pros and cons of the measure before him. We see,
in fact, that not only is his thinking determined by a complex
of whose action he is unconscious, but that he believes his
thoughts to be the result of other causes which are in reality
insufficient and illusory. This latter process of self-deception,
in which the individual conceals the real foundation of his
thought by a series of adventitious props, is termed
'rationalization.'

"The two mechanisms which manifest themselves in our example of
the politician, the unconscious origin of beliefs and actions,
and the subsequent process of rationalization to which they are
subjected, are of fundamental importance in psychology."
(Bernard Hart: _The Psychology of Insanity_, pp. 64-66.)]

Again, even where common-sense judgments are not particularly
qualified by such conditions, they are frequently
based upon the observation of purely accidental conjunctions
of circumstances. A sequence once or twice observed is taken
as the basis of a causal relation. This gives rise to what is
known in technical logic as the _post hoc ergo propter hoc_
fallacy; that is, the assumption that because one thing happens
after another, therefore it happens _because_ of it. Many
superstitions probably had their origin in such chance observations,
and belief in them is strengthened by some accidental
confirmation. Thus if a man walks under a ladder one
day and dies the next, the believer in the superstition that
walking under a ladder brings fatal results will find in this
instance a clear ratification of his belief. There seems to be
an inveterate human tendency to seek for causes, and by
those who are not scientific inquirers causes are lightly
assigned. It is easiest and most plausible to assign as a cause
an immediately preceding circumstance. Exceptional or contradictory
circumstances are then either unnoticed or pared
down to fit the belief.

Scientific method does not depend on such chance conjunctions
of circumstance, but controls its observations or
experimentally arranges conditions so as to discover what are
the conditions necessary to produce given effects, or what
effects invariably follow from given causes. It does not
accept a chance conjunction as evidence of an invariable
relation, but seeks, under regulated conditions, to discover what
the genuinely invariable relations are. This method of
controlling our generalizations about the facts of experience, we
shall presently exa