Book Reviews

John S. Haller Jr., Shadow Medicine: The Placebo in Conventional and Alternative Therapies, Columbia University Press, 2014

Placebos are much discussed in both the medical and philosophy of medicine literatures. Once narrowly defined as inert “sugar pills,” (Holman 2015), they now are now most often taken to be “treatments that appear similar to experimental treatments, but that lack their characteristic components” (Howick et. al. 2013). In addition to their use in the control groups of many clinical trials, placebos are also now widely recognized by medical practitioners to be powerful therapies in themselves, often outperforming conventional drug therapies in these studies.

Given this, I find Haller’s book, which is divided into six chapters, “Evidence Based Medicine,” “Postmodernist Medicine,” “The Powerful Placebo,” “Politics of Healing,” “Complementary and Alternative Medicine’s Challenge,” “Reassessment,” an introduction and an appendix, to be interesting, yet difficult to interpret with regard to his main thesis. As far as I can tell, Haller’s central claim is that, given the currently available evidence, we should understand complementary and alternative medicine (CAM) treatments to be placebos rather than actual treatments. In Haller’s view, CAM is an “opinion-based system,” which is “not unlike faith healing,” (p. 82) and is concerned not with mechanisms of action, but rather solely with patient outcomes. CAM is thus contrasted in the book with evidence based medicine (EBM), which Haller argues relies on scientific evidence (in particular the randomized controlled trial, or RCT) rather than personal belief and is concerned not just with whether or not a treatment works, but also with how it does. On his view, EBM is thus “rationalist” while CAM is “empiricist.”

This characterization of EBM vs. CAM, while central is the book, is problematic. In contrast to what Haller asserts, it is a hallmark of EBM that it is explicitly not concerned with mechanistic evidence. That is exactly the power (or the pitfall) of the much celebrated RCT. RCTs are designed to tell us whether or not a treatment works, not how it does (Kennedy and Malanowski 2018). On the EBM paradigm, mechanistic reasoning is considered to be a far inferior form of evidence to randomized trials, and in some cases is not even considered to be evidence at all. It seems fair to say, then, that both EBM and CAM are primarily concerned with therapeutic effectiveness and patient outcomes, rather than with mechanisms of action. (As an aside, this should mean that, contra to what many CAM practitioners argue, the RCT should in fact be a reliable method for testing the effectiveness of CAM therapies. On the other hand, this means that EBM practitioners cannot dismiss CAM treatments as “sham,” merely on the basis of the lack of a mechanistic explanation that describes how such therapies, such as homeopathy. By its own lights, EBM argues that mechanisms either don’t matter or don’t matter much. Instead, what matters in medicine is whether or a not a treatment is a) safe and b) effective.) The same, it seems, can be said of CAM, but this important similarity between EBM and CAM seems to escape Haller’s notice.

Further, in some parts of the book, Haller seems to contrast “placebos” with actual treatments, while in other parts, he seems to acknowledge that placebos are treatments. On the one hand, he claims that the placebo is “a product of postmodernist medicine,” which he describes a reaction to and against reductionist scientific medicine, because it interjects “subjectivity, uncertainty and ambiguity into the clinical encounter” (63). This seems to suggest that the way placebos work cannot (or at least should not) be understood scientifically. On the other hand, however, Haller argues that placebos are “real,” in that they “affect patients physiologically as well as psychologically. They alter blood pressure, heart, respiratory rate, and even body temperature” (73).

This tension, between the old view of placebos as inert and the newer view that acknowledges that they can be efficacious (and testable) treatments in themselves is certainly an issue well worth exploring – and Haller should be commended for doing so. The book could be much clearer, however, in exactly what it is arguing. Is the claim that placebos can and should be used in medicine (broadly construed so as to include both EBM and CAM)? Or is it that CAM shouldn’t be understood as medicine, properly construed, because it relies on a placebo effect” that cannot be scientifically measured or verified? Or is it that “Western science needs to advance beyond the current reductionist model to some blending of the subjective and social aspects of healing that includes the placebo?” (p. 157) Should or should not placebos be considered as potentially efficacious treatments in either CAM or EBM? And is this something that can be objectively decided? These questions, while hovering under the surface in the book are neither clearly explicated nor adequately answered.

The book does do a good job of giving a thorough history of the evidence based medicine movement and the advent and subsequent widespread acceptance of the randomized controlled trial (in the first chapter), as well as the history of homeopathy (in chapter 5). Readers interested in these topics will find a helpful resource here. For a more complete philosophical and medical exploration of placebos, however, they will likely need to look elsewhere.

Ashley Graham Kennedy

Florida Atlantic University

Boca Raton, FL



Holman, Bennett. (2015) “Why Sugar Pills are not Placebos” Philosophy of Science, 82 (December 2015) pp. 1330– 1343.

Howick J, Friedemann C, Tsakok M, Watson R, Tsakok T, et al. (2013) Are Treatments More Effective than Placebos? A Systematic Review and Meta-Analysis. PLoS ONE 8(5): e62599

Kennedy, Ashley and Sarah Malanowski. (2018) “Mechanistic reasoning and informed consent.” Bioethics p. 1-7.

Leave a Reply

Your email address will not be published. Required fields are marked *