st
materially satisfactory hypothesis. Mr. Joseph handles it by formal
logic purely, as if he had no acquaintance with the logic of hypothesis
at all.

Mr. Joseph again is much bewildered as to what a humanist can mean when
he uses the word knowledge. He tries to convict me[137] of vaguely
identifying it with any kind of good. Knowledge is a difficult thing to
define briefly, and Mr. Joseph shows his own constructive hand here even
less than in the rest of his article. I have myself put forth on
several occasions a radically pragmatist account of knowledge,[138] the
existence of which account my critic probably does not know of--so
perhaps I had better not say anything about knowledge until he reads and
attacks that. I will say, however, that whatever the relation called
knowing may itself prove to consist in, I can think of no conceivable
kind of _object_ which may not become an object of knowledge on
humanistic principles as well as on the principles of any other
philosophy.[139]

I confess that I am pretty steadily hampered by the habit, on the part
of humanism's critics, of assuming that they have truer ideas than mine
of truth and knowledge, the nature of which I must know of and can not
need to have re-defined. I have consequently to reconstruct these ideas
in order to carry on the discussion (I have e.g. had to do so in some
parts of this article) and I thereby expose myself to charges of
caricature. In one part of Mr. Joseph's attack, however, I rejoice that
we are free from this embarrassment. It is an important point and covers
probably a genuine difficulty, so I take it up last.

When, following Schiller and Dewey, I define the true as that which
gives the maximal combination of satisfactions, and say that
satisfaction is a many-dimensional term that can be realized in various
ways, Mr. Joseph replies, rightly enough, that the chief satisfaction of
a rational creature must always be his thought that what he believes is
_true_, whether the truth brings him the satisfaction of collateral
profits or not. This would seem, however, to make of truth the prior
concept, and to relegate satisfaction to a secondary place.

Again, if to be satisfactory is what is meant by being true, _whose_
satisfactions, and _which_ of his satisfactions, are to count?
Discriminations notoriously have to be made; and the upshot is that only
rational candidates and intellectual satisfactions stand the test. We
are then driven to a