Like philosophy of biology and philosophy of physics, philosophy of medicine is a branch of philosophy of science. This means that philosophy of medicine studies the scientific concepts, commitments, reasoning, and methods associated with medicine. These include the concepts of health, illness, and disease, commitments to various biological theories and medical models, the methods of scientific experiment and probabilistic reasoning. In Philosophy of Medicine: An Introduction (Routledge, 2018), R. Paul Thompson and Ross E.G. Upshur attempt to orient readers to the central philosophical discussions on each of these topics.

 

Chapter 1 surveys the scope of philosophy of medicine. The authors begin by situating philosophy of medicine in the broader disciplines of philosophy and philosophy of science. They offer two historical examples to draw out some philosophical questions unique to medicine: James Lind’s proto-experiment that led to the prevention and treatment of scurvy and Edward Jenner’s discovery of the smallpox vaccine. Apart from an unclear digression into Victor Bolie’s model of the relationship between glucose and insulin, they highlight important philosophical questions from these historical cases that set the stage for the remaining chapters.

 

In chapter 2, Thompson and Upshur discuss conceptual problems with defining health, disability, and disease. Despite extensive debate among philosophers and scientists, what counts as healthy is ambiguous (goiters result from the thyroid gland performing exactly as it is “supposed” to in response to low levels of iodine). And normal health is often relative to arbitrary factors (male horses have 42 teeth, while female horses have 40). Further, health and disease are value-laden concepts, which means that, whether a physiologic state is “good” depends on the question at hand. For instance, though mild sickling of red blood cells (sickle cell anemia) causes a number of health problems, including pain and fatigue, it also provides resistance to malaria and milder symptoms in those who contract it. This raises important questions about how symptoms are characterized by medical professionals, and thereby, what implications these have for treatment decisions and patient outcomes. Thomson and Upshur largely steer clear of ethical issues, however, they note that this debate has important social implications for whether some conditions are labeled “disabilities” or “diseases,” as in the cases of homosexuality and addiction.

 

Unfortunately, the authors do not mention the extent to which the fields of psychology and psychiatry have taken the problem of defining disease seriously. Later, in chapter 11, they explore controversies over defining and identifying mental illness, but they do not refer back to this general problem of defining disease. While psychology and psychiatry have a history of stigma related to this problem, it is noteworthy that other branches of medicine seem immune. For instance, while homosexuality is no longer regarded as a mental illness, some deaf communities embrace deafness as constitutive of a linguistic and cultural minority rather than a “disability group,” and there are serious concerns about how to treat patients who refuse life-saving amputation and patients who request amputation of healthy limbs.

 

Chapter 3 aims to introduce readers to the differences between scientific theories and models. The authors begin by distinguishing three approaches to theories: syntactic, semantic, and pragmatic. But after an extended discussion of the syntactic account and its deficiencies, the chapter loses coherence. Rather than discussing the semantic and pragmatic approaches to theories, Thompson and Upshur introduce Tarski’s semantic approach to models, offering Mendel’s work on pea plants and the standard account of menstruation as examples. Unfortunately, however, they do not define “model” or distinguish them from theories, so readers are left to infer that any scientific activity that specifies entities presumed to exist, identifies their properties, and explains how those entities change over time constitutes a model.

 

In attempt to tie the content of this chapter to medicine, Thompson and Upshur give a cursory description of immunology. Though they compare immunology with the syntactic theory of Newtonian mechanics, they describe it terms of their semantic account of models: entities, properties, and changes over time. Further, they do not explain the theory of immunology, they only describe in scientific terms what the immune system does, offering no medical example of how the system works. Though they close by saying the examples in this chapter show the importance of theories and models for medicine, readers are left unclear as to just what that importance might be and whether they raise or address any philosophical questions.

 

Chapter 4 contrasts materialism with explanations that “go beyond the material world” and reductionism with antireductionism. They claim that materialism is essential for the study of medicine, even if only as a heuristic, and they give no indication of what a non-materialist theory would look like or that anyone takes such views seriously. They offer a slightly more robust discussion of reductionism, distinguishing ontological reductionism from theory reductionism. On the former, every scientific entity can be explained in terms of some more basic scientific entity: molecules can be explained in terms of atoms, atoms in terms of sub-atomic particles, and so on. On the latter, every scientific theory can be explained by a more comprehensive theory: for example, population genetics can be explained in terms of molecular genetics. The only anti-reductionist view they entertain is holism, according to which a whole natural system exerts causal influence on its parts, and not simply vice versa. They ignore Thomas Nagel’s distinction between ontological and epistemological antireductionism and Fodor’s and Putnam’s accounts of multiple realizability. The relevance of reductionism to medicine is brief and underdeveloped. They note that medicine operates at both the population level and the molecular level, but they do not tell us whether or how these levels are related or whether there are philosophical problems associated with their operating at both levels.

 

Chapter 5 introduces theories of probability and their relevance for medical research. Unfortunately, this chapter is disorganized and misleading for readers not already versed in probability theory. The authors open by noting that probability is philosophically rich but support this with a casual speculation from Ian Hacking on how probability became an academic subject. They then attempt to explain the logical interpretation of probability by offering a series of quotes from John Maynard Keynes that only support the idea that he was attempting to develop a logical interpretation. To explain what a logical interpretation might look like, they spend two pages on the nature of axiomatic systems, rehearsing the history of non-Euclidean geometries.

 

Their account of the logical interpretation is followed by relatively clear expositions of the frequency and subjectivist interpretations of probability. But then they devote a mere two paragraphs on the propensity interpretation, in which they neither define propensity nor distinguish it from other interpretations except to say, erroneously, that it is “a variation on the frequency interpretation.” They then introduce the basic laws of probability without any indication that these laws apply to all interpretations. They explain Bayes’s Theorem for expressing conditional probability but then immediately apply the Theorem in the Bayesian interpretation of probability without distinguishing the Theorem from the interpretation, and without noting that it is a distinct interpretation, rivaling the frequentist and propensity interpretations. After a confusing tour of basic probability theory with no medicine in sight, this chapter ends with a promise that its material will be useful for the discussion of randomized control trials in chapter 7.

 

Chapter 6 discusses some of the philosophical problems with causation and induction. It opens with the promise that closed chapter 5—this material will be useful in chapter 7. And like chapter 5, it is a confusing romp through material that deserves much more space and better examples. The authors spend two pages explaining that the conclusions of deductive arguments follow necessarily from their premises. This might be appropriate for someone with little philosophical background. But they attempt this explanation with complex examples from mathematics: the debate over the Axiom of Choice in set theory and finding the domain and range of a real-number function.

 

Their conclusion to this point is that causes can only be identified inductively. But confusion mounts when they regard any attempt at inductive reasoning as fallacious because it is “invalid,” highlighting the hypothetico-deductive method of scientific reasoning as historically problematic for this reason. Inductive arguments are, indeed, invalid insofar as their conclusions do not follow necessarily from their premises, but this does not mean they are fallacious—they are simply a different form of reasoning. And it is important to note that no one regarded the hypothetico-deductive method as aspiring to anything as robust as deductive validity. Further, the authors are mistaken when they say Karl Popper attempted to address this “logical problem” by offering his method of falsification.

 

 

 

The authors are right, however, that there has long been philosophical controversy over the nature of causes. Hume argued that there are only constant conjunctions of perceptions, David Lewis argued that causes are simply what happens under similar conditions in close possible worlds, and so on. And given the complexity of the world and how we use language to individuate events, it is difficult to pick out precisely what effect is produced by what cause. Thompson and Upshur make much of this in chapter 7, arguing that we should not take randomized clinical trials to be the gold standard for medicine largely because they cannot identify causes to any strong degree of probability. In this chapter, they simply note this problem and give some medical examples of how causal language is used in explanations of disease. The chapter ends on a confusing note, however, when the authors reintroduce the problem of theories in medicine. They note that there is a “growing chorus of [scholars] who are challenging the lack of emphasis on theory in clinical medicine” (80), but they offer no hint as to the nature of this challenge or provide any citations for them. And they are critical of Carnap’s inability to formulate “a successful inductive logic” (80) but fail to explain that this is unrelated to their use of inductive reasoning in the rest of the book. Someone unfamiliar with the history of scientific explanation might assume that Thompson and Upshur are rejecting inductive reasoning in medicine while continuing to use it!

 

In chapter 7, the authors argue that medicine should reduce its reliance on randomized control trials (RCTs). While they rightly identify both conceptual and practical problems with achieving the randomization necessary for strong causal inferences in RCTs, they infer from these that medical researchers should only use the term “association” instead of “cause.” They also advocate for increased use of alternative experimental models that can sometimes accomplish what researchers need without the burden or impracticability of RCTs: systematic reviews, observational studies, and time-series studies, among others. Unfortunately, their descriptions of these alternatives are underdeveloped and include no philosophical discussion. They close by introducing what would appear to be one more strategy for identifying causes in medicine, proposed by Austin Bradford Hill. Oddly, Hill rejected the idea that his strategy could identify causes, and thus it is unsurprising that Thompson and Upshur conclude it is insufficient for that purpose.

 

Chapter 8 briefly describes different measurements in medicine, including odds ratio, incidence rate, and number to treat. However, this chapter is descriptive with no hint that there is philosophical controversy over them. Chapter 9 sketches a taxonomy of types of reasoning in which clinicians engage: prevention, diagnosis, therapy, etc. The authors highlight challenges that face each type of judgment, but their central point is that every type of clinical reasoning is fraught with uncertainty. The chapter ends with an odd discussion of Stephen Toulmin’s model of reasoning and Douglas Walton’s defense of argument as dialogic, though they provide no examples of how to use them or how they might enhance clinical judgment.

 

Chapter 10 introduces the idea that third-person, scientific perspectives are not the only relevant perspectives in medicine. Drawing on phenomenology and narrative methods, some argue that first-person perspectives on health and disease are crucial both for understanding and treating patients. Thompson and Upshur explain how qualitative research methods can be used to gather first-person perspectives to help inform clinical judgment. The central philosophical question is how first-person perspectives should be used in clinical judgment and whether they can be evaluated critically. While some might hold that these views are unimpeachable, Cheryl Misak argues that it is imperative that they be evaluated, at least for internal coherence and consistency with other evidence. A notable omission from this chapter is the growing philosophical literature on the ways some groups—especially patients and other historically vulnerable populations—are discredited as knowers, a phenomenon known as “epistemic injustice.” This literature adds important nuance to the discussion of using first-person accounts effectively in medicine.

 

Chapters 11 and 12 provide cursory sketches of theories of mind (dualism, materialism, Freudianism), how mental illnesses are defined in light of those theories and developments in the neurosciences, and distinctions among different ways of conceiving medical practice, including evidence-based medicine, personalized medicine, and values-based medicine. As with some other chapters, there is little philosophical analysis or discussion of philosophical debates.

 

Overall, the book is underdeveloped and poorly organized. Many of its discussions presuppose a stronger background in philosophy of science than could be expected of a reader of an introductory text, and in many cases the medical examples feel tacked on to general issues in philosophy of science. Further, a number of interesting topics in philosophy of medicine were not addressed, including statistical errors in clinical judgment, the demarcation between clinical medicine and science, the social implications of science denying in medicine (e.g., anti-vaxxers), medical error, medical research retraction rates, and off-label drug use. The authors offer a few teasers for interesting philosophical discussions, but readers will likely come away frustrated, and those will little background in philosophy of science should turn to a more comprehensive introduction to philosophy of medicine.

 

 

 

© 2018 Jamie Watson

 

 

 

 

Jamie Watson, Ph.D., Assistant Professor in the Department of Medical Humanities and Bioethics, UAMS