diminish what traces they bear of those marks of hard labour
which tend to remove the meticulous operations of much modern
research from the interests of men of taste. I have attempted to
keep a critical eye on such processes as have been appropriated
here. It is often suggested that questions of source and influence
have occupied the attentions of literary scholars too persistently
during the past forty or fifty years. I agree that such studies have
been disproportionately abundant in comparison with those that
show a perception of values in the minds and works at each end
of the line of influence. It is true, I fear, that discriminations of
quality and judgment have been held too long in suspense out of
excessive regard for scrupulous digressions into the 'backward
and abysm' of literary history. But what occurs to one more
frequently in ruminating over these matters is another sort of
doubt: not that investigations of this kind go too far, but
whether they go far enough? The role of influence as a deter-
minant in literature and the arts cannot be denied. It is discussed
or assumed universally nowadays, even by those critics who
deplore the emphasis. Yet what is meant by 'influence' in these
spheres? The word, in its most frequent use, is no master-key but
a counter standing for a vague category into which is thrust with
equal lack of discrimination samples of a complex phenomenon
which has never been examined in its variations or in its essence.
A large part of the problem presented by an influence in literature
lies in the nature of literary influence itself. Similar doubts assail
one about the methods habitually employed to detect or examine
a specimen. The quest tends to operate in one direction only--
away from the masterpiece. Yet the effort to retrogress toward
the source remains largely gratuitous if the procedure is not
inverted and pressed forward to appreciate the point at which
source and influence are subsumed in the final synthesis. Con-
sidered in its full implication the problem of literary influence is
an integral part of the problem of literary creation and this can

-viii-