of the _Divine Comedy_ than the untrained person. Or compare Pater's
interpretation of the "Mona Lisa" with Muther's. Can we say that certain
ideas and images belong properly to the work of art, while others do
not? With regard to this, we can, I think, set up two criteria. First,
the intention of the artist--whatever the artist meant his work to
express: that it expresses. Yet, since this can never be certainly and
completely discovered, there must always remain a large region of
undetermined interpretation. Now for judging the relevancy of this
penumbra of meaning and association the following test applies--does it
bring us back to the sensuous medium of the work of art or lead us away?
Anything is legitimate which we actually put into the form of the work
of art and keep there, while whatever merely hangs loose around it is
illegitimate. For example, if while listening to music we give ourselves
up to personal memories and fancies, we are almost sure to neglect the
sounds and their structure; we cannot objectify the former in the
latter; with the result that the composition is largely lost to us.
Naturally, no hard and fast lines can be drawn, especially in the case
of works of vague import like music; yet we can use this criterion as a
principle for regulating and inhibiting our associations. It demands of
us a wide-awake and receptive appreciation. The genuine meanings and
associations of a work of art are those which are the irresistible and
necessary results of the sense stimuli working upon an attentive
percipient; the rest are not only arbitrary, but injurious.

To this, some people would doubtless object on the ground that art was
made for man and not man for art. The work of art, they would claim,
should interpret the personal experience of the spectator; hence
whatever he puts into it belongs there of right. There are, however,
two considerations limiting the validity of this assertion. First, the
work of art is primarily an expression of the artist's personality