The Form and Function of Scientific Discoveries
Kenneth L. Caneva
Dibner Library Lecture, Smithsonian Institution Libraries, November 16, 2000
1 Caneva 1998a, 82-90. I intend on giving a much more detailed and extensive treatment of the topic of this lecture in a later paper.
2 Examples of early and modern canonical characterizations include Breger 1999, 30; Hermann 1971-1972, 2, 346 (by Lothar Suhling); Jungnickel and McCormmach 1986, 44; Ostwald 1896, 379; Siegel 1983,413; Whewell 1837, 3, 889-90; Whittaker 1951, 88-89. Even Nielsen's excellent account of Seebeck's work was headed "The discovery of thermoelectricity" (Nielsen 1991, 363-395; cf. 360). Tyndall (1873, 141)reported that "Thomas Seebeck, of Berlin, discovered that electric currents might be derived from heat," which gave rise to the construction of the "thermo-electric pile." John Herivel wrote that Seebeck showed "that temperature differences could produce electric currents" (in Williams 1982, 411).
3 Caneva 1974, 125-126; 1978, 92.
4 Caneva 1997, 44-47.
5 Examples of modern canonical characterizations include Muir 1994, 388; Porter 1994, 520; Root-Bernstein 1989, 136; Taylor 1941, 631; Whittaker 1951, 88; Williams 1982, 396 (by John W. Herivel). Cf. Breger 1999, 28: "Ørsted's discovery of the magnetic field of the electric current in 1820 was a convincing great success for the romantic program of research. . . . In the terminology of the time, which did not yet know the field concept, Ørsted had transformed electricity into magnetism." The anachronism of Breger's first sentence is obvious; the inappropriateness of his second claim is evidenced in Caneva 1998a, 54-55, 66,77, 79-80, 110, 113-114, where the distinction is noted between the transformation of forces and the calling forth of one force by another typically spoken of by Ørsted.
6 Ørsted 1820a, from the title.
7 Ørsted 1820a, 2, 4 = 1920, 2, 215, 218. He once spoke of the "flow [cursus] of the electrical forces in the connecting wire"(4 bzw. 218).
8 Ørsted 1820b, 365; 1820c, 78.
9 See the account in Caneva 1998a, 58-90, for details and references.
10 Ørsted 1805, 18-19 = 1920, 3, 103-104.
11 From among the many occurrences of Wirkungsform, see Ørsted 1812, 5, 236, 248, 252, 258 = 1920, 2, 38, 142, 147, 149, 151.
12 Ørsted 1821, 14 = 1920, 2, 448.
13 The extended process of redefining just what it was that Ørsted discovered has at length also credited him with the discovery of the magnetic field (Lindsay and Margenau 1936, 302).
14 On theories of transversal magnetism, see Rosenberger 1887-1890, 205-207; Caneva 1974, 61-65, 68-73, 75-80, 119-120; Caneva 1978, 78-83, 90-91.
15 Streit 1902, xv-xxii; Hermann 1971-1972, 2, 347; Caneva 1974,125; 1978, 92; Nielsen 1991, 382. Cf. Frankel's DSB article: "By far Seebeck's most significant discovery … was that of thermoelectricity-or thermomagnetism, as he called it-in 1822. … He did not, however, believe that an electric current was actually set up in the bimetallic rings and preferred to describe his effect as 'thermomagnetism'" (Frankel 1975, 281).Following Frankel's lead are Muir 1994, 464 (implicitly) and Breger 1999, 30 (explicitly). Breger added: "In the terminology of the time Seebeck had thus transformed heat into magnetism."As pointed
out with respect to Ørsted in note 5, above, such was not the way contemporaries typically spoke.
16 The date of Seebeck's discovery has been variously reported, reflecting the indirect and extended route by which different aspects of his work became known. I will not concern myself with this aspect of the discovery story.
17 Tilloch 1821, 462; Keferstein 1823, 4; Thomson 1822, 318.
18 Keld Nielsen's perceptive and detailed account of Seebeck's work includes a section called "Thermomagnetism becomes thermoelectricity" in which Ørsted's role is emphasized (Nielsen 1991,391-395).
19 Ørsted, letters of 2 December 1822 and 4 April 1823 to his wife, in Ørsted 1870, 2, 31-32 and 59.
20 Ørsted 1823a. Anonymous (1823, 315), reporting on the meeting of 3 March 1823, says Ørsted entertained the Academy "with the work that Seebeck has just done on electromagnetic phenomena. (See the preceding cahier.)"-i.e., Ørsted 1823a. The report was quickly translated into German (Ørsted1823c) and English (Ørsted 1823e).
21 Ørsted 1823a, 199 = 1920, 2, 263.
22 Ørsted 1823a, 199 = 1920, 2, 264. Ørsted was aware that Seebeck had another theory about these effects, but it is not clear that he had a very distinct idea of what that theory was (see his letter of 4 April 1823 to his wife, in Ørsted 1870, 2, 59-60).
23 Ørsted 1823a, 199-200 = 1920, 2, 264.
24 Ørsted 1823d, 9 = 1920, 2, 461.
25 Yelin 1823b, 4; 1823a, 419.
26 Yelin 1823c, 361.
27 Yelin 1823c, 363.
28 Yelin 1823b, 11. Yelin's reasoning was correctly captured in the editor's précis of his Der Thermomagnetismus published in the Bibliothèque Universelle, which may have introduced the term to the French-speaking world: "Yelin, having thus been led to consider the rupture of the equilibrium of temperature as the principal cause of the electromagnetic action of Seebeck's circuit, determined to try the effect of this rupture on a circuit, or on a piece of a single metal. Having even then obtained very pronounced effects, he thought it necessary to designate this class of phenomena by the name thermomagnetism" (Yelin, 1823d, 256 = Yelin 1824, 159).
29 As Seebeck explained, the published paper was an "extract" from four lectures delivered at the Academy of Sciences in Berlin on 16 August 1821, 18 and 25 October 1821, and 11 February 1822, plus later additions in the form of footnotes and an addendum (Seebeck 1825a, 265 = 1825b, 1). This clarification is omitted from Seebeck 1826.
30 Seebeck 1825a, 266-283 = 1825b, 2-19.
31 Seebeck 1825a, 283 = 1825b, 19.
32 Seebeck 1825a, 292-293 = 1825b, 28-29. Seebeck did not name Ampère explicitly. In a footnote to the German translation of A.-C. Becquerel 1826, Poggendorff noted with satisfaction Seebeck's objection to the notion of this identity (A.-C. Becquerel 1827, 353-354, citing Seebeck 1826, 140-141). He began the note with the comment that "[w]hat Becquerel here calls contact electricity is, as is well known, what we customarily call thermomagnetism" (353).
33 Seebeck 1825a, 296-297 = 1825b, 32-33 (quote on 297 bzw.33).
34 For a sampling of such usage, see Seebeck 1825a, 312, 330, 334, 338 = 1825b, 48, 66, 70, 74, among many occurrences.
35 See the works cited in note 4, above, which identifies a number of commentators who have analyzed Ritter's work with greater sophistication. Not being then sensitive to the distinction, my reference there (Caneva 1997, 95, n.60) to authors who cited "Ritter's discovery of ultraviolet light" as evidence for the influence of Schelling's Naturphilosophie included six citations (out of eight) that were in fact to Ritter's discovery of ultraviolet rays or radiation. For the "discovery of ultraviolet light," see Hermann (1987, 58) and Eichner (1982, 23). Berg paired Ritter's "discovery of UV radiation" with Herschel's prior "discovery of infrared radiation" even as the schema he drew up nicely displays the complexity of the polarities that underlay Ritter's understanding of the phenomena (Berg 1976b, 33, 32, respectively). Breger, among those who spoke of Ritter's "discovery of ultraviolet rays," noted Ritter's interpretation of the phenomena in terms of the polarity between oxidation and reduction processes without saying that Ritter saw his discovery as the chemical polarity of light (Breger 1999, 28). The body of McRae's account is faithful to Ritter's way of thinking, though in his bibliography he described Ritter 1806 as "his paper on his thought processes in the discovery of ultraviolet radiation" (McRae 1975, 474, 475). For Ritter's discovery of invisible chemically active rays beyond the violet end of the spectrum see Rosenberger 1887-1890, 67, and Whittaker 1951, 100. Three recent biographical dictionaries employ what appears to have become canonical language: "he discovered the ultraviolet rays in the spectrum by means of its darkening effect on silver chloride" (Muir 1994, 436); "from the darkening of silver chloride in light he discovered ultraviolet radiation" (Porter 1994, 583); "he discovered ultraviolet radiation by its darkening effects on silver chloride" (Williams 1982, 445; article by Colin Russell).
36 Ritter 1801b, 527.
37 Ritter 1801a, col. 123. Prompted by the translation of this passage in Caneva 1997, p. 45, Andreas Kleinert informed me in a personal communication that the obscure words "immerdasselbe" as printed in the original (in the last phrase quoted here) are a misprint for the (grammatically irregular) "inner dasselbe," as he confirmed by examining the corrected copy of the text in the Erlangen University Library.
38 Ritter 1802, 410-411, 414.
39 Wollaston 1802, 379 = 1803, 100, in a footnote.
40 Wollaston 1802, 380 = 1803, 100.
41 To the canonical luminous, thermal, and chemical rays Becquerel added phosphorescent (or phosphorogenic) rays, i.e., those that excite phosphorescence in certain substances (E. Becquerel 1842, 341, 342). On the basis of certain distinctive phenomena Moser argued explicitly against the identity of the rays of light and heat (Moser 1842d = 1843a).
42 The quoted words are from Melloni 1842, 136. Moser did not pronounce himself in favor of any theory as to the nature of light, though his mild support of the wave theory is suggested by his favorable citation of J. F. W. Herschel, who "considers it possible that some animals, viz. insects, do not receive an impression from any of those colours which are visible to us, but are affected by a species of oscillations which lie beyond the limits of our senses" (Moser 1842a, 199; quoted from 1843, 437).
43 Moser 1842b, 569; cf. 1842c, 2 ("invisible light rays"). Hermann Helmholtz echoed these views in an address of 1852: "[P]hysics teaches us that there is also light that we do not sense, invisible light, i.e. radiations that proceed from luminous bodies, notably the sun, have entirely the same laws of motion as light, are subject to exactly the same phenomena of interference, reflection, refraction, diffraction, polarization, and absorption, and, with regard to their whole physical behavior, differ from visible light only by a somewhat different magnitude of period of oscillation and refrangibility" (Helmholtz 1852, 13; quoted from 1882-1895, 2 , 603).
44 E. Becquerel 1842, 347; quoted from 1843, 542.
45 Brücke 1845a, 274 = 1845b, 605-606.
46 After having termed the phenomenon "dispersive reflexion," he added in a footnote that he was "almost inclined to coin a word, and call the appearance fluorescence, from fluor-spar, as the analogous term opalescence is derived from the name of a mineral" (Stokes 1852, 479). In the event he mostly used the term quot;internal dispersion."
47 Stokes 1852, 557.
48 J. F. W. Herschel had introduced the term "ultra-violet ray" in 1840, but he did so with such ambivalence that he himself did not further employ it, and no one else-Stokes included-seems to have picked up on his nonce usage (Herschel 1840, 20).
49 Eisenlohr 1854, 623-624.
50 Helmholtz 1855, 206 ("ultraviolette Strahlen"), 208 ("überviolettes Licht"); Esselbach 1855, 757 ("ultraviolettes Licht"; "das Ultraviolett"). The first usage I've found in French is Mousson 1861, 239 ("rayons ultra-violets"). Mascart (1863, 789) spoke of the "spectre solaire ultra- violet," and by 1867 Edmond Becquerel was regularly applying ultra-violet" and "infra-rouge" to the respective spectra and rays (1867-1868, 1 , 138-145). For a few years John Tyndall wavered between "extra-violet" and "extra-red" rays (Tyndall 1864, 329; 1866, 16) and "ultra- violet" and "ultra-red" rays (Tyndall 1865a, 44; 1865b, 5-6) before deciding in favor of the latter pair (Tyndall 1868, 437, 441; 1873, 127-141).
51 As of when I break off this story, Ritter had not yet become either the inevitably cited or the unique discoverer of ultraviolet rays. All I am confident in asserting is that, sooner or later, Ritter did attain that privileged status, as witnessed by his place in contemporary histories and the complete disappearance of Wollaston as in any way involved in the discovery. My purpose here has been to trace the historical preconditions for his having attained the status of discoverer of ultaviolet light.
52 The fullest account of certain aspects of the process described here is Augustine Brannigan's Social Basis of Scientific Discovery (1981), the more explicit consideration of which I must defer to the longer version of this paper I intend to publish. His very similar goal was to "explain how certain achievements in science are constituted as discoveries and not how they occurred to an individual" (11), whereby "discoveries are social events whose statuses as discoveries are retrospectively and prospectively objectified" (133). His answer involved an analysis of "the role of the social recognition in the constitution of a phenomenon's identity," as a result of which process "members of a society in one instance socially construct an event as a discovery only to later orient to it as a natural fact of life" (133).
53 All three men were inspired by the conceptual resources of Naturphilosophie to look for polarities, but that concern has left no traces on the canonized characterizations of their discoveries, in large part, it seems to me, because we do not regard polarities and dichotomies as one of the fundamental regulative or constitutive concepts necessary for the comprehension of phenomena. Significant in this regard is the fact that one of the most accurate and perceptive accounts of polarities in Ritter's work-one which described them in terms of the transformation of quantity into quality-came from an East German (Berg 1976a, 73). In a later essay, Berg and Richter commented that "[i]t has been too little noted up to now that Ritter was one of the first people during the time of the development of dialectics by German classical philosophy to recognize objective dialectics in nature-even if still in naturphilosophisch disguise-and to work according to the dialectical method" (Berg and Richter 1986, 10). What Ritter discovered looks different from the perspective of dialectical materialism.
54 "[U]ntil the scientist has learned to see nature in a different way-the new fact is not quite a scientific fact at all" (Kuhn 1962b, 53). Kuhn appropriately insisted on the importance of scientists' seeing things in a certain way without a conscious moment of interpretation.
55 One result of redescribing a discovery in a later vocabulary is "the linearized or cumulative histories familiar from science textbooks and from the introductory chapters of specialized monographs" (Kuhn 1984, 248 = 1987, 366).
56 "The metaphor of scientific discovery, the idea of dis-covery, is precisely that of uncovering and revealing something which had been there all along. … The crucial part is the prior existence of the discovered object. … The rhetoric of this ontology portrays the objects of discovery as fixed, but the agents of discovery as merely transient"
57 Kuhn interpreted the desire "to recast past developments in the language of modern concepts" in terms of scientists' resistance to entering into an "alien culture," seen as threatening because it "expos[es] the foundations of a previous life form as contingent" (Kuhn 1984, 250 = 1987, 368).
58 One of my reasons for preferring to speak of the collective-not social-construction of scientific knowledge is because 'social' tends to be understood as 'merely social,' in which case the inference against any kind of objective truth appears unavoidable. Although a concept of objectivity as unconnected to specific (and hence contingent) circumstances appears to me to establish its untenability by definition, I believe a reinterpretation of relativism as (precisely) connected to specific circumstances-which, in the case of scientific knowledge, typically means relative to a vast array of repeatedly tested experimental and theoretical claims-can go far to ground the authority of many (most?) scientific claims precisely in the cumulative history of each, which includes the personal experiences and physical experiments of a vast number of observers. One must, however, get past the imputation of 'merely relative' that customarily attaches to the notion of relativism. For a further discussion of these issues, see Caneva 1998b.
59 "[U]nit discoveries are the bricks from which, in the familiar image, the edifice of science is piecemeal built. … The concept of the unit discovery is constitutive of the scientific life as we know it" (Kuhn 1984, 251 = 1987, 369). "These stories are crucial to maintaining the values of the institution of science-the specificity and unique character of the knowledge it produces, for example" Pestre 1999, 205).
60 An important part of the process Fleck described involved the roles of what he called "vademecum science" (Handbuchwissenschaft)-that is, of the selective restatement and canonization of scientific knowledge in compendia, handbooks, and the like, prepared for the professional- and of the "popular science" prepared for the nonexpert, the interested amateur. Essential to popular knowledge-indeed for Fleck originating there-are certainty, simplicity, and vividness (Anschaulichkeit): clear and simple statements shorn of confusion and qualification that convey generally accepted truths about the world (Fleck 1979, 112-115; for the German see Fleck 1980, 146-152).