n other words,
that a science of society is possible. In England this idea was still a
novelty when Mill's System of Logic appeared in 1843.

The publication of this work, which attempted to define the rules for
the investigation of truth in all fields of inquiry and to provide tests
for the hypotheses of science, was a considerable event, whether we
regard its value and range or its prolonged influence on education.
Mill, who had followed recent French thought attentively and was
particularly impressed by the system of Comte, recognised that a new
method of investigating social phenomena had been inaugurated by the
thinkers who set out to discover the "law" of human progression. He
proclaimed and welcomed it as superior to previous methods, and at the
same time pointed out its limitations.

Till about fifty years ago, he said, generalisations on man and society
have erred by implicitly assuming that human nature and society will for
ever revolve in the same orbit and exhibit virtually the same phenomena.
This is still the view of the ostentatiously practical votaries of
common sense in Great Britain; whereas the more reflective minds of the
present age, analysing historical records more minutely, have adopted
the opinion that the human race is in a state of necessary progression.
The reciprocal action between circumstances and human nature, from which
social phenomena result, must produce either a cycle or a trajectory.
While Vico maintained the conception of periodic cycles, his successors
have universally adopted the idea of a trajectory or progress, and are
endeavouring to discover its law. [Footnote: Philosophical writers in
England in the middle of the century paid more attention to Cousin than
to Comte or Saint-Simon. J. D. Morell, in his forgotten History and
Critical View of Speculative Philosophy (1846), says that eclecticism is
the philosophy of human progress (vol. ii. 635, 2nd ed.). He conceived
the movement of humanity as that of a spiral, ever tending to a higher
perfection (638).]

But they have fallen into a misconception in imagining that if they can
find a law of uniformity in the succession of events they can infer the
future from the past terms of the series. For such a law would only be
an "empirical law"; it would not be a causal law or an ultimate law.
However rigidly uniform, there is no guarantee that it would apply to
phenomena outside those from which it was derived. It must itself depend