According to Sampson, "[from mechanistic] physics, philologists took the notion of describing the history of sound-changes occurring in a language in terms of 'laws' which apply uniformly to a whole range of examples, rather than discussing individual words in the anecdotal, case-by-case way in which a historian...treats individual persons or events" (pp. 16-17). The classic example of such a "law" is what is known as "Grimm's Law," which describes how certain consonants in Proto-Indo-European (the hypothesized ancestor of most modern European languages) changed in Germanic languages. Over the course of the 19th century, this notion of "sound-laws" modeled on the laws of mechanistic physics became widely accepted. The problem, though, was that linguistic "laws" did not really resemble the laws of physics. The laws of physics are truly universal (except at the quantum level)--they apply equally to the motion of a rolling ball and to the motions of the planets--and have been in force since the universe began. On the other hand, linguistic "laws" only apply to certain languages, not all languages at all times, and, as Sampson points out, "[one] would scarcely be impressed by a physicist who invented one law of gravity for seventeenth century Italy, another for modern England, and so on" (p. 28-29). Thus, it is not surprising that eventually the idea of "sound-laws" ultimately fell by the wayside.
Interestingly, Sampson's discussion of how mechanistic physics affected thinking in the field of linguistics strongly reminded me of something I remembered reading in another book--The Soul of Science: Christian Faith and Natural Philosophy, written by Nancy R. Pearcey and Charles B. Thaxton. In the fourth chapter of their book, Pearcey and Thaxton talk about how, during the 18th century, the "Newtonian" worldview (so called because it was based on the ideas of Isaac Newton about physical laws), "eventually became the paradigm for all human knowledge. Practitioners of political theory, ethics, psychology, and theology sought to restructure their disciplines in accord with mathematical physics in order to render them truly 'scientific'" (pp. 93-94). In other words, just as linguists in the 19th century had tried to apply the notion, borrowed from physics, of universal laws to language, so had political scientists, theologians, and other scholars sought to apply the notion of universal laws to their respective fields of inquiry. Moreover, in the end, just as in the case of linguistics, this effort proved less than successful. Still, that they should have even tried demonstrates the power of scientific paradigms outside the realm of science, and this is a phenomenon that has occurred more than once in the history of ideas--witness, for example, Karl Marx's efforts to discover economically-determined "laws" of history.
Image of Thomas Kuhn from Wikimedia Commons via tendencias21.net