Horizon for Scientific Practice: Scientific Discovery and Progress

Abstract

In this article, I introduce the notion of horizon for scientific practice (HSP), representing limits or boundaries within which scientists ply their trade, to facilitate analysis of scientific discovery and progress. The notion includes not only constraints that delimit scientific practice, e.g. of bringing experimentation to a temporary conclusion, but also possibilities that open up scientific practice to additional scientific discovery and to further scientific progress. Importantly, it represents scientific practice as a dynamic and developmental integration of activities to investigate and analyze the natural world. I use the discovery of the clotting factor, thrombin, and the experiments conducted by the Johns Hopkins physiologist, William Howell, on the enzymatic nature of thrombin to illustrate the notion of HSP. In a concluding section, I compare the notion of HSP to other notions for scientific practice proposed in the history and philosophy of science literature.

Acknowledgements

I thank especially Ron Anderson, Patrick Heelan, and two anonymous reviewers and the editor, James McAllister, of International Studies in the Philosophy of Science for insightful comments on earlier drafts of the manuscript, as well as Allan Franklin and Joseph Fruton for helpful discussions. I presented a version of this paper at the International Society for the History, Philosophy and Social Studies of Biology’s 2007 conference held at University of Exeter in Exeter, UK. Baylor University supported my research through a Summer Sabbatical award and a University Research Committee grant.

Notes

[1] The two other epistemological tasks discussed by Reichenbach (1938) are critical analysis task of science per se to understand it and the advisory task in terms of deciding whether evidence adequately supports a theory.

[2] Philosophers of science prior to Reichenbach also make a similar distinction, although he is credited for formulating the distinction in contemporary terms (Hoyningen‐Huene 1987; Losee 2001).

[3] Even Toulmin, for example, contended that when analyzing a discovery in science (that is in physics), the question to ask is ‘what sort of demonstration will justify us in agreeing that, whereas this was not previously known, it can now be regarded as known’ (Toulmin 1953, 17).

[4] For further discussion of Hanson’s logic of discovery, see Kordig (1977).

[5] Kuhn (1996) embedded his critique of the context distinction within a larger analysis of the structure of science. At the center of this analysis is the notion of paradigm, especially in terms of disciplinary matrix and exemplars.

[6] Certainly, this generation of philosophers recognized that scientists do discover objects; but such discovery is problematic, since scientists can only understand objects given theoretical matrixes in which philosophers locate them. The discovery of objects could only be accountable for in terms of their position within theories. And so, verification, confirmation, or corroboration of the theory was necessary to determine whether scientists discovered anything.

[7] For just a few examples of scholarly work in the history and philosophy of experimentation, see Chang (2004), Mayo (1996), Rheinberger (1997), and Weber (2005). For reviews of the field, see Chalmers (1999) and Radder (2003).

[8] L. R. Franklin (2005) extends Steinle’s role for this type of experimentation to include a broader role, and by his own admission more ambiguous, in the generation of scientific knowledge.

[9] Pickering (1995) takes Galison’s notion of constraint to task, emphasizing its negative dimension, as an inadequate concept for understanding scientific practice. Galison (1995) responds that the notion does have a positive dimension. For further discussion on this debate, see Stump (1996).

[10] Traditionally, philosophers and historians of science utilize metaphors to describe scientific practice. For example, Kuhn (1996) used the metaphors normal or paradigmatic science to describe accumulative, commensurable growth of scientific knowledge and revolutionary or non‐paradigmatic science to describe its episodic, incommensurable growth.

[11] The question here, which is seemingly uninteresting or trite with respect to traditional analysis of science, is why progress should accompany experimentation. The horizon for scientific practice helps to address that question in terms of both its temporal and visual dimensions. For, it includes not only a heuristic structure but also provides a regional area in which theories and experiments develop.

[12] For additional details on the historical reconstruction of these events, see Marcum (1996, 1998).

[13] Additional investigators, during the latter part of the nineteenth century, also examined the mechanism of blood coagulation, including Lea and Green (1883) and Halliburton (1888). For an in‐depth discussion of the history of blood clotting during this period, see Morawitz (1958).

[14] The statement, ‘the way the natural world is’, is an ontological statement, but it does not imply that the natural world is strictly or solely mind‐independent. On the other hand, the statement does not refer to the world as a social construction or to a completely mind‐dependent world. These designations are too ontic and fail to capture the ontological nature of the world that makes possible even a mind to investigate it. Rather, the statement refers to a mind‐independent world embedded in a mind‐dependent world view. For further discussion of the debate, often called the ‘science wars’, see Marcum (2008).

[15] For Heidegger, ontic refers to the individual instantiation of the ontological, which makes possible the ontic. Importantly, the ontological dimension of the world makes possible the characteristics of experimentation that allows scientists to discover objects and processes. For example, plasticity of experimentation depends upon a theory‐independent world, in that the world can outstrip theoretical predictions, while experimental reproducibility depends on the stability of natural phenomenon (Marcum 1996).

[16] Although Mulligan, Simons, and Smith (2006) are sympathetic to natural scientists attempting to appropriate ontology for doing science, they chastise contemporary philosophers for not assisting in their endeavours.

[17] See Fruton (1972, 95–131) and Leicester (1974, ch. 14) for a discussion of the development of these techniques.

[18] As part of an expanding horizon for discursive practices, new journals appeared during the nineteenth century. Schmidt and Hammarsten published in such a journal, Archiv für die Gesammte Physiologie des Menschen und der Thiere, founded in 1867.

[19] Physiological chemists knew long before Denis that salt inhibited blood coagulation. For example, as Hewson noted, ‘This property of one of the neutral salts [inhibition of blood clotting] has been long known amongst those who prepare the blood of cattle for food; for it has long been a practice with such people, to receive it into a vessel containing some common salt, and to agitate it as fast as it falls, by which means the coagulation is prevented, and the blood remains so fluid as to pass through a strainer, without leaving any coagulum behind: by this means they have an opportunity of mixing it with other substances for culinary purposes’ (Hewson 1772, 14–15).

[20] While paraglobulin for Schmidt played a significant role in blood clotting, Hammarsten (1877) challenged its role. The latter isolated fibrinogen by a salt precipitation technique that involved addition of an equal volume of a saturated sodium chloride solution to plasma treated with magnesium sulphate. He dissolved the product in water and re‐precipitated with a saturated salt solution. Addition of fibrin ferment, free of paraglobulin, to preparations of fibrinogen resulted in deposition of a fibrin clot. Moreover, Hammarsten isolated paraglobulin and found that the agent does enhance the amount of fibrin formed upon mixture of fibrin ferment and fibrinogen; however, the enhancement in the presence of paraglobulin was not specific, in that other proteins, such as the milk protein casein, had the same effect.

[21] For further discussion on the characteristics of experimentation, see Marcum (1996).

[22] According to Berzelius, ‘It is common knowledge that the conversion, for instance, of sugar into carbon dioxide and alcohol takes place during fermentation under the action of an insoluble body which we call ferment … I propose calling the power causing the conversions above mentioned under the conditions indicated catalytic, contrasting the word catalysis with the term analysis’ (Suner 1955, 58).

[23] Verso, in an analysis of Buchanan’s work, claims that the importance ‘of his second series of experiments, however, does not lie in the additional observations he made, but in the conclusions he drew from them’ (Verso 1960, 577).

[24] According to Fruton, ‘it was not clear whether the “chymosin” (from the calf stomach) … was identical with Schwann’s pepsin’ (Fruton 1972, 71).

[25] The development of the catalysis theory for fermentation was most intense from the middle to latter half of the nineteenth century (Fruton 1972).

[26] Laudan also describes theories as answers to questions or as solutions to problems. But, the approach herein as to how theories function as answers differs from Laudan’s approach, in one crucial detail. For Laudan, ‘empirical’ problems or questions are treated ‘as if they were problems about the world’ (Laudan 1977, 15). My contention is that such questions are empirical because scientists use experimental tests that depend on the ontological state of the natural world to answer such questions. Thus, there is an independence enjoyed by the empirical from the rational, as well as dependence. Currently, philosophers of science emphasize the latter almost to the exclusion of the former.

[27] Obviously, the progression from liquefied clot to thrombin is also an important example of scientific progress. However, I use Howell’s scientific work on thrombin because his theory of clotting is an excellent example rebutting the notion that only ‘good’ science is progressive (Popper 1963). Commentators considered his theory detrimental to scientific progress within the field of blood coagulation. For example, Jaques claimed that Howell’s theory of coagulation was not only misdirected but also hindered progress, in that editors of various journals refused to publish articles contrary to that theory. ‘The Quick prothrombin time is the basis for the various coagulation tests conducted in clinical blood coagulation laboratories all over the world … The original manuscript describing this test’, bemoaned Jaques, ‘was submitted to the editors of eight journals and in each case it was not accepted on the basis that it did not agree with the Howell theory!’ (Jaques 1988, 149–150). And yet, the theory was important for the discovery and development of heparin (Marcum 1990). Thus, Howell’s choice between two clotting theories illustrates how science progresses, warts and all.

[28] Losee also distinguishes between the descriptive and normative nature of theories of scientific progress. The notion of horizon for scientific progress is simply neither descriptive nor normative but rather, in terms of Darden (2006), ‘advisory’. In other words, the notion of progressive horizon does not provide an algorithm for how to progress scientifically, but it provides more than a description of scientific progress. If horizons are not being exposed, then science is not progressing.

[29] By calling Howell’s experiments a pedagogical exercise is not to belittle their importance. Obviously, such experiments are the beginning of a process by which a student gains the necessary experience and skill crucial to become a practicing member of a scientific community. Through education, both in terms of textbooks and laboratory manuals, students identify personal and community limits that serve to establish a horizon for scientific practice (Marcum 2008). For example, Howell was well aware of the immense problems associated with the blood coagulation field, prior to his re‐entry into this discipline, at the beginning of the twentieth century (Erlanger 1951).

[30] Such background information serves as a theoretical component of the notion of horizon for scientific progress in which the significance of previous work limits investigators. However, such information often serves as a limit that investigators must traverse by further experimentation.

[31] For a later review of Nolf’s series of experiments, see Nolf (1938).

[32] For example, Howell (1910a) did not consider thrombin to be a product of clot formation but an agent responsible for the process itself.

[33] Howell did not specify from which text of Oppenheimer he took this definition, but Oppenheimer (1900) is a very influential work on enzymes that went through several editions.

[34] For example, Bayliss claimed that catalysis is the primary characteristic of enzymes. ‘Apart from theory’, contended Bayliss, ‘it is useful to know what kind of properties to look for in a substance suspected to be an enzyme. An unknown body, if an enzyme, may be expected to manifest the general characters of catalysts to a greater or less degree’ (Bayliss 1911, 11). He also discussed two criteria to identify the catalytic nature of an enzyme: (1) altering the rate of the reaction to produce the final product and (2) not participating materially in the final product.

[35] Howell stated, ‘As in the case of enzymes, the reaction caused by the thrombin is incomplete owing to the fact that the end product of the reaction, the fibrin, inhibits in some way the further action of the thrombin’ (Howell 1910b, 465–466). However, he went on immediately to reject the notion that thrombin is an enzyme and to claim that further experimentation is necessary to examine the mechanism by which thrombin at higher concentration deposits fibrin.

[36] Seegers, Nieft, and Loomis (1945) showed that fibrin actively adsorbs thrombin during clot formation and proposed that this adsorption explains the quantitative relationship, observed by Howell (1910b) and by Rettger (1909), between thrombin and fibrinogen during fibrin formation. Interestingly, the date of this publication corresponds to the year of Howell’s death.