In Praise of a Healthy Iconoclasm

Americans trust scientists. Data show that 44 percent of U.S. adults say they have a great deal of confidence in the scientific community. Underlying this trust is likely the assumption that science will do what it has historically done ever since it developed its core methodologies—pursue truth through empirical means, guided by data rather than by other incentives, be they financial or partisan.

It is then worth asking—do we, in the scientific community, do this? On one level, the answer is obviously yes, we do, though perhaps imperfectly. But what if we modify the question, to ask: Do we do this all the time, or at least enough to fully justify the public’s trust in us? How often do we think for ourselves, guided principally by data, and how often are our thoughts shaped by other factors? I would argue that we are susceptible to other factors, though not necessarily in the sense of being unduly partisan or subject to financial incentives. Instead, science, it seems to me, has a weakness for groupthink, for being swayed by the consensus simply because it is the consensus.

In 2015, I wrote a piece for Fortune about public health’s complicated relationship with salt. The impetus for the piece was the conflicted attitude towards the potential adverse consequences of salt intake in the general population among public health authorities. Opinion about sodium intake had roughly settled into two camps. One camp felt that salt is a largely negative influence on health, and the other felt that the science did not support such a view. I became fascinated by this, not because it was directly linked to my area of study, but because I could not quite understand how such divergent views could co-exist, strongly held on either side. Both camps of the salt debate leaned heavily on their own body of research from which they drew conclusions. On one side, there was a body of knowledge that supported a dim view of population salt intake, and on the other there was a body of knowledge that seems to contradict this view.

How can there be two separate schools of thought, supported by two separate data pools, each supporting the opposite conclusion? What we found, when we dug into this, was that this was indeed possible when two schools of thought effectively sealed themselves within bubbles of unquestioned groupthink, embracing only lines of inquiry that reinforce the position to which they are most sympathetic. At some point, data on salt ceased being just data to those who engaged with it and became something else. The data became a narrative—competing narratives, in fact. As scientists began to perceive a narrative—“Salt good” or “Salt bad”—they started believing only research that supported the narrative they embraced. Most consequentially, perhaps, they started to only cite research that supported that narrative. This informed the creation of parallel scientific universes, in which conclusions were regarded, by their supporters, as close to self-evidently true, bolstered by what appeared to be a robust body of evidence.

To my mind, this teaches us that effective science should have—in addition to empirical rigor—a measure of healthy iconoclasm, as a means of resisting easy narratives and the groupthink that can support them. I do not mean a cranky iconoclasm that thrives on a reflexive rejection of consensus for its own sake, but rather a reflective iconoclasm that questions, that asks if what we think we know is as airtight as it may seem.

I suggest that such an instinct would serve us well in our efforts to find solutions to the problems that threaten health. Often, groupthink can weaken our efficiency in the face of such problems. A key challenge posed by groupthink is its tendency to give rise to untenable, absolutist positions on issues, which can stand in the way of real-world progress. During COVID, we saw a number of examples of science taking positions that emerged from dug-in, consensus-based thinking which arguably ignored critical nuance. I have written previously, for example, about the goal of “zero-COVID” (in which success against the virus is defined as nothing less than the elimination of all infections and deaths). This goal, to my mind, is impractical from a perspective rooted in pragmatic engagement with the realities of disease mitigation. However, from the perspective of scientific groupthink it makes perfect sense. For those of the opinion that COVID cannot in any meaningful sense be managed—and that the goal of public health is to eliminate risk rather than mitigate it—no other position could do.

The divide within the scientific community over COVID was well-captured by the dueling memoranda of the Great Barrington Declaration and the John Snow Memorandum. The Great Barrington Declaration argued for a “focused protection” approach: safeguarding the vulnerable, while minimizing the use of lockdowns and other restrictive measures, with an eye toward preventing the societal harms they can cause. The John Snow Memo argued strongly against such an approach, that nothing short of all-hands-on-deck efforts to contain COVID should be acceptable. These two perspectives have become, to many who hold them, more than perspectives. They have become narratives, and while there is merit in each approach to the pandemic, it has become difficult for partisans of either side to see this, once they become invested in their preferred story. A call for iconoclasm, then, is a call for being constantly aware of the other camp, so that narrative does not take precedence over data, allowing us to learn from alternative perspectives.

I realize that the pursuit of a healthy level of iconoclasm in science may be more easily recommended than done. It is one matter to think a bit more critically about positions we hold; it is another to pursue a research direction that is outside conventional wisdom about what constitutes a valid line of inquiry. And it is difficult, in any context, to publicly express thoughts which go against the grain of the consensus, particularly when that consensus is supported by one’s peers and friends. But when groupthink proliferates, it can cause us to miss important details, ultimately weakening our effectiveness as public health professionals, placing lives at risk.