Clinical trials and the origins of pharmaceutical fraud: Parke,Davis & Company,virtue epistemology,and the history of the fundamental antagonism

Introduction

In 1882 the American pharmaceutical manufacturer George Davis asked a young pharmacist and physician named Francis Stewart to write a medical article about a new remedy made from desiccated cow’s blood that Stewart had recently developed. Davis was the President of Parke, Davis & Company (Parke-Davis), which had been established in 1857 toward the end of the American Civil War and had since grown into a successful and reputable firm. Stewart had recently begun working with the firm, and Davis was interested in promoting the product to the medical community. Stewart wrote what he considered a scientific article. Davis, however, planned to use the article for promotional purposes and did not think the article was sufficiently laudatory. He then asked Stewart to revise it, and when Stewart refused Davis fired him. Stewart was deeply offended. Scientific articles cannot be written solely for “advertising purposes,” he noted in one angry letter during the exchange, and “must contain the truth, the whole truth, and nothing but the truth, to the extent that truth may be obtained.” Davis rehired Stewart after the two men reconciled, but tensions between them about the proper relationship between scientific and commercial practice continued until Davis was unexpectedly pushed out of the firm a decade later. ¹

From today’s perspective, this exchange is not particularly surprising. In recent years a large body of both scholarly and investigative work has documented troubling practices within the pharmaceutical industry, including the suppression of scientific evidence, illegal marketing, and ghostwriting. According to many critics, including the authors of this paper, such behaviors undermine the rational basis of scientific medicine and threaten the welfare of patients. At the same time, reformers in the medical community and other concerned parties have long sought to constrain the behavior of the industry in ways that align with their own views of ethical conduct and proper scientific practice. There is, of course, a significant amount of simplification in our description of this dynamic: the lines between industry and reformers are not completely clear, for example, nor is it true that troubling practices take place only in the pharmaceutical industry. It is also true, of course, that the pharmaceutical industry has long sought to advance scientific knowledge and that collaboration between the pharmaceutical industry and academic scientists, physicians, and others can and has resulted in important and useful new knowledge. Still, despite these complexities, unethical practices that violate contemporary scientific norms have become regrettably common within the pharmaceutical industry and among those who collaborate with it. It is also clear that such behavior has undermined the reliability of the clinical literature and harms patients. ²

In this paper we use Stewart and Davis to describe the origins of what the American Medical Association (AMA)’s Council on Pharmacy and Chemistry (CPC) in 1920 called “the fundamental antagonism” between “objectives that are largely commercial on the one hand and purely scientific on the other.” ³ Harry Marks and others have long noted the importance of the CPC in early twentieth century efforts to reform the American drug market along the lines of changing scientific norms and to suppress what they considerd fraudulent practices. ⁴ The CPC, like numerous other therapeutic reformers then and since, assumed that there is an important tension between scientific and commercial imperatives in the construction of pharmaceutical markets and, in doing so, operated according to an idealized vision of therapeutic rationality as they worked to change the world in which they lived. We share this assumption, and in this paper we argue that the tension that operates between scientific and commercial imperatives—at least as far as it relates to the behavior of both the pharmaceutical industry and therapeutic reformers in the United States—originated in the efforts of people such as Stewart and Davis to transform the drug innovation process along what they considered scientific and ethical lines. Epistemic norms and practices related to scientific knowledge about drugs changed rapidly in the decades immediately following the American Civil War as therapeutic reformers simultaneously embraced and promoted both laboratory science and increasingly systematized forms of clinical experimentation. This included reformers who worked in or with industry, such as Davis and Stewart, who tried to simultaneously improve pharmaceutical science and advance their own commercial interests. Yet the new framework these reformers helped usher in also fundamentally altered the set of epistemic virtues that reformers considered necessary to conduct reliable scientific inquiry regarding drugs. In doing so, this framework made possible not only new forms of knowledge. It also made possible new forms of fraud.

We draw the phrase “epistemic virtue” from the philosophical field of virtue epistemology, which addresses qualities which facilitate (virtues) or impede (vices) successful inquiry. ⁵ In this paper, we argue that epistemic virtues and vices are not fixed sets of traits but are instead historically contingent and reflect considerations that are often ignored by epistemologists. In this case we find that changes in epistemic virtue were motivated in part by the economic imperatives of drug discovery. Reputable manufacturers such as George Davis and those who collaborated with them, such as Stewart, were active participants in refashioning the epistemic virtues that applied to medical research. They did so, we argue, both by legitimizing the role of drug patenting in scientific practice and by contributing to the broader turn toward laboratory and systematized clinical experiment within medical science. Yet even as they did so, they also discovered that new products could be introduced and marketed in ways that did not fully conform to the changing normative regime that they themselves helped introduce. Indeed, manufacturers quickly discovered that violating the newly emergent norms governing scientific practice offered opportunities to expand markets and accumulate profit. At this moment, we argue, what we refer to as the fundamental antagonism—at least as it relates to the pharmaceutical industry in the United States—was born.

We thus describe a process in which both scientific practice and ethical reasoning were transformed. We also briefly describe how this new epistemic and normative regime was institutionalized toward the goal of rationalizing the market in new drugs. In doing so, we describe not only one possible origin point for many (although not all) of the fraudulent practices that now take place in the pharmaceutical industry; we also try to describe why we conceptualize such practices to be fraudulent in the first place. We thus seek a philosophical ground for normative claims about pharmaceutical markets that takes seriously the historicization of the claims themselves. For those of us who are concerned to preserve the place of normative claims within our descriptions of scientific practice it is difficult, perhaps impossible, to look at the recent history of the pharmaceutical industry without characterizing at least some industry behavior as corrupt. Yet in doing so, we find ourselves operating within the very framework that this paper seeks to explicate, one in which both laboratory science and rigorously designed clinical drug trials provide us with presumably valid scientific knowledge, and commercial forces, if not properly restrained, threaten to corrupt scientific practice. Our goal is thus to describe the origins of one specific type of fraud, and, in doing so, to provide a historical ground for our own critical work.

Drug manufacturing and early American medical epistemology

In order to understand the transformations that George Davis and Francis Stewart were a part of, it is first necessary to describe what they were reacting against. The second section of this paper therefore provides a brief overview of the normative and epistemic framework governing orthodox medical science in the early nineteenth century United States. In other work, one of us (Gabriel) has argued that American orthodox physicians during this period understood the relationship between scientific and commercial practices in ways that are both familiar and strange to us today. Medical investigation and discovery were rhetorically, if not always practically, positioned as being dedicated to the advancement of science and the betterment of humanity rather than toward the narrow interests of the investigator. As members of what one physician in 1825 called “the republic of medical science,” orthodox physicians assumed that scientific knowledge about healing goods should circulate freely among their peers to the benefit of all; the free circulation of knowledge was therefore rhetorically juxtaposed to the apparently selfish and unscientific practices of the quack. ⁶ Although the experimental testing of drugs and other cures has a long history that dates back to at least the fourteenth century, Elaine Leong and Alisha Rankin note that during the early modern period there was little consensus about what counted as proof or evidence. ⁷ Furthermore, both medicine and pharmacy were understood as experientially based forms of practice, and, as Leong and Rankin suggest, even the most theoretical work in these areas was “developed from empirical and sensory modes of knowing like taste and smell.” ⁸ Orthodox physicians in the early United States also confronted a medical market filled with competing sects and practices which they discounted as both irrational and dangerous. All this meant that from the perspective of early American orthodox science, the validity of knowledge claims produced through drug testing was assumed to be intimately linked to the reputation and character of the individual investigator. The claims made by reputable physicians of supposedly good character were accepted within the broader orthodox medical community as both trustworthy and true in a way that the claims of disreputable physicians and so-called quacks simply were not, even if the two engaged in very similar practices. ⁹

Orthodox physicians in the early nineteenth century United States thus embraced a series of epistemic virtues that are familiar to us today, including honesty and humility. Yet their tendency to juxtapose science and self-interest also meant that they embraced an antimonopolistic selflessness as a core epistemic virtue and, in doing so, rejected both drug patenting and the use of secret ingredients as unethical forms of quackery. The early American drug market was dominated by the sale of raw botanicals and manufactured goods made from known ingredients, such as “oil of wormseed” and “essence of peppermint.” ¹⁰ By the 1830s, however, a robust trade had also developed in so-called “patent medicines” that were either made from secret ingredients or, less commonly, actually patented. Although popular, these products deeply offended the sensibilities of orthodox physicians because they seemed to prioritize the self-interest of the manufacturer over the advancement of medical science and the good of the public. Reformers in the medical community assumed that both patenting and secrecy interfered with the ability of physicians to freely share information about the products in question and to use them to help their patients as they saw fit. From this perspective, patent medicines were a dangerous form of medical irrationalism that undermined medical science and threatened the public welfare. Proper character, and thus the ability to make trustworthy scientific claims, therefore depended on their rejection. A prohibition on patenting pharmaceuticals and prescribing patented products, or remedies made with secret ingredients, was thus incorporated into early codes of medical ethics in the United States, including the AMA’s first Code of Ethics in 1847, which noted that it is “derogatory to professional character. . .for a physician to hold a patent for any surgical instrument, or medicine.” ¹¹

Orthodox physicians thus understood the monopolization of therapeutic discoveries as a form of quackery and juxtaposed it to the “professional character” needed to make trustworthy scientific claims. ¹² More generally, they considered the discovery and introduction of new remedies first and foremost to be a scientific, rather than a commercial, activity: the discovery of new medicinal plants, the isolation and development of new chemicals and formulas—these and other improvements to the therapeutic armature were understood as one part of the supposedly cooperative and benevolent scientific process rather than the products of commercial self-interest. Physicians were thus expected to publish the results of their studies for the benefit of all, rather than restrict knowledge about their discoveries for their pecuniary benefit. At the same time, marketing new remedies before the medical community had reached a consensus on their therapeutic value was frowned upon and at times equated with outright quackery. It was considered both unethical and unscientific to promote and sell a product before the medical community had decided upon its usefulness, a reversal of the proper order of things in which commercial interest was placed ahead of the common good. From the perspective of orthodox physicians, the problem with quackery was therefore not just its dangerous nature, or even the way it monopolized scientific knowledge. The fundamental problem was the manner in which narrow self-interest seemed to trump the cooperative nature of a supposedly benevolent science. ¹³

As a result, even if a monopolized drug might be useful in certain cases it was still considered both unethical and unscientific to recommend it to a patient. We are accustomed to thinking that a drug’s utility can be evaluated through rigorously designed clinical trials and that, in general, such utility can be assumed to apply across a range of clinical cases in roughly the same manner. Yet in the early decades of the nineteenth century, orthodox physicians generally did not think of drugs as operating in this way. Instead, remedies were assumed to have general effects that might help restore balance to the bodily system, but the balance of the system itself depended on many factors and—outside of a few drugs known as specifics—remedies were typically not assumed to have the same effect across different cases in the way that we hope they do today. ¹⁴ Case reports of clinical experiments thus provided a general sense of how a remedy might be useful, but they did not imply that the remedy had the potential to be helpful for any particular individual case; the job of the physician was to integrate knowledge about the general effects of the drug in question with knowledge about the specific characteristics and problems of the patient he sought to treat. Patented drugs contradicted this framework because—in addition to promoting the narrow self-interest of the patent holder—the act of monopolization was assumed to interfere with the ability of physicians to use the drug as they saw fit for the benefit of their individual patients. Medicines made with secret ingredients, for their part, could not be tailored to individual cases due to the simple fact that physicians did not know what they were actually made of. At the same time, patent medicine manufacturers typically marketed their products as useful in all cases of a given problem, and often as useful for all cases of multiple problems. From the perspective of orthodox medicine this was deeply irrational since it contradicted their understanding of each case as fundamentally unique. The trouble with quack remedies thus lay both in their monopolized status and in the claims of generalized utility made about them. Indeed, these were understood as interrelated problems since the ability of manufacturers to claim universal action was typically based on their supposed possession of secret knowledge gained from relatives, Indians, or other sources that they monopolized in order to promote their own self-interest. ¹⁵

Still, there was an important ambiguity in all of this. The small handful of drugs known as “specifics” that appeared to have the same effects across all cases pointed to the emergence of a new scientific framework in which drug action was conceptualized as tightly linked to the nature of the substance in question. From the perspective of orthodox physicians, patent medicines made from such substances might actually be useful for everyone with a particular type of problem, as their manufacturers claimed. The possibility of universal action, in turn, opened the possibility that—despite their monopolized nature—these products were not fundamentally quackish. During the antebellum period, however, most orthodox physicians rejected this idea and described monopolized products as quackish even if they were also recognized as being useful in a general sense. Indeed, in such cases the orthodox medical community tended to describe the manufacturers in question not as making useful therapeutic advances but instead as selfishly and unethically expropriating medical science for their own personal gain. In 1832, for example, a physician named John Sappington introduced a product he called “Sappington anti-fever pills” in Missouri. Sappington had been trained as a physician by his father, a situation which was not unusual at the time, and had maintained a busy medical practice before introducing his pills. Sappington’s pills became extraordinarily popular and many of his customers testified to their effectiveness. Yet despite their apparent usefulness, the orthodox medical community in Missouri denounced Sappington as a quack and refused him entry to the local medical society. Perhaps in response, in 1844 Sappington revealed that his pills were made from sulfate of quinine—a specific for the treatment of fever. Yet this did little to rehabilitate either Sappington’s reputation or the reputation of his pills within the orthodox medical community. “The efficacy of Dr. Sappington’s pills, which in the general I do not deny, depends on this, that the sulphate of quinine, as it is well known, will seldom fail to arrest the paroxysms of intermittent fever,” noted one critic. “This was not a discovery by Dr. Sappington. All that he can claim is, that he has made the discoveries of others instrumental in his own wealth.” ¹⁶

William Morton’s 1846 patent on ether serves as another important example of these dynamics. Morton’s discovery of general anesthesia was quickly recognized as a scientific triumph, but his willingness to monopolize his discovery was considered by most observers to be a gross violation of medical ethics. As a result, some physicians initially resisted the use of ether as a form of quackery, dismissing it as a patent medicine that was, by definition, outside the domain of medical science. Yet the dramatic and obvious therapeutic utility of the discovery made the equation between monopolization and quackery difficult to maintain. In response, physicians quickly decided that Morton was not really responsible for the discovery at all, that his status as a dentist meant that he could hardly be expected to follow the lofty ideals of medicine, and that the discovery had in fact grown out of the scientific work of numerous earlier physicians. Doing so characterized Morton as a threat to the scientific enterprise due to his selfish effort to patent the procedure, but it also worked to redeem general anesthesia itself from the charge of quackery. The procedure might have been patented by an unethical dentist, this line of argument went, but its discovery grew out of the broader efforts of the scientific community and it remained a tremendously important discovery nonetheless. The re-articulation of Morton’s patent as an unfortunate example of commercialism, but one that did not disqualify the use of ether itself, thus worked to maintain the broader conceptual framework of antebellum medical science in the face of an obviously useful, yet monopolized, discovery. ¹⁷

Taken together, both Sappington’s pills and the ether case illustrate how the equation between monopoly and quackery—and the conflation between patented medicines and patent medicines—operated as a powerful part of orthodox medical thought over the course of the antebellum period. In both cases, efforts to monopolize therapeutic innovation toward commercial self-interest were denounced as quackery and juxtaposed to the epistemic virtues of orthodox medical science—and, in particular, to the virtue of selflessness. This was a powerful critique, and one that had a tremendous effect on the shape of the antebellum drug market. Although many manufacturers continued to use secrecy to protect their commercial interests, and a small number continued to patent their goods, in the two decades before the American Civil War some manufacturers began to self-consciously follow the scientific norms of the medical community. These so-called “ethical” manufacturers refrained from monopolizing their products through patenting or the use of secret ingredients, and they only marketed their goods to physicians and pharmacists. Most importantly for our purposes, in general they refrained from commercially introducing new products in a way that violated the norms of the elite orthodox medical community; to do so would be to risk the label of quackery and therefore the potential collapse of their markets. By the 1850s so-called ethical manufacturers had established themselves as a distinct wing of the drug manufacturing industry. ¹⁸

These companies were decidedly limited in their ability to commercially introduce new products. Manufacturers in the so-called ethical wing of the industry might introduce new plant remedies discovered in a distant lands, for example, but in general they were careful to publish what they knew about such remedies before selling them due to the assumption within the medical community that scientific discoveries should not be monopolized for private gain. Scientific innovation in this sector of the industry, in other words, generally preceded commercial introduction and, due to the normative prohibition on monopolization, so-called ethical manufacturers all tended to produce essentially the same products—after all, if a company discovered a new remedy and published what they knew about it, other manufacturers were soon able to enter the market in the product as well. There was some ambiguity in all this due to the fact that these manufacturers competed with one another around the quality of their products, and at times changes in quality could be dramatic enough to change the utility of the good in question. Yet manufacturers were careful to describe such products as simply high-quality versions of familiar goods rather than as fundamentally new products, in order to avoid the charge of prioritizing their own narrow interest over the interests of the scientific community as a whole. Innovation as a commercial strategy in the so-called ethical wing of the drug manufacturing industry was thus limited by the ethical framework through which its markets were constituted. ¹⁹

Clinical drug trials and the reconceptualization of markets

The situation changed dramatically in the decades immediately following the American Civil War as manufacturers in the so-called ethical wing of the drug industry established themselves as an important part of the new industrial landscape. In doing so, they laid the foundation for the eventual emergence of the modern pharmaceutical industry that we know today. Yet they also faced an important problem: the normative prohibition on monopolizing scientific discoveries and introducing new remedies in a manner that benefitted the interests of a single firm meant that so-called ethical companies were decidedly limited in their ability to harness scientific and technological innovation without risking the charge of quackery—a charge that could seriously harm a company’s reputation, and thus its ability to market its products to the medical community. Indeed, in the two decades immediately following the American Civil War it made little sense for most so-called ethical manufacturers to invest significant resources in developing new products, since they were expected to divulge information about whatever discoveries they might make and were unable to monopolize those discoveries for their own ends. Scientific innovation did take place in this wing of the industry during this period, but in general it was not commercialized in the way that we would expect today. Instead, manufacturers—many of whom were themselves trained as either physicians or pharmacists—understood their scientific efforts as contributions to the general storehouse of medical and pharmaceutical knowledge. The results of such efforts were circulated within the medical and pharmacy communities, and thereby worked to enhance the overall reputation of the firms in question and, perhaps, indirectly increase the sales of their products. In general, however, such efforts were not directly linked to the commercial introduction of new goods. Explicitly promoting new products risked the charge of quackery. ²⁰

Yet manufacturers in this wing of the pharmaceutical industry also faced intense competition from a range of manufacturers that did not always conform to these norms. This included the patent medicine industry, of course, but also chemical manufacturers that sometimes discovered therapeutically useful products, small companies that manufactured specialty goods, and, most importantly, German and other European manufacturers that had no real allegiance to the ethical norms of the medical community in the United States. These companies innovated and commercially introduced new products as they saw fit, and they often used both patents and secrecy to protect their markets in ways that orthodox physicians found deeply troubling. Their willingness to do so grew out of a variety of complex factors, including the fact that their markets did not depend on the attitudes of orthodox physicians and, in some cases, that the obvious utility of their products meant that they could not easily be dismissed as quackish goods. In the 1880s, for example, German manufacturers introduced a wave of synthetic chemical antipyretics that proved both surprisingly effective and tremendously popular. ²¹ Manufacturers in the so-called ethical wing of the domestic pharmaceutical industry thus found themselves facing a highly competitive market yet unable to innovate and monopolize new products in the same manner as their competitors.

They responded by embracing—and helping to create—a new scientific framework. As historians of medicine have long noted, in the decades following the American Civil War orthodox physicians turned toward what historian John Harley Warner has called “physiological therapeutics.” Under this new framework, physicians increasingly assumed that therapeutic knowledge should be based on laboratory science, and that the laboratory basis of scientific knowledge meant that treatments acted the same across individual cases. In this sense, all drugs were conceptualized as “specifics” in the way that only a handful of drugs had been in the antebellum period. At the same time, this change in perspective underlay gradual efforts to systematize drug trials in order to generate generalized knowledge about the actions of specific drugs in specific diseases. Over the course of the late nineteenth century physicians thus worked to systematize clinical trials and struggled with the question of how to link laboratory and clinical science together in a coherent framework. They also debated the role of clinical experience in the new therapeutic framework and sought to place treatments that had long been in use, such as bloodletting, on a sound scientific basis. ²²

What is important here for our purposes is that both the physiological turn and the effort to establish generalized knowledge about drug action through increasingly systematized clinical trials pointed to the detachment of medical science from the traditional normative framework of orthodox medicine. As Warner has argued, the physiological revolution was intertwined with a broad revolt against orthodox medical ethics and, in the terms we are using in this paper, the normative framework that linked physician character with epistemic warrant. ²³ Proponents of physiological therapeutics believed that since medical science was properly grounded in the laboratory, the link between medical science and a physician’s character embedded within formal codes of medical ethics—and by extension the broader scientific framework these codes tried to articulate—was both antiquated and un-scientific. Although the relationship between laboratory findings and clinical utility was not always clear, the embrace of both meant that, increasingly, scientific knowledge about drugs was understood as having little to do with seemingly antiquated norms of conduct. Proponents of both physiological therapeutics and systematized clinical trials thus worked toward establishing new epistemic practices and a related set of epistemic virtues that were associated with the ability to conduct science in what they considered a modern sense, including attention to detail, careful observation, and the disregard of one’s own personal preferences in the outcome of an experiment. Indeed, according to this new perspective, the personal characteristics of the investigator had no necessary relationship to the validity of the claims being advanced at all. As we suggest below, this included a reinterpretation of drug patenting from an epistemic vice to being epistemically neutral or, perhaps, even a virtue.

George Davis and Francis Stewart played an important role in these changes. Under Davis’ leadership, during the 1870s and 1880s Parke-Davis pushed the boundaries of what was considered acceptable behavior for an ethical firm. In the late 1870s, for example, the company introduced a variety of new botanicals from California to the drug market, prompting accusations of quackery by physicians concerned that the company was trying to undermine the scientific process for its own gain. In 1881 Davis also established a small laboratory at the company and hired chemist Albert Lyons from the University of Michigan. Lyons developed a method of standardizing the company’s fluid extracts, which the company then marketed under the term “normal liquids.” It was an important breakthrough that brought both therapeutic and commercial rewards. The company was careful to describe normal liquids as an improvement on fluid extracts rather than as a new therapeutic substance in their own right, thereby avoiding the charge of quackery that had followed their introduction of the California remedies. In 1889, however, Davis began an effort to introduce standardized fluid extracts—under the term “normal liquids”—into the United States pharmacopoeia. From his perspective, this would advance medical science in an important direction and, simultaneously, benefit the interests of his firm. Critics, however, saw the effort as an unethical intrusion of private interest into the scientific process and denounced it as such. As one critic put it, “All the prevalent cant about ‘standardization’ looks suspiciously like an attempt to boom one firm of manufacturing druggists in the Pharmacopoeia.” ²⁴

An important part of this effort was Stewart’s rejection of the prohibition on medical patenting and, with it, his argument that clinical utility mattered more than monopoly status when evaluating a remedy. Stewart was a remarkably prolific writer, and over the course of his career produced a tremendous outpouring of articles arguing for both the ethical legitimacy of patent rights in drug manufacturing and the value of pharmaceutical manufacturers commercially introducing new products. Grounded in an emergent epistemological framework that prioritized knowledge generated in the laboratory and the clinical trial, Stewart and other reformers argued that clinical utility was what really mattered: scientific evidence that demonstrated the utility of a good justified both the commercial introduction of that product and its adoption into widespread medical practice. At the same time, patenting did not, by definition, define a remedy as unscientific—whether or not a product was protected by patent, in other words, was irrelevant to the question of its utility, and therefore had little to say about its ethical status. The embrace of clinical utlity as a justificatory mechanism for the commercial introduction of new goods, and with it the rejection of monopoly status as the defining boundary between legitimate and illegitimate remedies, thus worked to reconfigure the relationship between markets, epistemic virtue, and scientific practice. Under this new formulation, patenting was epistemically neutral—or, perhaps, even an epistemic virtue since it supposedly allowed manufacturers to invest resources in developing new products and thereby advance medical science. As Stewart put it in 1882, the purpose of patents “is to promote progress in science and the useful arts; and this end is secured by encouraging authors and inventors to write and discover, protecting capital invested in the product of their brains, creating a valuable industry, and making this industry a great knowledge-producing power engaged in original research and publishing the results for the benefit of humanity and the cause of truth in the world.” ²⁵

Toward this end, in 1880 Stewart developed what he referred to as “the hospital plan,” in which Parke-Davis sent samples of its new products to physicians working in hospitals and other locations with an invitation to investigate them. The company promised to publish reports of the results, “for good or ill,” in a series of so-called “working bulletins.” Working bulletins would also contain the results of laboratory investigations and other information about the treatment in question in an effort to present the current state of knowledge about the remedy in a single document. Davis embraced the plan wholeheartedly, and the company issued its first working bulletin on Berberis aquifolium in 1880. The company issued at least forty working bulletins over the next decade, with the program continuing until at least World War I. ²⁶ Certainly, this was not the first nor the only effort to organize clinical experimentation toward the goal of creating generalized knowledge; as Harry Marks has noted, the working bulletin system was part of a broader and haphazard effort in late-nineteenth-century American medicine to organize the experiences of general practitioners into collective investigations. ²⁷ What was groundbreaking about the plan, however, was the scope of the project, the wide variety of substances included, and the fact that it was an effort to organize both laboratory investigation and clinical investigations conducted at multiple sites into systematic overviews of new remedies. Perhaps most importantly, it did so in a way that clearly combined scientific and commercial motives without appearing to undermine the lofty goals of science. As Stewart later described it, the goal of the hospital plan was to “harmonize the interests of science and commerce, so that one may aid the other without jeopardizing the interests of either.” ²⁸

For Stewart, the hospital plan thus served as a means to reconcile scientific and commercial practices without subordinating the former to the latter. The company’s commitment to publish the results of investigations into its products “for good or ill” also suggests the manner in which the plan worked to transform the epistemic virtues associated with drug discovery. For Stewart, the most pressing aspect of this process was the need to replace what he considered an archaic attitude toward patenting with a focus on establishing the utility of goods through collective investigation. Patenting was thus rendered epistemically neutral, or perhaps turned into an epistemic virtue due to the fact that it—supposedly—allowed manufacturers to invest resources into the process of drug discovery. At the same time, the hospital plan promoted a set of epistemic virtues that, supposedly, distinguished the production of evidence from the interests of any particular actor. Foremost among these was the partial detachment of epistemic warrant for claims about drugs from the character or standing of individual investigators through the prioritization of collective action. According to Stewart’s plan, Parke-Davis would not publish the results of clinical investigations until “at least twenty-five cases have been treated in different hospitals to constitute a fair test for each drug.” A “sufficient” number of cases, in other words, was necessary in order to give a drug a “fair test,” and neither the interests of any particular investigator nor the interests of the firm would determine the results. At the same time, “competent physicians and surgeons” would conduct the tests according to a set of epistemic virtues that were based on the goal of establishing equivalency of test across investigatory sites, including virtues based on both the tamping down of one’s own personal predilections and the establishment of standard doses, shared vocabulary, and other mechanisms of standardization. The fact that competency was positioned as an epistemic virtue, of course, suggests that only some attributes of an investigator’s personal self were to be tamped down; it also suggests that reputation remained an important consideration in this process. Despite this, Stewart’s plan was an important contribution to the emergence of a new set of epistemic virtues based on the effort to collectively produce evidence. As the company’s 1881 announcement of the working bulletin plan to the medical community put it, “science proposes to exhibit what is actually known or may be learned by exact observation, precise definition, fixed terminology, classified arrangement, and rational explanation.” ²⁹

As the controversy over Davis’ efforts to introduce “normal liquids” suggests, this was a highly contested process. Critics such as the one quoted above who denounced “all the prevalent cant about ‘standardization’” were concerned not only about the commercial intentions of the firm but more generally about what they saw as the changing norms underpinning medical science that facilitated these intentions. Yet neither Davis nor Stewart were cynical actors here. It is clear from their correspondence that both men cared deeply about the science of therapeutics—Stewart, for example, was a vocal critic of what he considered unethical practices among drug manufacturers and he regularly denounced the patent medicine industry for supposedly subverting medical science to their own narrow interests. ³⁰ Stewart and Davis both understood themselves as advancing the cause of medical science even as they worked to advance their own commercial interests. Both men recognized that the hospital plan enhanced the company’s reputation within the medical community and worked to forestall charges of quackery that might otherwise be leveled against them for commercially introducing new products. They also recognized that informing the medical community about positive experimental results had the power to increase sales of their products and therefore that writing scientific articles had commercial implications. The hospital plan should be understood in this context: it was an effort to both scientifically establish the therapeutic utility of new remedies according to newly emergent epistemic norms and to simultaneously promote those remedies to the medical community. For both Stewart and Davis, the two were deeply intertwined projects. ³¹

Yet in hindsight, we can also see that the new linkage of scientific and commercial forces in the so-called ethical wing of the industry had an unexpected result. The possibility of commercially introducing new products, and the value of scientific information about them in promoting their sale, created a new tension between scientific and commercial practices for manufacturers such as Parke-Davis that they previously did not face. The linkage between scientific and commercial practices translated into a new pressure for scientific claims to conform to commercial needs and, at times, for commercial imperatives to violate the new epistemic virtues. Here we see the origins of the dispute between Davis and Stewart mentioned at the beginning of this paper: Davis initially asked Stewart to write the article for promotional purposes, but he also assumed that doing so was a reasonable part of the broader scientific process of informing the medical community about the product. When Stewart found himself unable to write a significantly laudatory article and stay within the bounds of what he considered “the truth, the whole truth, and nothing but the truth”—in other words, to write a sufficiently laudatory article while still tamping down the commercial interests of the firm for which he worked—the commercial imperatives of the firm came into conflict with the newly emergent epistemic virtues of collective investigation, both by requiring Stewart to prioritize the firm’s interests through the writing of what he considered untrue statements and by threatening to make his reputation relevant to the evaluation of scientific claims once again—in this case, in a negative sense. The anger in the correspondence between Davis and Stewart over the issue points to the significance of the stakes involved.

Parke-Davis’ efforts to promote a new anesthetic named “chloretone” is another example of this tension. Chloretone (now sold under the generic name chlorobutanol) is a mild anesthetic with sedative and hypnotic properties made by combining chloral hydrate and acetone. First produced in 1881 by the German chemist Conrad Willgerodt, the first animal tests on the drug were conducted in 1894 by John Jacob Abel, with a handful of clinical tests following shortly thereafter by a physician named Arthur Oppenheimer. The drug held a great deal of commercial promise—E.M. Houghton, the director of the company’s scientific laboratory, believed that it could be used not only as an anesthetic but also as a general pain reliever and in the treatment of insomnia, epilepsy, and other conditions. By late 1899 the company had distributed more than 30,000 tablets for clinical testing to physicians; in the same year, Houghton published an article in the Journal of the American Medical Association describing the new drug and its “encouraging” potential. ³² The company also informed its salespeople to “push” the drug and suggested that it could be used with “absolutely no danger” and “apparently no ill after effects.” ³³ This was not quite correct. Abel, in fact, had grown concerned about possible toxic reactions to the drug following Oppenheimer’s initial clinical tests due to the fact that one patient had developed “alarming symptoms” after its use. ³⁴ The company was also concerned that sustained use of the drug might “produce a fatty degeneration of the various organs,” and by 1899 had begun animal testing to determine whether or not this was the case. ³⁵ However, Parke-Davis does not appear to have disclosed these concerns to physicians as they distributed it for testing or in the company’s detailing work. Instead, the company emphasized the safety of the drug and promoted it as “the ideal hypnotic.” ³⁶ They also suggested that physicians continue to experiment with it in the treatment of a wide variety of clinical problems, including conditions such as epilepsy that implied long-term use. ³⁷ These investigations led to the publication of a large number of laudatory articles about chloretone in the medical press. ³⁸ At the same time, however, a small but growing number of critics raised concerns about the toxicity of the drug; as early as 1901, for example, one physician described chloretone as “a dangerous hypnotic” and noted the “somewhat startling” conclusion that it appeared to be two and a half times as toxic as chloral hydrate. ³⁹ Indeed, concerns about toxicity appear to have limited the commercial success of the drug; as Joseph Remington and Horatio Wood put it in 1918, chloretone “failed to achieve the confidence of the profession.” ⁴⁰

The development and promotion of chloretone illustrates how scientific and commercial practices were increasingly intertwined in the construction of pharmaceutical markets in the decades immediately before World War I. Scientific knowledge about new drugs was now to be established through both laboratory investigation and clinical experiment. At the same time, the antimonopoly position that had formerly served as a core epistemic virtue softened considerably and the legitimacy of new products was increasingly understood as hinging on their utility—as demonstrated through laboratory and clinical studies—rather than on their monopoly status. By the first decade of the twentieth century the medical community had largely come to terms with pharmaceutical patenting and increasingly accepted the idea that—as Francis Stewart argued—patenting allowed drug manufacturers to invest resources in the development of new products. Not surprisingly, ethical manufacturers cautiously turned toward patenting their products around this time. Parke-Davis’ 1903 patent on adrenalin is the most notable example. Adrenalin, despite its patented status, was widely considered a therapeutic breakthrough and earned the company a tremendous amount of money. ⁴¹ Its embrace by the medical community suggests the extent to which physiological therapeutics, clinical trials, and the pursuit of profit through drug innovation had begun to replace the older scientific framework in which a commitment to an anti-monopoly position was a core epistemic virtue. At the same time, however, the simultaneous willingness of the company to promote chloretone in ways that appear to have downplayed concerns about its dangers also suggest the manner in which the new linkage between the collective production of scientific evidence and commercial forces could lead companies to subvert the epistemic virtues of the former in the interests of the latter. The newly emergent framework thus made possible a new form of epistemic vice, one in which laboratory and systematized clinical data produced through collective investigation is distorted or ignored in pursuit of commercial success.

Clinical drug trials, regulatory objectivity, and fraud

We now briefly discuss the regulatory environment that governed pharmaceutical markets in the United States in the early twentieth century. By the latter decades of the nineteenth century a wide variety of reformers had begun efforts to rationalize what they considered a dangerous and predatory drug market. These included state and federal authorities that worked to enforce a wave of new laws intended to improve the quality of products on the market, prevent deceptive advertising, and suppress recreational and other forms of use that violated the sensibilities of the medical profession. These efforts culminated in the passage of the 1902 Biologics Control Act, the 1906 Pure Food and Drug Act, and the 1914 Harrison Narcotics Tax Act, each of which established significant regulations over the pharmaceutical market, albeit in complex and not always consistent ways. A variety of private and quasi-private organizations also sought to improve the market, including the AMA’s CPC. ⁴²

For our purposes, the key dynamic in the emergence of this new regulatory regime was the embrace of scientific claims made through both laboratory science and clinical experiment as a justification for regulatory enforcement. Regulatory and epistemological practice overlapped with one another in significant ways, whether in government laboratories, expert reports that were written and read by reformers, court trials in which scientific findings were debated, or other venues. These practices were themselves increasingly systematized along both formal and informal lines and required the participation of multiple parties in different institutional locations; the emergent regulatory regime that sought to rationalize the manufacture, sale, and clinical use of pharmaceuticals was thus an early example of what Cambrosio and colleagues have called “regulatory objectivity” which, as they note, is “based on the systematic recourse to the collective production of evidence” in which protocols and conventions are developed that take place outside the narrow domain of the clinical encounter. ⁴³ This process hinged on the idea that both laboratory and clinical science created forms of knowledge that were not dependent on the character or preferences of the investigator. Methodological practices that sought to systematize clinical experiment and control for the “personal equation” thus sought to promote a rational and benevolent therapeutic market. ⁴⁴ These practices were then institutionalized across the drug-development landscape in a variety of ways, including through adoption by drug manufacturers and clinical researchers in academic settings. They also increasingly served as the basis for evaluating claims made by manufacturers about their products by governmental bodies and other actors intent on reforming the drug market.

Stewart’s hospital plan was an important part of this new system. Stewart stopped working with Parke-Davis soon after Davis left the company in the late 1880s, and over the next several decades Stewart worked for a variety of other manufacturers, including Frederick Stearns & Company, Merck & Company, and H.K. Mulford & Company. In each of these, he established a working bulletin system similar to the one he had established at Parke-Davis. ⁴⁵ Given that these were among the largest and most successful ethical manufacturers in the United States at the time, his influence on the movement to systematize clinical research and organize it toward the collective production of evidence was considerable. Central to his vision was the idea that the collective production of evidence would act to eliminate the interests or biases of individual investigators. As one observer described his work in 1912,

“Dr. Stewart devised the working bulletin system more than thirty years ago. This collects evidence for and against the remedy, presents it to those best equipped for studying it, pharmacologically and therapeutically, adding new evidence as it comes in. The greater the number of competent observers, the more nearly the personal equation is eliminated.” ⁴⁶

Although Stewart was certainly not alone in this idea, the timing and broad influence of the hospital plan suggests that he was an early and influential proponent of an idea that underlay later efforts to control for investigator bias in clinical drug trials.

Yet Stewart also recognized the need to have some sort of independent authority to validate the claims made by manufacturers about their products to ensure that they were warranted by this process of collective investigation. Over the course of his career Stewart advocated for the establishment of what he sometimes referred to as a “board of control” that would evaluate claims made by manufacturers in order to ensure their trustworthiness. For Stewart, as for many other reformers at the time, the rationalization of the drug market depended in part on the ability of manufacturers to successfully develop and promote new products based on the collective production of evidence, yet it also required some type of arbitrating authority to ensure that manufacturers did not unfairly distort scientific evidence toward their own commercial ends. Even as the character of investigators was replaced as a ground for epistemic warrant, in other words, increasingly systematized clinical trials became an important means through which the commercial interests of drug manufacturers were advanced. Independent evaluation of the evidence produced through these trials, however, also became necessary to ensure that it represented, as Stewart had earlier put it, “the truth, the whole truth, and nothing but the truth.” ⁴⁷

In 1901 Stewart published an article in The Journal of the American Medical Association (JMA) in which he argued for the need for an independent body of scientific experts to evaluate new drugs on the market. This would serve two functions: first, to “collect knowledge of materia medica products . . . and publish it for the benefit of science;” and second to “aid the manufacturers . . . in the introduction of their brands to commerce by advocating that the medical profession in prescribing shall specify only those brands which comply with scientific and professional requirements.” Stewart’s proposed body, in other words, was intended to distinguish between trustworthy and untrustworthy manufacturers and to promote the interests of the former and suppress those of the latter. George H. Simmons, the editor of the JMA, found his arguments convincing, and after several years of effort established the CPC in 1905. ⁴⁸ The purpose of the CPC was to evaluate newly introduced drugs and determine whether they should be considered legitimate products that might be properly used in medical care; the CPC regularly published reports of its findings, and products that were determined to be quack remedies were prohibited from being advertised in the JAMA and other journals published by the AMA. The CPC was thus one important part of the broader effort to establish, institutionalize, and enforce changing epistemic norms. In doing so, it helped establish regulatory objectivity through the collective production of evidence.

The AMA’s Committee on Therapeutics was particularly important in this regard. As one of us (Holman) has argued, under the auspices of the CPC the Committee on Therapeutics sometimes sent products out to physicians for clinical testing and gauged the results against the claims made by their manufacturers. In doing so, the Committee on Therapeutics sought to advance both the methodological and normative framework of the new epistemic regime. As member Torald Sollmann noted in 1912:

A fact is a fact, whether it is discovered by an academician or by a hod-carrier. . . The man who makes an empirical discovery, who believes it to be a fact, has the moral obligation to establish by exact observation that it is a fact and not the figment of his imagination. . . The essential conditions are not in the material, but in the mental equipment, of the investigator. What is needed is, first of all, a frank dislike of cant—an inability, native or acquired, to accept superficial impressions and hasty conclusions as demonstrated fact.

What Sollmann makes clear is: (a) the professional status and reputation of an investigator is immaterial to whether or not that person can discover facts; (b) the source of objectivity is in both the methods and the epistemic virtues—the “mental equipment”—of the investigator; and (c) that investigators have a “moral obligation” to manifest epistemic virtues such as personal disregard, careful observation, and skepticism to claims not so established. ⁴⁹

As Stewart’s work suggests, manufacturers in the so-called ethical wing of the industry played an important role in establishing the new epistemic practices that increasingly governed drug discovery. In the decades before World War I, manufacturers increasingly invested resources in establishing and maintaining in-house scientific laboratories, for example, a trend which accelerated following World War I. Manufacturers also sometimes used clinical trials as a means to promote sales—in 1903 the Mallinckrodt Chemical Company thus proposed a sophisticated series of clinical experiments at Mount Sinai Hospital, New York City, in order to demonstrate the superiority of their brand of ether as compared to Squibb’s product. ⁵⁰ Between World War I and World War II, manufacturers in the so-called ethical wing of the drug industry embraced the use of increasingly systematized forms of clinical trials in order to both determine and demonstrate the utility of new drugs. During the 1920s, for example, Parke-Davis worked with physicians in Detroit to clinically test a serum remedy for tuberculosis at a sanatorium near the city. The company also maintained a close working relationship with regulators at the Public Health Service as they sought to investigate the utility of the product. In other words, manufacturers such as Parke-Davis were active participants in the broader process of establishing regulatory objectivity through the collective production of laboratory and clinical evidence. ⁵¹

Yet it is also clear that manufacturers quickly recognized that violating the evolving epistemic norms that underlay the emergence of regulatory objectivity might make financial sense. Manufacturers in the ethical wing of the industry sometimes preferred to refrain from sending their products to the CPC, for example, because they knew that the CPC might make a bad evaluation of their product. They also sometimes kept their manufacturing methods secret, intentionally obfuscated findings about their goods, and otherwise acted in ways that violated the epistemic norms of the developing therapeutic regime. ⁵² The development, marketing, and legal battle over a product named Sargol is a good example of this dynamic. Sargol was first introduced in 1908 by a patent medicine company run by a man named Wylie Jones. The product was widely advertised to the public as an amazing discovery that would help people gain weight—as the company put it in one newspaper advertisment, “Sargol makes puny, peevish people plump and popular.” ⁵³ In 1916 the federal government prosecuted the company for mail fraud based on the assertion that their advertising was deceptive. The resulting trial lasted twelve weeks, over two hundred witnesses were examined, and over two thousand pieces of evidence submitted. ⁵⁴ Not surprisingly, the trial was waged along the lines of the new epistemic regime, with the government making its case by arguing that Jones’ claims about his product were not supported by scientific evidence. As the CPC later noted in an overview of the case, “the medical testimony submitted by the government proved beyond controversy that the pretenses set out by the Sargol concern were absolutely worthless so far as flesh or tissue building is concerned.” The Sargol Company was eventually fined $30,000 for fraud and the drug was banned from being sold or advertised through the mail. ⁵⁵

In many ways this was a fairly standard regulatory action for the time. Jones had, from the perspective of the new therapeutic regime, engaged in a dangerous form of quackery by making claims for his product that were not supported by scientific evidence. His conviction for mail fraud was one small part of the broader effort among reformers to suppress the sale of what they considered dangerous nostrums, and the Sargol case featured prominently in the CPC’s Nostrums and Quackery. More surprising, however, is that during the trial it was revealed that Parke-Davis had in fact manufactured the pills for the Sargol Company from a secret remedy. Indeed, the company had an entire department dedicated to producing “private formulas” for both individuals and other manufacturers. ⁵⁶ This was a significant breach of ethics on the part of the company: among the defining features of a so-called ethical firm was, supposedly, the refusal to manufacture goods made from secret ingredients and the refusal to advertise directly to the public. Orthodox physicians had long considered both to be hallmarks of quackery. Yet the revelation did little to damage Parke-Davis’ reputation. After a brief flurry of criticism the issue died down and was forgotten. Parke-Davis retained its status as a leading ethical firm despite its overt breach of the scientific norms that it claimed to uphold. The company had thus come a long way from the moment in 1882 when George Davis suggested that Francis Stewart change his scientific article to better promote one of Parke-Davis’ products. A willingness to violate the scientific and ethical norms of orthodox medicine had been operationalized at the firm as one business strategy among many. The company would, unfortunately, continue to pursue this strategy in the years to come. ⁵⁷

Conclusion

This paper has sought to describe a transformation in the epistemic framework governing the investigation and sale of pharmaceuticals. We have argued that a transformation took place in the latter decades of the nineteenth century, in which the assumptions behind the emergent framework of physiological therapeutics and clinical experiment were linked to a need among manufacturers to commercially innovate new products. Under this new framework, clinical trials operated as a justificatory mechanism for the sale of new goods; unlike an earlier time, the key question of ethical legitimacy of a product now hung on the question of demonstrable and generalizable utility rather than monopoly status. Yet the efforts of therapeutic reformers such as Stewart to establish clinical utility as the primary framework through which to understand the ethical legitimacy of a product, and with it the epistemic virtues that underlay the ability to successfully create this type of knowledge, also had an important unintended consequence: it made scientific claims about drugs a primary vector for marketing and, in doing so, made the suppression or distortion of such evidence a potential part of profit maximizing behavior for drug companies. The effort to establish regulatory objectivity had a similar effect. Even as a regulatory regime based, in part, on scientific claims generated through laboratory and clinical science was established, the pharmaceutical industry discovered that suppressing knowledge about its practices and products could be used to evade regulatory goals.

One of us (Holman) has argued that the tension between regulatory processes and pharmaceutical industry efforts to evade them should be understood as an epistemic arms race, in which regulatory practices are constantly evolving to constrain epistemically deleterious industry practices while industry practices evolve to evade regulation that cuts into corporate profit. ⁵⁸ This perspective helps explain the struggle between the effort to establish regulatory objectivity and the effort to maximize profit through practices that evade epistemic constraints. Furthermore, this perspective also helps explain the changing normative framework that governs how therapeutic reforms have long evaluated pharmaceutical industry behavior. We have suggested that epistemic virtues and vices are not fixed sets of traits but are instead historically contingent, and in this case we have demonstrated that changes in epistemic virtue were motivated in part by the economic imperatives of drug discovery. One of the many ironies in the story that we tell is that the new epistemic virtues embraced by therapeutic reformers such as the CPC originated, in part, in efforts by drug manufacturers such as Parke-Davis to expand their markets and maximize their profits.

We have thus sought to describe a moment in which the epistemic and normative framework that many critics of the pharmaceutical industry now operate within came to pass. As we mentioned at the beginning of this paper, for those of us who are concerned to preserve the place of normative claims within our descriptions of scientific practice it is difficult, perhaps impossible, to look at the recent history of the pharmaceutical industry without characterizing at least some of its behavior as unethical. We thus find ourselves operating within the very framework that Stewart and other reformers in the late nineteenth century sought to establish, one in which both laboratory science and rigorously designed clinical drug trials that provide us with presumably valid scientific knowledge and commercial forces, if not properly restrained, operate as a threat to that knowledge. We thus share the assumption that the “personal equation” should be removed from clinical drug trials as much as possible in pursuit of evidence that applies across different cases, that scientific claims “must contain the truth, the whole truth, and nothing but the truth, to the extent that truth may be obtained,” and that—as Stewart noted in 1901—“the exploitation of new products by exaggerating their merits and repressing knowledge of failures is one of the most dangerous forms of quackery.” ⁵⁹ The fact that we share Stewart’s perspective on these issues indicates that the epistemic virtues that he helped introduce continue to shape the way that we think about knowledge-making practices in the construction of pharmaceutical markets. So too does our assumption that commercial and scientific practices can be usefully conjoined as long as the behavior of the pharmaceutical industry is properly restrained. Finally, our willingness to make normative claims about the relationship between science and markets—and to label some behavior as unethical—is itself based on the changing nature of epistemic virtue and vice that we have sought to elucidate and, in turn, an idealized vision of therapeutic rationality that we use to ground our claims. This does not mean that such claims are not warranted. It simply means that we, like Stewart and Davis, take seriously the possibility of a future better than the past.