff notation in a printed page, and to do so is
still to remain unintelligible to the unmusical reader. The use of
musical notation involves printing volumes of large size, which are
heavy to handle; and it is not at all convenient to enter on a card-
index which can be shuffled and compared. To compare such
tunes it is necessary to transpose them to a uniform key, since only
then will the likeness of the contours become apparent to the eye.
But it is hard to avoid a suspicion that transposition may injure in
some way a musical text already fixed with some hesitation. There
is, as we have seen, some doubt whether staff notation suffices to
represent the subtler nuances of certain kinds of folk-song. It also
implies—through its connexion with mechanical instruments like
the piano—a greater precision of measurement than is found in the
free narrative style of the human voice. It would be better, then,
having fixed the standard transcription as accurately as possible,
to leave it alone, and seek for some auxiliary notation, less costly
and cumbrous, and more intelligible to the musically uneducated,
more capable, also, of being entered on cards which can be shuffled
and compared.

Such a system has been devised by S. B. Hustvedt, of the Uni-
versity of California at Los Angeles,r to serve to classify the English
and Scottish ballad tunes. There are two things to be done: we
must record the tune and we must characterize it. For purposes
of record it is sufficient to name the notes, give their length, and
add the key and time. A reference to the source of information
will help in later verifications. By using ordinary letters it is
1 See Note A, at the end of the book.