234

INTERNATIONAL TRADE

shall be in operation, all others excluded. Therein our situation
is unlike that of the chemists and physicists, or that which the
biologists have succeeded in attaining within the last half-century.
It is essentially like that of the geologists and geographers. These
also are limited to the method of observation; they also cannot
experiment. The reason for their limitations is indeed different
from that which the economist must face. The geologists cannot
experiment because the length of time over which their forces
operate stretches far beyond the possibilities of man’s operations;
the same is true in the main of the geographers. The economist
cannot experiment because he deals with human beings, and
moreover with human beings in the mass — with large numbers
and with mass phenomena. He cannot subject thousands of
people to a given set of conditions, rigidly prevent the intervention
of any other than the chosen factors, and unconcernedly let the
consequences emerge. He can do no more than observe what takes
place in the unregulated and changing world of affairs. He must
watch with attention the confused march of events, discern the
significant phenomena among the negligible, and patch together
fragments of evidence which, tho they may tend to confirm a
hypothesis, will but rarely constitute a continuous chain of evidence.
 The same is true of the geologist. He finds strata that
appear and disappear, with gaps as they vanish across valleys or
are buried beneath mountains and plateaus. It is thru long and
patient observation, skilful hypothesis, visualizing of an imagined
situation, excision of the accidental and the non-pertinent, that
he establishes his generalizations or at least gives them plausibility.
Such is the position of the economist, and such are his limitations.
With all the refinements of observation which the development of
statistical technique may bring, with all the accumulation of data
thru economic history and description, the economist’s generalizations
 can never have the exactness or the certainty which come
from experimental confirmation.
Occasionally it happens, however, that there is a simple situation:
 a train of economic phenomena in which one cause alone
is in operation, or is so predominantly in operation that others can