This article examines an often neglected topic in the history of science, namely clinical observation, specifically the objectivity and knowledge production associated with therapeutic trials. It will describe an eighteenth and nineteenth century pharmacological concept of objectivity and exemplify that concept using late nineteenth century European cocaine research. As conceived within clinical drug research, this concept of objectivity does not correspond with those described by Daston and Galison in their seminal book Objectivity (2007). I will explore the implications of this “new” concept of objectivity—as compared to the “old” ones postulated by Daston and Galison—for their core propositions.
Pethes et al. (2008) have described the lack of systematic examination of the “experimental observation” (pp. 11f) of humans in different academic and social realms since the eighteenth century as an important desiderata in the history of science. While there are important studies on scientific observation and objectivity (Daston and Galison 2007; Porter 1996), they tend to ignore clinical observation.1 This article is devoted to the neglected themes of clinical observation, objectivity, and knowledge production in the context of therapeutic experiments.2 The much acclaimed and criticized book Objectivity (2007) by Lorraine Daston and Peter Galison (Ash 2008; Pettit 2009; Haskell 2010; Stoliarova 2010; Shah 2017) has opened a fascinating perspective on scientific objectivity and subjectivity, which I intend to critique and modulate by introducing a pharmacological concept of objectivity. That concept influenced clinical drug trials in the eighteenth and nineteenth century and does not correspond with the objectivity concepts described by Daston and Galison. While both Daston and Galison deal with the natural sciences, they ignore clinical research. And yet, in spite of this lacuna, their statements lay claim to general validity. The concept of objectivity explored in this article challenges that claim to validity. Beyond other more widely debated aspects, such as scientific atlases as historical sources or epistemic fear, this article is intended to contribute to the debate on Objectivity and its Critics (Daston and Galison 2008; Padovani et al. 2015) and, in the spirit of constructive criticism, relativize and expand the interpretation expounded by Daston and Galison (2007). Contrary to Hacking (2015), who claims that the only sensible approach to studying objectivity is to study the word in the context of its usage, this article examines an enduring eighteenth century concept of objectivity that was not articulated as such until the late nineteenth century. That no specific term for a concept or phenomenon (such as “objectivity”) yet existed does not necessarily mean that they did not yet exist (in the consciousness of the historical actors); even though—as a result of, among other things, different modes of expression—they were likely not represented in an entirely identical manner.
In this article, I will explore a different concept of objectivity used in clinical drug research, examining first the trial rules as described in late eighteenth century theoretical works on pharmacology. I will engage in a close reading and explication of the rich source material in order to avoid the historiographic problem of more synthetic narratives, which usually tend to convey more of one’s own terminology, viewpoints, and style of thought than of the historical actor’s. I will then draw on late-nineteenth century therapeutic cocaine trials to discuss what this “new” concept of objectivity—as compared with the “old” one posited by Daston and Galison (2007)—implies for their core propositions.
2. Theoretical Foundations of the Clinical Experiment: The Eighteenth Century Pharmacological Concept of Objectivity
In order to develop a better understanding of the continuity of the eighteenth century pharmacological concept of objectivity, some introductory remarks are needed about nineteenth century coca and cocaine research, in which the concept lived on, before returning to its origins and historical legacy. From the middle of the nineteenth century, coca leaves had been systematically studied. One result of this research had been the extraction of cocaine, which the young chemist Albert Niemann (1860) likely first achieved. Thanks to the ready availability of cocaine, which began to be produced by the pharmaceutical company Merck (Darmstadt, Germany), the overall number of trials increased after 1862 and, for the first time, researchers also conducted animal trials in their coca research. But human trials nevertheless continued and both medical students and pharmacological experts continued to conduct experiments on themselves using cocaine and other coca preparations (Schroff 1862; Bennett 1873; Christison 1876; Anrep 1880). Broader awareness of cocaine as a new medical drug arose in the 1880s, following Sigmund Freud’s cocaine studies (especially 1884) and Carl Koller’s3 discovery of its properties as a local anesthetic for use in ophthalmology (Koller 1884). Thereafter, in Europe and North America, cocaine was widely studied as a remedy for all kinds of symptoms (Lindemann 2018).
In European coca and cocaine research in the second half of the nineteenth century, we can discern objectivity criteria that, as regulative ideals and to differing degrees, influenced cocaine trials. Certainly, the requirements for reliable observation, and methodological aspects in general, were rarely stated explicitly in scientific publications on coca and cocaine trials, but we can nevertheless draw conclusions about the implicit norms of practice. These conclusions are based on 1) isolated methodological remarks made by different coca and cocaine researchers that reveal their scientific ideals, specifically their objectivity ideals (Aschenbrandt 1883; Dowdeswell 1876; Erlenmeyer 1887); 2) justifications for breaking the rules that signify unadhered to norms (Obersteiner 1885; Wallé 1885); as well as 3) the neurologist and morphine expert Albrecht Erlenmeyer’s critique about the lack of objectivity in the clinical cocaine trials of his colleagues, from which we can also garner information about the disregarded objectivity criteria (Erlenmeyer 1885, 1886, 1887).
The implicit norms of practice acquired in this way consist of individual objectivity criteria, that together comprise a coherent concept of objectivity.4 The key objectivity criterion is to capture the phenomenon being studied as it “really” is, in other words to capture the actual effect of the remedy, in this case cocaine (Erlenmeyer 1887, p. 155).5 This aim and core criterion implies and necessitates all other objectivity criteria that, in turn, must be satisfied by both objective observation as well as objective interpretation (observation-based, general validity, reproducibility6, verifiability, truthfulness, completeness, comprehensibility7). The objective observation is independent of the individual observer, and thus reproducible and therefore repeatable and verifiable. The objective knowledge that is derived from the objective observation and that captures the phenomenon being studied as it ‘really’ is (without distortion), is generally valid. Repeatability and verifiability demand in turn a full and comprehensible report that is honest with regard to all relevant aspects of the trial (Erlenmeyer 1887, pp. 448–49). In the 1880s, this concept of objectivity was already established and can be traced back at least to the end of the eighteenth century. At the time, the specific objectivity criteria were articulated as trial rules by various authors of theoretical works in pharmacology (without using the term “objectivity”). Over the course of the nineteenth century, these trial rules, i.e., objectivity criteria, were repeatedly formulated, modified, and expanded upon by various physicians in medical handbooks and textbooks (at least in the tradition of leading German-language pharmacological research).8 These objectivity criteria were the standards used in contemporary drug testing. I turn now to eighteenth century trial rules in order to explicate this pharmacological concept of objectivity.
The physicians Johann Christian Reil ( 1817)9 and Adolph Friedrich Nolde10 (1799) criticized the current state of pharmacology, in particular the unreliability of clinical observations and fought to establish generally valid, normative standards for drug trials. With this aim in mind—and independently of one another—they devised similar rules and principles, proceeding from the assumption that:
The only way to enhance pharmacology is therefore to conduct trials, to carefully document the results, and to subsume the isolated experiences under higher laws. The principles designed to achieve these things must be set out in advance so that our efforts may have a compass and our results a rule, according to which they can be assessed. (Reil 1817, pp. 150–51; see also p. 144)
2.1. 1) Rules on the Conditions or Prerequisites for Reliable Trials
To acquire reliable observations about a medication’s effects, one needed a suitable observer, suitable probands, and a suitably prepared medication. The suitable, i.e., good observer was a trained physician and talented scientist with unimpeded abilities of perception and an adequate ethical bearing:
In order to observe, the observer must have good senses, intellect, judgement, erudition as well as a sufficient degree of skepticism. He must be guided only by the love of truth and not by self-infatuation, prejudice, tradition or by devotion to a system or school. (Reil 1817, p. 138; see also Nolde 1799, part 1, pp. 84–85).
Beyond sufficient knowledge in his field,… he must also possess a deep-seated empirical spirit and … knowledge of human nature …, although an agreeable and pleasing demeanor is also very often helpful. (Nolde 1799, part 2, p. 85).
For drugs to be suitable in trials, they “must be of the best quality, authentic and uncorrupted” (Nolde 1799, part 1, p. 57). Reil (1817, pp. 138–39) insisted that the medication being tested not change over the course of the trial; it’s composition and quality had to remain the same. To ensure these criteria, the observer had to “be equipped with sufficient botanical, chemical, and pharmaceutical knowledge” (Nolde 1799, part 1, p. 63). He had to be knowledgeable about “simple as well as composite substances” (p. 58) in order to evaluate them and not be duped by the widespread fraud and errors of the apothecaries (pp. 57–63). Ludwig Büchner (assistant physician at the university hospital in Tübingen) described the unreliability of the medicinal preparations—being rarely inspected even in hospitals and private practices—as a source of error in therapeutic experiments (1854, p. 304). Likewise according to the pharmacologist Karl Damian von Schroff (1873, pp. 49–50), the authenticity of the remedy to be tested had to be ensured and the physician had to be familiar with its type, provenance, and chemical characteristics. Furthermore, the observer needed to select suitable, i.e., similar probands, make the correct diagnosis, and be familiar with the illness he wished to treat in trials with the new remedy:11 “If we experiment on patients, they must be as similar as possible to one another. The illness must be true and not hypothetical, discernible on the basis of sensory evidence, and fully recognized by the experimenter as being actually present” (Reil 1817, p. 139); “To be able to correctly describe the effect of a trial remedy, one must above all know which patients and medical circumstances are liable to produce meaningful results” (Nolde 1799, part 2, p. 75).
2.2. 2) Rules for Conducting Trials
For the trials to be conducted adequately, “each trial remedy had to be applied as simply and purely as possible” (Nolde 1799, part 1, p. 67). Nolde’s claim referred to the composition of the drug in question, not to its exclusive use. Furthermore, “to examine and establish a drug’s effects on an illness, [we] must also appropriately prescribe it” (p. 64). Appropriately prescribing the drug involves the form in which it was administered, as a powder, pill, or liquid (pp. 74–6), as well as its dosage (pp. 64–5). Schroff (1873, pp. 49–50) too remarked on the need to consider the appropriate amount and the proper application—in terms of form and timing—of the prospective drug. Furthermore, the drug to be tested had to fit the patient’s prescribed diet and lifestyle, and vice versa (p. 72), meaning, for example, that a patient with high blood pressure would not be an ideal candidate for testing an especially potent stimulant. Like Georg Friedrich Hildebrandt12 (1786, pp. 83–4) and Schroff (1873, p. 49), Reil too devised a rule to ensure that only the drug to be tested be used: “The substance must be tested alone and not in combination with other ones, because otherwise it remains uncertain which of the applied substances produced the elicited effect” (Reil 1817, p. 140). Nolde (1799, part 2, pp. 98–101) likewise believed that reliable observations depended upon no other drugs being used, although he emphasized that the necessity of treatment did not always permit this course of action.13
The rules were designed to ensure quality research, without disrupting treatment or endangering the patient’s well-being. Nolde raised the fact that the demands of research and treatment were often contrary and that treatment often took priority, for example his rule that “One shouldn’t undertake trials with new drugs if one can still rely on known and proven ones, and in any case every new remedy should be applied with the greatest of care” (Nolde, part 1, p. 86). Schroff (1873, p. 49) too devised the same rules, namely, that a trial not harm the patient, that new remedies be applied carefully, and that no trials be conducted if existing remedies were available. More so than Reil and Nolde, Schroff stresses the necessity of conducting preliminary trials in animals and healthy people as part of an “ideal trial sequence,” as Störck (1731–1803) had already called for (Winau 1986, pp. 89–91).
A further rule formulated by both Reil and Nolde stipulated that, to produce reliable results, the trials be repeated frequently under the same conditions:
The trials must be repeated frequently under exactly the same conditions and, in each case, generate the same results. Only this can convince us that the results are the effects of the medication. If, after one or several trials, a certain phenomenon … sometimes occurs and sometimes doesn’t, it always remains a possibility that, after applying the substance, the phenomena occurred not because of it, but because of some another arbitrary cause. (Reil 1817, p. 140; see also Nolde 1799, part 2, p. 102)14
If one is to be certain that a drug is useful in treating some diseases, then just as many similar cases as are treated with the remedy would have to remain untreated under the same conditions. Only then, having two such sets of observations, would a satisfactory difference—to the benefit or detriment of the tested remedy—emerge. (Schroff 1873, p. 50)
Nolde noted that the trials also needed to be repeated under altered conditions: “In order to be as certain as possible about the effects of a drug, we must repeat our trials multiple times under similar, and sometimes even dissimilar conditions” (1799, part 2, p. 102). The Italian physician and natural scientist Felice Fontana (1787), who conducted animal trials with snake venom and curare and who significantly influenced the theory of experimentation (Winau 1986, p. 91), also insisted upon the reproducibility of observations, i.e., the repetition of the trials under the same and different conditions, as well as their simplicity:
The first [means for avoiding errors] is to repeat the trials as often as possible.… The second involves modifying the trials in countless ways, altering the conditions as demanded by nature and by the trials’ arrangement, and ensuring they be as precise and simple as possible. This second one demands far more skill and acumen from the observer than the first does; and there are only a few observers … who can boast to have always been so diligent. The third is to … also discover the source of other’s errors. (Fontana 1787, pp. 68–9)
One must be well acquainted with the patient’s character, the illness being treated by the remedy …, and the stage of the illness when the remedy was applied. One must also duly consider all other relationships and circumstances of the patient’s life. Only then … will [the observer] be able to judge, after the substance has been taken, which subsequent changes can be attributed to it (Schroff 1873, p. 50)
On the morning of the 19th, entering the room at an entirely unusual time, I examined him: he sits in front of the nightstand with a bottle of Eau de Cologne on it, his thigh is exposed and just as I arrive he throws something into the corner. I go toward this discarded object and discover a hypodermic syringe (Pravazʼsche Spritze).… I say “that is a morphine solution” and the patient’s face turns ashen, he breaks down, trembles, and stammers “yes, yes, go ahead and take it.” (Erlenmeyer 1887, pp. 376–77)
2.3. 3) Rules Governing the Documentation of the Trial
Finally, not only did the trials have to be exactly observed, they also had to be exactly logged and documented, requiring a record of all relevant circumstances, truthfulness, and unambiguous terminology. Nolde (1799) emphasized the necessity of a complete and truthful description of the trial, accounting for all relevant issues: “When announcing new remedies or recommending old ones, one must withhold nothing that can in any way influence a correct assessment of a drug and be ashamed of fabricating or distorting observations at the expense of the truth” (part 2, p. 108). Reil (1817) chiefly stressed the necessity of precisely describing the drug’s effect: “the effects of the substance must be provided not in general, but in detail …. It is not a question of whether it works, but of what the effect is and how it is achieved” (p. 141). Fontana (1787) states that it is the “duty for every observer … not only to faithfully repeat prior trials that contradict him, but also to recount his own trials in a way that leaves no room for doubt in his reader’s mind” (Fontana 1787, pp. 68–9). Jörg too calls for a “conscientious and truthful documentation of the trials” (Jörg 1822–24, cited in Gerken 1977, p. 14). In nineteenth century pharmacological reference works and textbooks, the rules governing trials rarely included any documentation of the trials, but those rules were manifested in the institutional documentation practices of hospitals and clinics.
3. The Pharmacological Concept of Objectivity and the Propositions of Daston and Galison
Eighteenth century pharmacologists were already complaining about the gap between theoretical norms and clinical trial-practice, a gap that for the most part persisted until the standardization of the clinical drug trial in the late twentieth century. To be precise, in a sense this gap persists, despite standardization and new research contexts. Researchers today still have some leeway in applying modern trial rules (Stegenga 2011). However, in the context of nineteenth century coca and cocaine research, the trial rules and objectivity criteria that I have just described were used by various asylum physicians, and to varying degrees, as regulative ideals. Their use depended on the extent of their institutionalization, i.e. on common practices of treatment, research, and documentation, as well as on the specific possibilities and interests of the researchers using them (Lindemann 2018, chapter 2.3). But the use of objectivity criteria in research practice in not my primary focus. Instead, I will now compare the pharmacological concept of objectivity in normative terms with that described by Daston and Galison (2007).
Let us begin by asking how the efforts of coca and cocaine researchers differed from other contemporary efforts at generating objectivity, in order to determine what was specific about their concept of objectivity. It’s worth noting that, in the clinical trials to test cocaine’s use by morphine addicts (Richter 1885; Obersteiner 1885; Smidt and Rank 1885, Erlenmeyer 1885), objectivity would be achieved not primarily by quantitative procedures, nor by laboratory tests or precise measurements. Such would have been the approach adopted by other contemporaries in order to produce objective clinical observations and results (Matthews 1995; Porter 1996; Hess 1997; Daston and Galison 2007; Ledebur 2015). Instead, objectivity would be attained solely by rule-governed experience, requiring—at least in theory—that both observations and interpretations satisfy objectivity criteria. Systematic measurements designed to generate objectivity were fairly prominent in Erlenmeyer’s work (1885), but he was an exception. In the context of the cocaine trials, measurements (for example of pulse and temperature) comprised an essential component of objective observations, but they were not the primary means for producing objectivity. This plurality of approaches aimed at generating objectivity again points to the heterogeneity and diversity of clinical research on drugs in the late nineteenth century (Hall 1997). This contradicts the claim that “mechanical objectivity” (see below) was the “ubiquitous epistemic norm of the age” (Ledebur 2015, p. 41). We can further illustrate this by comparing it with the research results of Daston and Galison (2007).
What cocaine-researchers understood “objectivity” to mean does not accord with the concepts of objectivity that Daston and Galison (2007) describe as characteristic of the time. While both Daston and Galison deal with the natural sciences, they ignore clinical research; and yet, in spite of this lacuna, their statements lay claim to general validity. They describe “mechanical objectivity” as the earliest form of objectivity, not emerging until the middle of the nineteenth century: “Objectivity the thing was as new as objectivity the word in the mid-nineteenth century” (Daston and Galison 2007, pp. 34–5). In contrast to “mechanical objectivity,” which fundamentally sought to eliminate subjectivity (“objectivity is blind sight, seeing without inference, interpretation, or intelligence,” p. 17), the objective clinical observation as understood and practiced by cocaine researchers was based on interpretation. Perceiving and grasping any relevant evidence depended entirely on the continuous interpretation of symptoms.21 Thus, clinical observation meant intelligent sight. The objects being observed were, after all, themselves subjects who were able to present themselves to the observer in non-authentic ways, to (dis)simulate, to themselves disregard or play-up their symptoms. The second concept of objectivity that Daston and Galison (2007) describe is called “structural objectivity.” This has nothing to do with clinical observation because “structural objectivity … has no truck with any kind of seeing, be it four-eyed sight or blind sight” (p. 256; see also p. 259). The ideal and the practice of the objective clinical observation undertaken by the cocaine researchers corresponded far more with a “code” (p. 18) or paradigm that, according to Daston and Galison, did not emerge until the twentieth century and that they call “trained judgement” (“judgement-inflicted vision as a goal for scientific sight,” p. 311).
The cocaine trials and the pharmacological concept of objectivity they rely on call into question the general validity of several key claims of Daston and Galison (2007), namely 1) that scientific objectivity doesn’t emerge until the middle of the nineteenth century,22 2) that “mechanical objectivity” is the earliest form of objectivity (pp. 197–98),23 as well as their claims 3) about the formative sequence, transient dominance, and distinctiveness of the postulated codes or paradigms that result from fixing the temporal origins of specific paradigms in sequential decades or half-centuries.24 For example, “truth-to-nature” is supposedly a code of the eighteenth and nineteenth centuries, “mechanical objectivity” is supposed to have emerged only in the mid-nineteenth century, “structural objectivity” at the end of the century, and “trained judgement” in the early twentieth century.25 The rules governing eighteenth century pharmacological trials already point, at least in theory, to a comprehensive concept of objectivity that, according to Daston and Galison (p. 29), did not yet exist and that had nothing in common with the “truth-to-nature” paradigm. This concept of objectivity, which was still present in the cocaine trials of the 1880s, contradicts the concepts of objectivity that Daston and Galison claim existed at the time, exhibiting instead characteristics of the “trained judgement” paradigm that, by their account, had not yet emerged. But Ledebur (2015, p. 49) has described emblematically a paradigm shift from “mechanical objectivity” to “trained judgement” in psychiatric terminology at the beginning of the twentieth century. The clinic was influenced by these paradigms, although the existence of alternative concepts of objectivity suggest that the paradigms described by Daston and Galison cannot be taken to have been the only existing or operative ones.
This being said, I need to make another point concerning the role of “subjectivity” in the emergence and existence of scientific objectivity: According to Daston and Galison (2007), scientific objectivity and subjectively mutually condition and define each other: “there is no objectivity without subjectivity to suppress” (S. 33). “Subjectivity is as historically located as objectivity; they emerge together as mutually defining complements” (p. 5; see also pp. 37, 258). According to Daston and Galison, scientific (mechanical) objectivity arose in the mid-nineteenth century in response to a new epistemological problem: the subjectivity of the researcher:
The emergence of scientific objectivity in the mid-nineteenth century necessarily goes hand in glove with the emergence of scientific subjectivity. Subjectivity was the enemy within, which the extraordinary measures of mechanical objectivity were invented and mobilized to combat. It is no accident that these measures often appeal to self-restrained, self-discipline, self-control: it was no longer variable nature or the wayward artist but the scientific self that posed the greatest perceived epistemological danger. This untrustworthy scientific self was as new as objectivity itself; indeed, it was its obverse, its photographic negative. (p. 197–98; emphasis added)
If one compares the pharmacological and mechanical concepts of objectivity, an obvious difference arises in the relationship to subjectivity. The ideal of pharmacological objectivity doesn’t rest on the elimination of every subjective aspect of the observer, and it should be applied using different means than the ideal of mechanical objectivity, namely the observer’s scientifically informed ability to perceive and interpret (supported by, but not in lieu of measurements and mechanical evidence). And yet the two concepts do have something in common: the pharmacological ideal is based on the elimination of certain aspects of subjectivity; furthermore, its aim of capturing the research object as it really is, without distortion, matches the aim of mechanical objectivity (Daston and Galison 2007, p. 20). With regard to the drug’s effect, this means that it cannot be captured without (the unavoidable) alterations caused by observers and probands. These alterations, which an adherence to the trial rules is designed to avoid, are effects of a phenomenon that we would call projective perception. This becomes clear, for example, in Hildebrandt’s warning about the fallacy Post hoc, ergo propter hoc:
One overlooks in the case of false experience that … phenomena can occur simultaneously or in quick succession of one another without one having any connection whatsoever with the other: that phenomenon b [for example, recovery] can have entirely different causes than phenomenon a [for example, the application of a remedy], even though we have only just observed it. Incidental causes can prevent one from seeing the other phenomenon: or prejudice can impede us by blinding us so that we only see what we want to see and nothing else. (Hildebrandt 1786, p. 42)
Staring in the mid-nineteenth century, men of science began to fret openly about a new kind of obstacle to knowledge: themselves. Their fear was that the subjective self was prone to prettify, idealize, and, in the worst case, regularize observations to fit theoretical expectations: to see what it hoped to see. Their predecessors a generation or two before had also been beset by epistemological worries, but theirs were about the variability of nature, rather than the projections of the naturalist. (p. 34–5; see also pp. 135, 203, 242; emphasis added)
Unlike mechanical objectivity, in eighteenth and nineteenth century clinical drug research subjectivity is not the source of error. The trial rules are principally designed to minimize all sources of error, i.e., to counter all potentially confounding effects that, in addition to the illness itself, could alter the drug’s impact. These effects can have causes other than the subjectivity of observer and proband, such as other as drugs or elixirs taken in addition to the remedy being tested. One difficulty of drug trials is that numerous potentially confounding factors can be neither generalized nor anticipated and can only be discovered, if at all, by a rigorous examination of the proband (including their physical and mental state) and their living conditions (see above, pp. 9–10). Nevertheless, the observer’s subjectivity was deemed to be an especially dangerous source of error. All of the terms that, in the broadest sense, are related to the phenomenon we call projective perception are negatively connoted, such as “egotism” (Fontana 1787, p. 67), “preconceived opinions” (pp. 70–1), “prejudice” and “partiality” (Hildebrandt 1786, pp. 84–5), “authorities” (Reil 1817, p. 138), “fallacy” (Nolde 1799, part 1, p. 56), and “self-deception” (part 2, p. 97). The observer’s projective perception is entrusted to form and dominate the observation itself, to give it content and structure (“prejudice can impede us by blinding us so that we only see what we want to see and nothing else,” Hildebrandt 1786, S. 42). Observers were often reminded that they should conduct their trials “without prejudice, and with the intent of seeing correctly” (Fontana 1787, p. 67). Avoiding projective perception is also considered to be a precondition for finding sources of error that do not arise from subjectivity.26 In addition to admonishing discipline, the main trial rules serve to curb precisely this source of error in the observer’s projective perception. Fontana, for example, illustrates this when he explains the design of his trial:
The first question that I now endeavor to examine is whether smelling salts are a certain remedy against snake bite, that is, whether smelling salts ensure the survival of an animal that would have died without this remedy.… I have repeated my initial trials on this first question so often, that for several of my readers it will seem unnecessary. But I know what a preconceived opinion can do to a beloved hypothesis and to the reputation of a famous name. (Fontana 1787, p. 70)
The epistemological subjectivity problem that Daston and Galison (2007) describe was already so relevant in the 1780s, that 1) it had ex negativo decisively influenced the notion of a good observer (“He must be guided only by the love of truth and not by self-infatuation, prejudice, tradition, or by devotion to a system or school,” Reil 1817, p. 138: see above, pp. 5–6), and that 2) it influenced or should have influenced research practice.27 Although not the only source of error, projective subjectivity was nevertheless considered to be a fundamental source of error that counteracted rule-driven, verifiable, and reproducible (objective) observation and interpretation. And so subjectivity is also constitutive—alongside other aspects—of this concept of objectivity, even though here the desired objectivity was not as exclusively determined and defined as mechanical objectivity was. But with what justification can I speak of “subjectivity” and “objectivity,” given that these terms were not even used by early pharmacologists, who instead distinguished between “true” and “false” experience (Hildebrandt 1786, p. 40)? I do so because the scientific ideal they strove for manifests characteristics that were later described as objective, and because the opposite of that ideal corresponds with what, from about the middle of the nineteenth century onward, was called subjective. And so one finds in Meyers Konversationslexikon in the 1880s, that “objective” is that “which relates to an O.[bject], belongs to or accords with it, as opposed to the subject, as the perceiver, and the subjective, as that which belongs to the subject. An objective realization or representation accords with the true nature of the thing, a subjective one accords only with the perception that the observer has of it” (pp. 303–4).28 It is precisely in the sense of this dualism that prior to 1800 the “subjective” (the supposedly distorted perceptions of scientists) was counterposed to scientific values and aims (and the key criterion of grasping the phenomenon being studied as it really is): “Only the error one deems to be true obstructs scientific progress and leads us not to truth and nature, but to dreams and illusions” (Fontana 1787, S. 65). Sometimes fear of the subjective went so far that all scientific products of the mind were negatively connoted and juxtaposed with truth or reality, for example “to reason” and draw “rational conclusions,”29 “hypothesis,” “system,” and “theory.”30 Here too we can identify a close relationship between science or specific scientific subjects (scientific selves) and ethics and epistemology as described by Daston and Galison (2007), except that this code of “epistemic virtue” (p. 18) is not at all one of the paradigms they describe. Now one could argue that the pharmacological concept of objectivity manifests elements of “trained judgement” and “mechanical objectivity”; with a good deal of imagination, perhaps even a trace of “truth-to-nature,” if one puts much credence in Reil’s idea of a numerically derived “normal condition” of the probands (Reil 1817, p. 139; see footnote 17). In this regard, Savoia (2015) describes Sigmund Freud’s approach to hysteria as a mixture of three ideals distinguished by Daston and Galison (2007). This kind of description, which relies on comparative concepts, can enhance our understanding of the pharmacological concept of objectivity, but again points to the relativity of the Daston-Galison paradigms which are described as specific and distinct. If one finds their individual components at work in different mixtures, then the power of the paradigms as such is diminished. What’s more, the fact that in pharmacology prior to 1850 the terms “objectivity” and “subjectivity” were not used is explained by Daston and Galison themselves: “Sometime circa 1850 the modern sense of ‘objectivity’ had arrived in the major European languages, still paired with its ancestral opposite ‘subjectivity’. Both had turned 180 degrees in meaning” (p. 31).
If one compares the pharmacological concept of objectivity with say the mechanical one, it’s richness, among other things, stands out and is most likely attributable to its different research tasks and objects: compared to the production of images (Atlases), research on and treatment of living objects, that in turn are themselves subjects, appears to pose a series of divergent challenges. Thus it is hardly surprising that we find no identical concepts of objectivity. The eighteenth and nineteenth century pharmacological concept of objectivity calls into question the core propositions of Daston and Galison (2007), some fundamentally – the emergence of objectivity in the nineteenth century, as well as the contemporary ‘novelty’ of the epistemological problem of subjectivity—and others in their general validity—the sequential emergence of “epistemic virtues” (p. 18) as well as their existence as separate, distinct paradigms. The concept of pharmacological objectivity effectively demonstrates what some reviewers have already theoretically identified as troubling: the generalization of results derived (solely) on the basis of atlases (Ash 2008; Pettit 2009). Although using atlases as sources to investigate objectivity has some advantages (Daston and Galison 2008), the extent to which they can be considered representative is clearly limited. As the philosophical and history of science contributions to a collection of articles published by Padovani et al. (2015) suggest, further research is bound to reveal a more heterogeneous view of scientific objectivity and subjectivity than Daston and Galison (2007) depict. Furthermore, it would be interesting to learn more about the emergence of other ideals of objectivity and knowledge. Although Daston and Galison claim to answer the question of “how and why objectivity emerged as a new way of studying nature, and of being a scientist,” p. 17), many questions still remain unanswered about “the double reformation of self and sight” (p. 122) and its presumed offspring, scientific objectivity.
Assistant physician in the ophthalmology department at the general hospital in Vienna.
For (extended) documentation of this objectivity criteria, see Lindemann 2018 (Chapter 2.3).
On these first three criteria, see Erlenmeyer’s criticism of his colleagues Hermann Smidt and Carl Rank: “Rank and Smidt would necessarily have arrived at precisely the same results as Erlenmeyer, 1885, 1886 in their individually and jointly published articles, if they hadn’t placed the conclusions derived from these articles in incomprehensible opposition to the actually published evidence on which the articles were based” (Erlenmeyer 1887, p. 170).
On these last four criteria, see Erlenmeyer’s criticism of Wallé, a country doctor and advocate of cocaine: “Where and how these ‘most recent observations’ were carried out remains unsaid. There is not the slightest sign of scientific, objective observation; only subjective impressions, enthusiasm, and imagination” (Erlenmeyer 1887, pp. 448–49).
Reil, Johann C. 1799. Beitrag zu den Prinzipien für jede künftige Pharmakologie. Magazin zur Vervollkommnung der theoretischen und praktischen Heilkunde 3: 26–64. Reil’s publication will be cited henceforth using the reprinted edition of 1817. From 1788 Reil was professor of medicine and director of the Clinical Institute in Halle; in 1789 he also became a municipal physician.
From 1784 Nolde was professor of obstetrics in Rostock, and from 1797 also a district physician. The translations are the author’s own.
As of 1786, Georg Friedrich Hildebrandt was professor of anatomy at the Collegium medicum in Braunschweig.
“Yet it would again be going too far if one followed Mr. Hanemann’s recommendation of using only one single remedy at a time for an illness. For not every illness is so simple that we can flatter ourselves in believing that applying one remedy at a time will lead to success” (Nolde 1799, part 1, p. 68–9).
See Husemann (1874): bedside examination “can be of some use only if it focuses solely on phenomena that can be related to the remedy in question and that constantly arise” (p. 17).
“It’s undeniable that the more a remedy resulted in a favorable outcome, the greater the trust it deserves and the more justifiable it seems to attribute outcome to cause” (Schroff 1873, p. 50).
“As a rule, we must insist that success in one or a few cases allows us to draw no correct conclusions; only by examining hundreds and thousands of cases of the same illness can we make plausible assumptions about what effect the remedy had on a successful outcome. Thus, the statistical method proceeds directly from the empirical method, shares nearly all its defects, but precludes its optimism by using brutal numbers drawn from large sets of observations to scrupulously assess the empirical method’s favorable outcomes” (Nothnagel and Rossbach 1884, pp. 3–4).
“But as regards the human body, no equality is possible. Every individual differs from the others, and the same individual does not remain the same over time. And so a normal condition Normalzustand must be established, so that prior to the trials the proband’s own deviations can be precisely determined. This will ensure that differences in the results can be compensated for by the proband’s individual deviations from the assumed norm” (Reil 1817, p. 139). The rule requiring probands to be as similar as possible resembles current evidence-based principles of randomization and efforts to ensure that subjects in different groups are (on average) as similar as possible (e.g., Worrall 2002).
“The sex, age,… and way of life, the habits, idiosyncrasies …. The quality of food and drink enjoyed by the patient, also their quantity, the air he breaths,… the apartment, … his psychological mood (Seelenstimmung), the manner in which he is treated by his acquaintances, his various circumstances and a hundred other things. All of these factors may be no less overlooked by the attending physician, than the success that these individual effects or an ensemble of effects had on the state of the illness he is medicating” (Nolde 1799, part 1, pp. 83–4).
Professor for physiology and pathology of the nervous system at the University of Vienna and director of a private mental asylum in Ober-Döbling near Vienna. Alongside Erlenmeyer, Obersteiner was an authority in the field of morphine research.
“To guard as much as possible against such trickery on the part of his patients, and to preserve the certainty of his empirical science,… the physician must above all carefully study the character of his patients in this regard. One must examine … one’s patients from every angle …, from time to time shifting the conversation with him unobtrusively to other topics, paying attention to the patient’s regular behavior in his business affairs, his home, and in raising his children.… Always look him directly in the eye and interrupt him with a few unexpected questions …” (Nolde 1799, part 2, pp. 77–9).
“These paranoid delusions even persist for quite some time. And even though patients are masters at denying it, a psychiatrically trained eye will be able easily to recognize their continued existence” (Erlenmeyer 1887, p. 185).
“The history of scientific objectivity is surprisingly short. It first emerged in the mid-nineteenth century and in a matter of decades became established … as scientific norm” (Daston and Galison 2007, p. 27; see also p. 29).
For Daston and Galison (2007), “mechanical objectivity” is the first comprehensive concept of objectivity (pp. 197–98), as implied in this claim: “In this fashion, it is not difficult to tote up a long line of forerunners of objectivity—except that none of them operates with the concept in its entirety, to say nothing of the practices. The aim of a non-teleological history of scientific objectivity must be to show how all these elements came to be fused together (it is not self-evident, for example, what emotional detachment has to do with automatic data registration)” (Daston and Galison 2007, p. 29; emphasis added).
The authors dispute a linear sequence of the codes or paradigms, but their account nevertheless posits a de facto master narrative based on their claims about the emergence and predominance of the codes at certain times: “There was a science of truth-to-nature before there was one of objectivity; trained judgement was, in turn, a reaction to objectivity. But this history is one of innovation and proliferation rather than monarchic succession. The emergence of objectivity as a new epistemic virtue in the mid-nineteenth century did not abolish true-to-nature, any more than the turn to trained judgement in the early twentieth century eliminated objectivity” (Daston and Galison 2007, p. 18).
See Daston and Galison (2007), pp. 18, 197–98, 261, 311.
“Furthermore, as the patient uses the remedy in question, we must exactly, and with an impartial eye, blinded by no prejudice, observe everything that at once both affects him, and that is affected in him and that visibly manifests itself. In doing the former, we will be in a position to judge whether or not things other than the examined remedy contribute to the success; in doing the latter, we will come to know how the examined remedy manifests its effectiveness. Without observing both of these circumstances, the attempted trial is not worth the smudge of ink one uses to write it down and to contribute one’s supposed experience to the world” (Hildebrandt 1786, pp. 84–5).
The trial rules were only partially implemented. According to Tröhler (2003, p. 148, 144), calls appealing to Bacon and demanding controlled (clinical) trials in the second half of the eighteenth century were only slowly heeded. And the eighteenth century works cited here bear witness to the attempt to systematically expand orderly (true) experience in Bacon’s sense of a scientific norm. But Bacon’s reception was indirect, through the work of the physician Johann Georg Zimmermann (Hildebrandt 1786, p. 40; Nolde 1799, part 2, p. 92), who relies on Bacon to criticize the status quo of therapeutic research (Tröhler 2003, p. 148).
This quote is taken from the entry on “Objékt” in Meyers Konversationslexikon (1885–1890), vol. 12: Nathusius-Phlegmone, pp. 303–4.