A PORTRAIT OF STEPHEN CRANE
359
sight and his writer's sense of craftsmanship and intention, Berryman
is generally excellent. He is at his best in the chapter on Crane's art;
here he winnows the major work from the trifling, distinguishes sug–
gestively the peculiar quality of Crane's poetry in an intriguing analogy
with the primitive spell of the medicine man, defines freshly the im–
pressionism of his prose style, not to mention three sub-varieties of it,
and manages to suggest at the same time the mordantly independent
character of the mind that lies behind the whole work. He does what is
the business of a descriptive practical critic: he suggests the uniqueness of
Crane without labeling, and he achieves discrimination without loss of
clarity.
As to the distinctively biographical parts, a distinction must be
made. One can only admire the intelligent industry which has gone
into the establishment of fact; but concerning the scheme of person–
ality it is possible to have grave doubts. Berryman seems to have been
scrupulous in running down every source of information, both published
and unpublished (including three important ones which, he tells us, he is
not permitted to cite) . That is not to say that the record is altogether
a satisfactory one. Crane was often laconic about himself to the point
of mystification; he was hopelessly careless as to facts, and spirited (as
Berryman says) as to dates in what he did say; and legend multiplied
mystery and contradiction about him wherever he went. Some of the
puzzles a:bout the mere events of his life will doubtless never be un–
raveled; insofar as an outsider can tell, Berryman seems
to
have gone
about as far as diligence in investigation and imagination in collation
can take one.
These are the materials of the portrait; the portrait that Berry–
man shapes from them is, I should say, much less likely to be definitive.
His treatment proposes the wider question of the uses of psychoanaly–
sis to the biographer of the dead writer, or perhaps more exactly, of
the nature of the evidence which will justify such uses. I suppose it
is beside the point, where truth is concerned, to confess to a certain
sense of anticlimax in arriving at promised revelation in the form of an
Oedipus complex', even a special and ingenious pattern of an Oedipus
complex. What is more serious is the question of the criteria for proof.
Much depends, in Berryman's account, on a single incident whose source
is Thomas Beer (not elsewhere regarded as particularly reliable) about
Crane's fright at seeing at the age of twelve a white girl stabbed
by her Negro lover. But the conscientious analyst can do little with an
isolated episode. Indeed even with the full co-operation of a living pa–
tient it is not always easy
to
reach a confident conclusion; when the
