Charles Lyell, Historian of Time's Cycle - Time's Arrow, Time's Cycle: Myth and Metaphor in the Discovery of Geological Time - page 119

entists, and embraced warmly by all geologists; the substantive claims were
controversial and, in some cases, accepted by few other geologists.
Lyell then pulled a fast one—perhaps the neatest trick of rhetoric, measured
by subsequent success, in the entire history of science. He labeled all these
different meanings as "uniformity," and argued that since all working
scientists must embrace the methodological principles, the substantive claims
must be true as well. Like wily Odysseus clinging to the sheep's underside,
the dubious substantive meanings of uniformity sneaked into geological
orthodoxy—past an undiscerning Cyclops, blinded with Lyell's rhetoric—by
holding fast to the methodological principles that all scientists accepted.
We shall probably never know whether Lyell perpetrated this ruse
consciously—I suspect that he did not, since his strong commitment may
have engendered a personal conviction that all meanings must be true a
priori. In any case, Lyell's rhetorical success must rank among the most
important events in nineteenth-century geology—for it established an
"official" history that enshrined, as the earth's own way, a restrictive view
about the nature of change. If any scientist ever tries to convince you that
history is irrelevant, only a repository for past errors, tell him the story of
Lyell's rhetorical triumph and its role in directing more than a century of
research in geology.
Using my two categories of methodological and substantive claims, Rudwick
(1972) has discerned four distinct meanings of uniformity in Lyell's
Principles.
1. The uniformity of law. Natural laws are constant in space and time.
Philosophers have long recognized (see, in particular, J. S. Mill, 1881) that
assumptions about the invariance of natural law serve as a necessary warrant
for extending inductive inference into an unobservable past. (Induction, as C.
S. Peirce noted, can be regarded as self-corrective in an observable present,
but we can never see past processes, and no amount of current repetition can
prove that present causes acted in the same way long ago—hence our need
for a postulate about the invariance of nature's laws.) Or,