How to learn about past research practice through iHPS

Before addressing a metascientific research problem by integrating historical and philosophical analyses, it is necessary to establish that the question being dealt with does, in fact, constitute a suitable research topic for iHPS. That is, the research problem must draw attention to both philosophical and historiographical questions. We saw that opponents of iHPS believe that the very selection of philosophical issues and historical events is problematic and leads to methodological flaws. 1 will counter this criticism by showing that such selections can be justified.

Selecting historical instances and philosophical frameworks

In the previous section, I argued that the new mechanical philosophy offers a promising explanatory framework to scholars interested in uncovering the motivation behind the decision of researchers to combine physico-chemical and biological methods. Of course, we should be careful to avoid historical anachronism, a historically incoherent interpretation of the writings and deeds - and we might add. the actions - of the historical figures. According to Jardine (2000, p. 252), being aware of the ‘material, psychological, social and institutional conditions of the production of deeds and works’ helps us avoid vicious anachronism. If nothing else, as Jardine explains, the presuppositions of our modern analytical categories - for example the modern category mechanism as characterised in the new mechanical philosophy - should be realised by the group of actors under consideration. Jardine (2000, pp. 254-5) demands that ‘the categories we apply to past deeds and works should not be at odds with the entire range of the significances attached to them in the past societies in question’. As to our example, we need to establish that the presuppositions of the new mechanical philosophy hold good for the researchers under consideration (ibid., p. 266). We should be able to show, for instance, that botanist Went did indeed believe that in order to explain plant growth he needed to describe the relevant entities involved in plant growth, work out their physico-chemical properties and activities, and figure out how they were organised in such a way that the phenomenon was produced.

The influence of epistemic norms on research practice can best be assessed in cases where the research has been planned from scratch, since the influence of contingent factors will then be minimal. Setting up a new research laboratory comes remarkably close to a state of tabula rasa. When Thomas H. Morgan agreed to organise Caltech’s biology department in 1927, for example, his choice of goals, resources and abilities was relatively unconstrained.³⁷ He was free to choose which fields (and therefore which goals and norms) to include, which instruments to buy (and thereby what resources to provide) as well as who to employ (that is, he could decide the kind of abilities that were needed), and he would have made these choices in accordance with his normative views on what constitutes good research. Moreover, we can be pretty certain that Morgan made his decisions cautiously, knowing that his choices would determine research at Caltech for the next few years at least.

Morgan (1927, p. 86) was eager to establish research projects at the intersection of the biological and physical sciences. He offered the post of Assistant Professor of General Physiology to a young plant physiologist from the Netherlands. Herman E. Dolk. A former assistant to Went, Dolk knew how to perform an intricate plant test that was crucial for isolating growth hormones.³⁸ Another European scientist engaged by Morgan was the English biochemist Kenneth V. Thimann. After their arrival in Pasadena in the spring of 1930, Thimann and Dolk worked together on plant growth substances until Dolk’s untimely death two years later.³⁹ Analysing the early phase of such a cross-disciplinary research project can be illuminating. Collaborative work can generate historical sources that are especially revealing about the researchers’ goals, norms, and capacities: in the above case, for example, the scientists needed to clarify and align their theoretical assumptions, methodological norms, and expertise in order to ensure that joining forces would be effective; and they had to explain their new concepts and methods of working to scientists working in their own discipline.

In sum, philosophical frameworks and historical actors are not selected arbitrarily: we deliberately choose philosophical frameworks with high explanatory power regarding the metascientific research problem to be solved; these frameworks’ general assumptions need to be in accordance with what we know about the goals, beliefs and actions of the scientists under consideration; while we focus on historical actors for whom the phenomena we wish to examine can be expected to be well documented.

Identifying scientists’ research problems, actions, goals, capacities and norms

Once a selection of research projects has been made, the work practices of the scientists and the factors that determined their research problems can be uncovered. The historical sources that allow us to reconstruct scientists’ resources include publications, correspondence, laboratory inventories, receipts, instruments, photographs and films. Morgan, for instance, wrote to Went describing the small laboratory with two underground rooms - ‘suitable for carrying out ... work on growth substances’ - that he had arranged, while Dolk’s correspondence with Went even included floorplans of these facilities.⁴⁰ Occasionally, researchers discussed their abilities explicitly. In his inaugural lecture of 1930, Kogl explained that, unlike chemists, biologists are able to develop physiological tests. Organic chemists, on the other hand, have an idea of what needed to be done to purify hormones.⁴¹ In other cases, one can derive the importance of abilities indirectly: after Dolk died in a car accident in 1932, Morgan declared that they would not be able to continue their phytohormone research ‘unless someone as competent as Doctor Dolk could be found to take his place’.⁴²

To map a field’s goals at any given time, we can consult contemporary textbooks, programmatic lectures and conference programmes. More specific goals can be identified in publications, especially in reviews. In a review of 1927. botanist Peter Stark (1927, p. 91) referred to the isolation of growth substances as a task ‘whose fulfilment ranks among the most urgent research desiderata’. Moreover, researchers’ conceptualisation of problems can. for example, be gleaned from their grant proposals and publications. In their first joint article, Dolk and Thimann (1932, pp. 30-31) wrote that in order to get a better idea of the mechanism of the growth substance’s action, it was necessary to find out more about its chemical structure.

Fortunately, norms can be taken from facts, and a discipline’s abstract norms can be extracted from method discussions, especially those in textbooks.⁴³ Plant physiologist Wilhelm Pfeifer (1900, p. 4), for example, stated:

In explaining the relations between cause and effect, Physiology ... must determine the nature and properties of the different parts affected, as well as the accompanying external factors which may be involved. Both the character of the stimulus to action and the necessary mechanical means by which the resulting phenomenon is affected require to be known ....

Went (1933, p. 137), for his part, wrote: ‘Investigating a factor’s impact on a life process seems to be comparatively simple. All that has to be done is to keep all other factors constant, vary the one. and see the result.’ Passages like these reveal scientists’ views on the criteria that needed to be met to support a claim with adequate evidence. As Jutta Schickore (2017, p. 5) puts it. ‘methods-rela- ted concepts, statements, and reflections as they are presented in experimental reports are significant because they reflect the authors’ understanding of the structure and organization of good experimental research'. Certainly, we cannot take scientists’ programmatic methodological pronouncements as a reliable account of the norms that shaped their actions. To identify the latter, we need to focus on the scientists’ actions and subsequently try to establish that these were governed by specific norms. This kind of procedure is not uncommon in historiography.⁴⁴ In turn, researchers’ actions can be worked out with the help of sources such as laboratory notebooks, correspondence, internal reports and publications. The latter can contain helpful clues concerning researchers’ intentions. Take, as an example, one of Went’s students, who described his experimental work and the reasoning that provided him with

'data concerning the mechanism of the action of the [plant growth] hormone’ as follows:

By applying the substance which promotes elongation and simultaneously preventing the cell from elongation, it proved to be possible to concentrate on the first phases of the process of cell elongation .... Hence it was possible to distinguish between causes and effects of elongation, which had always been the greatest difficulty in the analysis of growth.

(Heyn 1933, pp. 78-79)

Assessment of the findings’ relevance to History of Science and Philosophy of Science

It is hard to see how investigating scientists’ research actions and their conceptualisation of research problems cannot be relevant to the History of Science. Such investigations provide insights into the material, institutional, operational and social conditions of past scientific practice, while also emphasising the role of certain abilities and the significance of new technologies in answering long-standing research questions. It is also interesting - particularly for philosophers - to learn how robust the epistemic norms under consideration have been over time. This can be evaluated by comparing the norms that shaped phytohormone research with the norms discussed in earlier scientific work on similar problems. The fact that Went’s work was so well received in turn substantiates the claim that the standards concerned were widely accepted at the time. We can thus leam about the prevalence of the norms exposed by the new mechanistic philosophy in actual scientific practice.

Crucially, iHPS studies do not end with the investigations of particular local cases of past scientific work. As Scholl (2018, p. 237) puts it, we strive ‘to know how broadly, and in what variations, certain abstract descriptions apply to concrete instances’. Therefore, we should try to obtain more abstract descriptions from our concrete historical studies, since by studying additional concrete historical cases we will be in a better position to assess the empirical scope of these abstract descriptions.⁴’’