Postscript on the Realism-Instrumentalism Debate

The body of the monograph has throughout skirted around the philosophic literature comprising the “realist-instrumentalist” debate. This Postscript does not take sides in this debate, but offers suggestions intended to make the debate more tractable. One suggestion concerns two largely ignored distinctions: the first between theoretical claims that enter into the design of an experiment constitutively versus only heuristically; the second between intermediate standings a hypothesis can have between its being a mere conjecture and its becoming deeply entrenched through the success of research predicated on it. The second half of the Postscript explains why, of all elements of science, the equations in theory-mediated measurement that authorize values for target quantities to be obtained from values of more accessible quantities and the values so obtained can, under identifiable conditions, have the strongest claim to permanence in the face of both new data and theory change.

Our Initial Issues, Revisited

In addition to giving summary answers to the two issues singled out in Chapter 1—concerning (1) what Perrin established about Brownian motion itself and (2) how the standing of molecular theory changed between 1905 and 1913—the chapter answers three further questions raised by the intervening chapters: (3) What did Perrin and Brownian motion contribute to the new standing? (4) Given that Ostwald had already come to accept molecules before he encountered Perrin’s work, what led him to do so and what did the new standing amount to in his eyes? (5) How did the success in determining values for the interlinked principal molecular constants during the period of 1905 to 1913 contribute to the reliance on evidence from complementary theory-mediated measurements of interlinked “fundamental constants” that has been at the center of research in microphysics ever since?

The Historical Background: Molecular Theory as of 1900

The task of this chapter is to explain the sense in which molecular theory—both molecular-kinetic theory and chemical-molecular theory—were still viewed as hypotheses as of 1900 and why the evidence bearing on them during the second half of the nineteenth century was insufficient for them to have achieved standing beyond this. The chapter reviews the strengths and limitations of the evidence in question, taking advantage of two widely read textbooks in physical chemistry published in the 1890s by Wilhelm Ostwald and Walther Nernst and a uniquely comprehensive review of the evidence pertaining to the kinetic theory of gases, by O. E. Meyer, published in 1899. This background defines the historical context within physics and chemistry for the developments covered in the remainder of the monograph.

Perrin’s Brownian Motion Experiments

Between 1908 and 1911 Perrin published values for Avogadro’s number—the number of molecules per mole of any substance—on the basis of theory-mediated measurements of the mean kinetic energies of granules in Brownian motion. The umbilical cord connecting these energies to Avogadro’s number was the assumption that they are the same as the mean kinetic energies of the molecules in the surrounding liquid. This, as van Fraassen has argued, seems to presuppose that molecules exist, thereby undercutting Perrin’s claim to be proving their existence. This chapter reviews Perrin’s four theory-mediated measurements, showing, on the one hand, that none of them in fact depended on molecular theory yet, on the other, that, by virtue of being exemplars of theory-mediated measurement at its best, they managed to establish several extraordinary landmark conclusions about Brownian motion in its own right.

Introduction

Lore has it that research on Brownian motion, spearheaded on the theoretical side by Albert Einstein, but then strongly supported by Jean Perrin’s experimental efforts, finally ended the controversy over whether molecules exist. That view has nevertheless been challenged on more than one occasion, most recently by Bas van Fraassen. A discussion of the history of the standard view and challenges to it leads to two issues that the remainder of the monograph addresses: one concerning just what Perrin established about Brownian motion itself, and the other concerning how the standing of molecular theory had changed from 1900, first to Einstein’s initial paper of 1905 and then between that year and Perrin’s Les atomes of 1913. At the center of both of these issues is evidence resulting from theory-mediated measurements of aspects of Brownian motion—hence the subtitle of the monograph.

Implications for Molecular-Kinetic Theory

The question in this chapter is what evidential bearing Perrin’s extraordinary findings on Brownian motion have on molecular-kinetic theory without having to assume that his measured kinetic energies match those of molecules in the surrounding liquid. Three possibilities are assessed: (1) they provide a successful test of the claim that Brownian motion is caused by molecular motion; (2) they “ground” molecular-kinetic theory by establishing values for its fundamental parameters; and (3) they add legitimacy to certain hypothetical molecular phenomena through establishing that Brownian motion exhibits counterparts to them. The conclusion in each case is that Perrin’s results provide some evidence, but of less moment than has generally been claimed. The chapter concludes that his results nevertheless do provide conclusive evidence not that molecules exist, but that sustained, highly localized random fluctuations are occurring in the liquid, fluctuations that are incompatible with its acting as a continuum.

The Historical Background: Brownian Motion as of 1905

The mystery of Brownian motion had been announced with its discovery by Robert Brown in 1828: the persistence of the motion of solid particles in liquids for indefinite periods of time instead of sinking as sediment to the bottom. Once molecular-kinetic theory emerged more fully a few years later, it was the obvious candidate for explaining the phenomenon. Nevertheless, those developing kinetic theory in the second half of the century, Maxwell and Boltzmann, appear to have ignored it. The chapter summarizes research on Brownian motion during the nineteenth century, indicating why leading physicists ignored it, and what developments in the first five years of the twentieth century led to its suddenly becoming so important to kinetic theory. This background supplements that of Chapter 2, completing the historical context for the developments covered in subsequent chapters.

Converging Values for Avogadro’s Number

Perrin’s values for Avogadro’s numbers presuppose that the mean kinetic energies in Brownian motion match those of molecules in the surrounding liquid. In support of these values and the presupposition, Perrin turns to values of Avogadro’s number and mean kinetic energies of molecules obtained by means of theory-mediated measurements by Planck (from blackbody radiation), Rutherford and his colleagues (from α‎-particle radiation), and Millikan (from ionization). While Perrin’s values differed from these others, they all collectively yielded values for the molecular constants within a theretofore unachievable window of ±5 percent. This chapter assesses first the evidence for Perrin’s values from his appeals to each of these “agreeing” complementary determinations, and then the evidence that molecules exist from all of them taken together. The conclusion is that the latter warranted taking molecules to exist in ongoing research into microphysics even though the referent of the word “molecule” remained seriously underspecified.