56 ECONOMIC ESSAYS IN HONOR OF JOHN BATES CLARK 
EACH 
a Hi 
time or energy for earning a good living or enjoying the fruits 
of one’s labors. It is rational not to look after one’s interests 
perfectly in every respect and every relation of life; and this 
fact has real significance in judging the effects of an economic 
system which is built on the supposition that every individual 
does look out for his own interests in all his relations with his 
tellowman. Rational decision can attain perfection only in 
dealing with things familiar and customary, but it is only 
needed in dealing with things new and not yet reduced to cus- 
tom or routine. And those strategic decisions called “marginal” 
include many and significant departures from the static norm 
of rationality. 
The so-called “instinct of workmanship” is another paradoxical 
trait, for it is essentially one whereby any means may become 
an end in itself: a worker gains interest in the technique of any 
process which the attaining of his ends make necessary, and 
having done so, he may lavish his efforts, rather than economize 
them, or even sacrifice the end to the technique. Yet this waste 
and possible perversion is the price of that direct interest in the 
work as such, without which the most effective work is not 
possible. Here again, perfect efficiency, conceived after rational 
models, is an ideal which is not in accord with human nature 
as it is actually constituted. Waste of some sort is inevitable. 
Since intelligent choosing is so largely a matter of “trial and 
error,” it is important to ask how the errors operate, how they 
correct themselves (if they do so) and what happens if they 
do not. For our purposes “error” is probably an unfortunate 
term, suggesting as it does a mathematical calculation or the 
determination of an objective fact, in which there is one accurate 
result and departures from it can be definitely determined. This 
is true in many cases, especially in the field of business deci- 
sions, where it is a question of cheapening production or increas- 
ing profits. This also applies to consumption, so far as it is a 
matter of economical use of particular means to attain a definitely 
given end. But where it is a case of choosing between different 
ends, the case is altered. Here there are two great classes of 
choices: those in which it is possible to sample alternatives and 
then follow for the future the one which experience leads one 
to prefer, and those in which such sampling is difficult or 
impossible and the individual may be disappointed in his choice 
without knowing that another course would have produced greater