8

πλ
8

mentally associated with the object. In this statement, however,
perfect accuracy is sacrificed to perspicuity. To these three kinds of signs
I give the names, icon, index, and symbol, respectively. (The division
was published in 1867.) I next show that no study, experiment, or
other operation upon a symbol of an object can lead to anything
more than a different form of thought about that object; and
that no operation upon an index of an object, though it may yield
new truth about the object, can ever furnish apodictic
proof of that truth. The study of an icon, on the contrary,
though its results can no more be infallible than is the summation
of a column of figures, may disclose necessary, i.e. essential,
truths regarding the object. I never consider what can
be effected by the connected study of different signs, and thus find that while