reconcilable assailants. Nor
do I think that Mr Arnold would have much relished the apology made, I
think, by Mr Leslie Stephen since his death, that its critics "mistake
an epigram for a philosophical definition." In the first place, the
epigrammatic quality is not clearly apparent; and in the second place,
an epigram would in the particular place have been anything but
appropriate, while a philosophical definition is exactly what was
wanted.

Mr Arnold himself never attempted any such defence. He pleaded, with
literal justice, that the phrase "a criticism of life" was only part
of his formula, which adds, "under the conditions fixed for such a
criticism by the laws of poetic truth and poetic beauty." But this
does not make the matter much better, while it shows beyond
controversy that it _was_ a philosophical definition that he was
attempting. It merely takes us round in a circle, telling us that
poetry is poetical, that the archdeacon performs archidiaconal
functions. And while it is not more illuminative than that famous and
useful jest, it has the drawback of being positively delusive, which
the jest is not. Unless we are to assign some quite new meaning to
"criticism"--and the assignment of new meanings to the terms of an
explanation is the worst of all explanatory improprieties--poetry is
_not_ a criticism of life. It may be a passionate interpretation of
life--that has seemed to some not a bad attempt at the
unachievable,--a criticism it cannot be. Prose fiction may be and
should be such; drama may be and should be such; but not poetry. And
it is especially unfortunate that such poetry as answers best to the
term is exactly that poetry which Mr Arnold liked least. Dryden and
Pope have much good and true criticism of life: _The Vanity of Human
Wishes_ is magnificent criticism of life; but Mr Arnold has told us
that Dryden and Pope and Johnson are but "classics of our prose." That
there is criticism of life _in_ poetry is true; but then in poetry
there is everything.

It would also, no doubt, be possible to pick other holes in the paper.
The depreciation of the "historic estimate," instead of a simple hint
to correct it by the intrinsic, is certainly one. Another is a
distinct arbitrariness in the commendation or discommendation of the
examples selected. No one in his senses would put the _Chanson de
Roland_ on a level with the _Iliad_ as a whole; but some among those
people who happen to possess an equal acqua