1. The publication–citation–reputation (PCR) system as a discovery procedure

Our argument presupposes a process-oriented model of the normal operation of science in the absence of government involvement or, more generally, in the absence of exogenous intervention. According to this model, science is an institutionalized form of interaction between generally self-interested individuals each of whom publishes their own work and, in turn, uses and cites the works of others. Under normal circumstances, the various incentives that individuals encounter when engaged in this mode of interaction, including the incentive provided by the norm that emphasizes empirical correspondence, promote the emergence of scientific knowledge about the world.Footnote ¹

Normally, the scientific process manifests bidirectional feedback loops that constrain scientific activity to promote the emergence of scientific knowledge. A publication is evaluated by other scientists who might either adopt or criticize it in the form of a (positive or negative) citation, which feeds back upon the reputation of the original author, thereby affecting the reception of their publications and prospects for professional advancement. This system of institutionalized incentives ensures that what is ultimately accepted as scientific knowledge has been exposed to the daylight of rational scrutiny. The scientific knowledge that emerges from the PCR process has been subjected to, criticized from the perspective of, and deemed in acceptable conformance with, the assessment standards of the relevant scientific community. It is through the PCR process that scientific error (or, worse yet, deliberate fraud) is identified and corrected.Footnote ²

However, where this process is either absent, superseded, or perverted, so too are its error identification and correction functions, and, other things being equal, the knowledge that emerges from such science is less actionable than the knowledge that emerges from a normally functioning PCR process. Any plan of action that presupposes a proposition resulting from a negated or perverted PCR process is more prone to failure than any otherwise identical plan that presupposes no such proposition.

As antecedents for this model of the system of science, we particularly cite R. K. Merton, D. L. Hull, M. Polanyi, and F. A. Hayek. Merton (Reference Merton1996: 333), summarizing a lifetime of work in the sociology of science, observed:

Hull (Reference Hull1988: 309), in a detailed study of two competing groups of research biologists, underlined with first-hand observation the seriousness with which the researchers regard the reputational effects of the citation of publications: “Individual scientists want credit for their publications, as much credit as possible. … [They] readily acknowledge that they crave for recognition[.]” Polanyi (Reference Polanyi1962; Reference Polanyi1967: 54–55) conceived of science as a dynamic order of freely coordinating individuals who are “in fact cooperating as members of a closely knit organization” by virtue of “the adjustment of the efforts of each to the hitherto achieved results of the others,” and argued passionately (against much opposition) for science to be free from external direction, pointing out the inherent contradiction between such direction and the process of discovery. Hayek (Reference Hayek1960: 388–389) who, contemporaneously with Polanyi, saw science as a spontaneous order, also argued that the advance of knowledge requires that scientists have the freedom to pursue their work autonomously within the evolved conventions of science and unencumbered by “political interference.”Footnote ³ However, the most important antecedent for our model was Hayek's (Reference Hayek and Caldwell2014b) exposition of the epistemic characteristics of the market's price system, which inspired the identification of its analog, the PCR process, in science.

2. An epistemic theory of social order

This model of the emergence of scientific knowledge is a particular instance of a general epistemic theory of social order that, like our assumed epistemology, has its roots in the work of F. A. Hayek. Scheall (Reference Scheall2015) argues that an epistemic theory of economic order can be synthesized from various parts of Hayek's methodological writings (see Hayek, Reference Hayek and Caldwell2014a, Reference Hayek and Caldwell2014b, and, especially, Reference Hayek and Caldwell2014c).Footnote ⁴ Our goal is to extend this theory of economic order into a more general epistemic theory of social order.

According to this epistemic theory, social order emerges endogenously in some domain – be it the economy, science, politics, or society itself – to the extent that the presuppositions of the plans (i.e. “beliefs”) of the individuals active in this domain are internally and externally coordinated, i.e. mutually compatible and adequate to environmental circumstances. This is to say that, barring changes in either the relevant beliefs of other individuals or in environmental conditions, the individuals active in this domain will all be able to effectively implement their respective plans – in other words, that these people know, in our sense of knowledge (Scheall, Reference Scheall2016). Social order obtains in some domain to the extent that the plans of the relevant individuals are actionable.

It follows from this conception that social disorder in some domain is also epistemic, i.e. a matter of the beliefs of the individuals active in the respective domain being uncoordinated either with each other or with other aspects of the environment. Social disorder in some domain constitutes a lack of knowledge in our sense: it means that the plans of some individuals in the relevant domain are bound to fail. The marks of social disorder are disappointed expectations and failed plans, and the measure of disorder is the extent of disappointment and failure.

It should be obvious that the realization of perfect social order is practically impossible. Simply put, the data do change, and typically too rapidly to afford perfect coordination by anyone. However, our epistemic conception of social order does not assume or imply either that any individual ever knows anything entirely – actionability is a continuum, not a dichotomy – or that perfect social order in some domain is possible. That some expectations are disappointed is part and parcel of social life.

Given this epistemic theory of social order, the central problem of the social sciences is the discovery of mechanisms operative in society that either promote or impede the coordination of individual knowledge.

Hayek (Reference Hayek and Caldwell2014b) argued that, with respect to the economy, a system of freely adjusting prices is a necessary element in minimizing the disappointment of expectations. Hayek's (Reference Hayek and Caldwell2014b: 99–100) famous “tin example” illustrates the role of the price system in the coordination of knowledge in the economic realm, and thus its central part in the maintenance of economic order. Hayek describes how changes in supply and demand, although known first hand by only a few, spread through the economic system, affecting the prices of substitutes and complements, without downstream market participants needing to know more than the prices relevant to their local plans.Footnote ⁵

The PCR system plays an analogous (but by no means identical) role in coordinating the beliefs of scientists. An unfettered PCR system is the mechanism whereby the presuppositions of individual plans of action in the scientific domain come to be coordinated both with each other and with other facets of the external environment.

We can describe the operation of the PCR process in terms mirroring Hayek's tin example. Imagine that a novel theoretical explanation of some phenomenon has been published. Just as price signals may or may not be acknowledged and put to use by market participants, “publication is a signal inviting response, which may or may not be recognized and acted on by other scientists” (McQuade and Butos, Reference McQuade and Butos2003: 140). The interaction between the publishing scientist and the broader scientific community will serve to coordinate plans only if some among the latter find the explanation useful in furthering their own work. Scientific custom typically requires that the use of someone else's work be acknowledged in the form of an explicit citation to the author of the original publication. However, it is important to recognize that:

Provided the PCR system is functioning without interference – that is, provided that scientists privy to a publication are not hindered from registering their impressions via the PCR system – then many members of the scientific community require no immediate first-person knowledge of the contents of the respective publication. If only some scientists know the contents of a publication and respond by citing it (positively or negatively), and if the scientists who receive this signal respond accordingly, by either accepting the new explanation, remaining neutral, or rejecting it in favor of some other, then, as Hayek (Reference Hayek and Caldwell2014b: 99–100) noted, “the effect will rapidly spread throughout the whole … system and influence not only all the uses” and criticisms of the new explanation, but also those of its relatives and rivals. “The whole acts as one” scientific community, “not because any of its members survey the whole field, but because their limited individual fields of vision sufficiently overlap so that through many intermediaries the relevant information is communicated to all.” Thereby, a consensus concerning scientific knowledge is either created, maintained, or destroyed.Footnote ⁶

Our general theory of social order is epistemic in several different senses. It starts from the conception of economic order as well-coordinated knowledge, but it also builds upon Hayek's epistemic explanation of the advent of economic disorder, i.e. as a consequence of “humans [typically, for Hayek, policymakers] acting on the basis of knowledge that they don't in fact possess in such a way that interferes with the functioning of a freely adjusting price system, and which thus hampers the operation of the tendency” for price changes to facilitate knowledge coordination (Scheall, Reference Scheall2015: 104–105). At the highest level of generality, our theory of social order predicts that, wherever there is exogenous interference with the epistemic mechanism that serves to coordinate knowledge in some social domain, failed plans tend to follow – disorder results. Other things being equal, a plan of action is more prone to fail if based on a presupposition that is the consequence of a superseded or distorted price system, or a superseded or distorted PCR system.

A thorough investigation of the various ways in which the PCR process of science might be overridden or distorted by exogenous forces arising from either the public or private realms is surely in order. However, such an inquiry would take us too far afield in the present context. As anyone familiar with the infamous Lysenko affair in mid-century Soviet genetic science can attest, the institutions of the state can be used to manipulate and, in the extreme, negate the normal operation of the PCR process. This is not to deny that there are circumstances in which private non-scientific actors might exert a similarly deleterious influence on the emergence of actionable scientific knowledge, but the most obvious cases should surely be handled first. Even ignoring Lysenko-like cases, the government's role as primary funding source, principal external regulator, and prominent beneficiary of both scientific and technological research make it an obvious place to begin an investigation of the effects on science of the impingement of exogenous forces.

3. Policy-induced social disorder

Economic policymakers falsely believe that they possess the knowledge necessary to make explicit and deliberate policy so as to either maintain economic order or ameliorate economic disorder when it appears. By themselves, such false beliefs would be relatively harmless were they never made the basis of policy. But economic disorder emerges because policymakers hold false beliefs both about the economy and about their false beliefs about the economy:

The consequences of economic policymakers’ ignorance of their own ignorance are clear: “when policymakers pretend to possess the relevant economic knowledge and make policy on the basis of this pretense, their decisions typically impede, either directly or indirectly, the price system's knowledge-coordinating function” (Scheall, Reference Scheall2015: 109). Economic disorder is a condition in which the plans of individual economic actors are uncoordinated – most often, in virtue of policies made on the basis of a pretense of knowledge.Footnote ⁷

Our explanation of social disorder as it emerges from the scientific domain will ultimately be similarly epistemic: acting on the basis of the false presupposition that they possess the counterfactual knowledge of what would emerge from an unfettered PCR process without the possibility of recourse to such a process, politicians make policy that impedes (directly or indirectly) the PCR system's epistemic functions, namely, the coordination of knowledge and the identification, and correction, of scientific errors. The scientific knowledge, such as it is, that emerges from a politically perverted PCR process is more likely to encounter resistance from the world, either from the beliefs of others or from further elements of the environment. For this reason, a plan that presupposes a proposition that resulted from a negated or perverted PCR process is more prone to failure than an otherwise identical plan that presupposes no such proposition.

Social disorder, then, is a consequence of policy made on the basis of policymakers’ false belief in the adequacy of their knowledge for the purposes of maintaining social order. The knowledge that emerges from a politically distorted PCR process is, for anyone who would presuppose it as part of their individual action plan, less actionable (less reliable) than the scientific knowledge that emerges from an unfettered PCR process. Just as economic disorder is a consequence of economically ignorant policies that override the epistemic function of the price system, such social disorder as emerges from the scientific domain flows from scientifically ignorant policies that override the epistemic function of the PCR system.

4. Acknowledging our own ignorance

To keep the discussion as tractable as possible, we are deliberately ignoring many considerations that are almost certainly relevant to explaining why policies that interfere with the PCR process get enacted despite policymakers’ manifest ignorance of the knowledge required to realize the goals of the relevant policies. We are purposely setting aside all considerations specifically concerning public choice and any questions related to the motivations of, and incentives confronted by, policymakers. If science policy is made on the basis of considerations of what is best for the policymaker rather than in the interests of the public, then – to the extent that policy-relevant presuppositions enter the plans of members of the public – this self-interested political behavior is the best explanation of the failure of these plans.Footnote ⁸ However, accounting for the epistemic complications that contribute to the failure of plans based on presuppositions related to science policy while keeping an eye on the motivations of political actors would make this already complex analysis more convoluted still. For this reason, we have adopted the admittedly inadequate assumption that policies which impact the PCR process, regardless of their consequences, are made on the basis of what policymakers honestly believe to be in the public interest.Footnote ⁹ Moreover, there are examples (like the case study below) in which this does seem to be the case. While this does not make issues of the motivations of political actors irrelevant in such cases, it does provide some cover for our decision to ignore these complications in the present context.

The positive value of the theory is that it provides some clues where to find instances of both scientific disorder and social disorder caused by science policymaking, namely, in cases of disappointed expectations among the users of scientific knowledge; it also provides a first approximation of an explanation of such cases that might be extended by integrating some of the public choice considerations we have opted to avoid herein.

We acknowledge the limited testability of the theory: it is extremely difficult, if not plainly impossible, to observe either the presuppositions of an individual action plan or the plan itself, and, thus, its success or failure. However, there is an important sense in which its limited testability is not as damaging as might seem at first glance. Indeed, given its basis in the work of Hayek – who did more than anyone to point out the very limited predictive possibilities (and, thus, limited policy relevance) of the social sciences – the model would be self-undermining were it capable of predicting particular events with any precision. We are constrained to the prediction of patterns, which is as it should be in the social sciences.

The specific pattern prediction that we are making here is that external interference in the operation of the PCR process of science compromises the reliability of the knowledge produced by that system, thereby increasing disorder, both in the science itself and in the greater society. We are not making a claim of optimality for the unhampered process, but we are noting that, like the price mechanism in its domain, it harnesses individual incentives and individual knowledge to produce knowledge of the system's environment that is useful, but inevitably limited and contingent, and that there are no credible alternative institutional arrangements for producing that type of knowledge. In describing the involvement of government in nutrition science, we are presenting a detailed specific illustration supporting our pattern prediction.Footnote ¹⁰

5. An overview of trends in nutrition science

To the extent that there was a medical-scientific consensus at the start of the 1970s concerning the dietary causes of coronary heart disease (CHD), obesity, and diabetes, it stressed their connection with the consumption of carbohydrates (Butos and McQuade, Reference Butos and McQuade2012). The prevailing medical wisdom held that a diet including too many carbs, especially refined sugars and flours, contributed to tooth decay, as well as to weight gain and the various maladies associated with it. However, within a few short years, Americans were being advised to consume fewer calories in the form of dietary fat, especially the saturated fat primarily found in red meat and animal-based foods. Since carbohydrates and dietary fat are nutritional rivals – for any given caloric load, consuming less of one must mean consuming more of the other – and since this novel advice encompassed a recommendation to at least maintain, if not decrease, caloric intakes, the implication was that Americans should avoid red meat and indulge in bread, pasta, rice, fruits, and vegetables. This change was a consequence of the US federal government's interference in the PCR process of nutritional science. The government placed the political cart before the scientific horse. In an ostensive effort to provide Americans with healthful dietary advice, the government short-circuited the PCR process from which a consensus on nutrition science would have otherwise been likely to eventually emerge.

There are several indications that scientific as well as public opinion is circling back in favor of the view that carbohydrates are to be convicted (i.e. the “carbohydrate hypothesis”). Government officials are gradually acknowledging this reversal in the fortunes of the carbohydrate hypothesis. The eighth and most recent (2015) iteration of the US Department of Agriculture's (USDA) Dietary Guidelines for Americans takes a few, albeit slight, steps back from the more trenchant injunctions on dietary fat proclaimed in past editions. If our argument is sound, the deleterious consequences of the government's pronouncement of a nutritional consensus without recourse to a daylighting PCR process have made this turnabout prudent. Recent scientific research suggests that there is no decisive evidentiary basis for the “fat hypothesis” that the dietary fat (in particular, the saturated fat) we consume contributes to ill health.

We take no stance concerning the adequacy of any particular dietary hypothesis. Indeed, we maintain that such a judgment requires the sort of scientific knowledge that can only emerge from an unfettered PCR process. But we do claim that the ascendance of the fat hypothesis is a case in which the PCR process was essentially countermanded by government mandate.

Our model of social order bears interesting implications for the fact that the USDA treats Guidelines as a living document. Each subsequent modification implies that the knowledge upon which each of the previous iterations was based was inadequate. To respond that this is because what counts as scientific knowledge is constantly changing is to beg the question, for it assumes, rather than establishes via argument, that the knowledge that enters into each restatement of Guidelines is the product of a process that ensures as far as possible its actionability. As a matter of historical fact, the knowledge that has figured in the various versions of Guidelines has always been less a product of an open PCR process than of the unreliable caprices of political whim.

6. The “fat hypothesis” and government interference in nutrition science

The fat hypothesis emerged mid-century in the context of two widely accepted social beliefs. First, there was the perception that Americans were in the midst of a heart disease “epidemic.” Not coincidentally, Congress in 1948 created the National Heart Institute and the National Heart Council, precursors to today's National Heart, Lung, and Blood Institute (NHLBI). The US federal government is the biggest player in heart disease research, and the NHLBI is the main vehicle through which it funds research on this and related disorders.Footnote ¹¹ Second, there was the belief that the increasing industrialization and urbanization of America in the latter half of the 19^th and first part of the 20^th centuries coincided with a radical “nutritional transition” away from the widespread consumption of cereals, fruits, and vegetables (i.e. carbohydrates), toward increased consumption of animal foods (i.e. dietary fat). It was but a short step from this purported correlation between the “nutritional transition” and the “epidemic” to the hypothesis that the first was the cause of the second, i.e. that the increased intake of dietary fat at the expense of carbohydrates in the first few decades of the 20^th century led to a rampant increase in the incidence of CHD a few decades later.

However, the alleged epidemic and the nutritional transition are both more myth than fact. Taubes (Reference Taubes2007: 5–13) argues that the perception at mid-century of a heart disease epidemic was largely an artifact of new medical-screening techniques that led to a marked increase in the diagnosis, but not necessarily the incidence, of CHD. Further complicating matters was the fact that, by the middle of the 20^th century, early death from infectious diseases and inadequate nutrition had effectively been eradicated in the United States. CHD is common only among those over 50. The marked decrease in the number of premature deaths “left Americans living long enough to die of chronic diseases – in particular, cancer and heart disease” (Taubes, Reference Taubes2007: 7). Another factor propagating the perception of a heart disease epidemic was the proliferation of new medical categories of heart disease, which led to more deaths being classified as due to heart disease (Taubes, Reference Taubes2007: 7–8). The evidence is similarly suspect for the “changing-American-diet story,” according to which Americans switched their eating habits en masse in the latter part of the 19^th and first half of the 20^th centuries from allegedly healthy carbohydrates to dangerous animal foods.

The main cheerleader for both the changing-American-diet story and the hypothesis that fat is responsible for fattening, diabetes, and CHD was cardiologist Ancel Keys of the Laboratory of Physiological Medicine at the University of Minnesota.Footnote ¹² Keys initially based his argument on USDA data that Americans at the turn of the century had consumed 25% more starches and cereals, 25% less fat, and 20% less meat than in the 1950s and afterward. These claims were not as definitive as Keys imagined. The first systematic reckoning by the USDA of food consumption in the US was not published until 1949.Footnote ¹³ Basic statistics of food production in that report were generated by the Census of Agriculture taken every five years, while intermediate years were estimated from sample data. The estimates of meat consumption only referred to interstate commerce of meat production. Given the absence of data for “intermediate years” and non-federally inspected meat, the consumption data were precarious. Interestingly, the data for foodstuffs not subject to interstate commerce, and especially for meat consumption, beginning in 1909 through 1943, were not only used by Keys but have been regularly reported in more recent data analyses. Taubes (Reference Taubes2007: 11) notes:

The 1942 Report noted the “basic data for computation of estimates of retail slaughter up to 1943 were even more incomplete” than the “year-to-year changes shown by sample data” (Taubes, Reference Taubes2007: 19).Footnote ¹⁴

Keys played the central role not only in publishing research that supported the fat hypothesis but in convincing the American Heart Association (AHA) to adopt the hypothesis as its central operating assumption, and – perhaps most importantly – in acting as a persistent advocate for the hypothesis both among the relevant communities of researchers and in the public eye. Keys’ (Reference Keys1953) hypothesis consisted of two sub-theses. First, that the consumption of fat – originally total fat, but later, especially saturated fat – increases the levels of total blood (or serum) cholesterol. Second, that, since cholesterol is the main component of the atherosclerotic plaques associated with CHD, it is total cholesterol that induces the formation of these plaques. Thus, Keys invoked a purported mechanism from eating saturated fat to increased total cholesterol in the bloodstream, to the formation of atherosclerotic plaques and, thus, to heart disease, elevated blood pressure and hypertension, heart attack, stroke, etc.

Keys (Reference Keys1970) based his hypothesis on data for 1958 through 1964 from seven countries which, at the time, were still emerging from the catastrophe of World War II.Footnote ¹⁵ Given that postwar conditions included both famine and food rationing, Keys’ data may not have constituted the best evidence for or against any hypothesis concerning the health effects of different diet regimens under normal conditions. What might have appeared to be a correlation between fat consumption and heart disease could have been caused by any number of other hardships related to the war. Moreover, the evidence upon which Keys based his hypothesis appeared selective and led to severe criticism of his methodology. Before the field work even began, Yerushalmey and Hilleboe (Reference Yerushalmey and Hilleboe1957) pointed out that the strength of the association observed by Keys depended on the countries selected for the study. If countries other than those included by Keys had been selected, the observed association would have been weak.

In 1957, the AHA insisted that the available evidence failed to support Keys’ hypothesis of a causal relation between dietary fat and heart disease. However, though no new evidence had appeared in the meantime, three years later an ad hoc AHA committee issued a report, co-authored by Keys, that reversed the Association's earlier skepticism. From this point forward, the AHA advocated increasingly stringent recommendations to reduce the consumption of dietary fat and none of its members was a more indefatigable crusader than Keys.

Keys’ persuasiveness with the media, the American public, and his scientific peers, together with the AHA's increasingly restrictive recommendations concerning dietary fat, fed the growing acceptance of the fat-cholesterol hypothesis over the third quarter of the 20^th century. However, the evidence for the hypothesis remained stubbornly recalcitrant. Had it not been for George McGovern's Senate Select Committee on Nutrition and Human Needs, which both elevated the alleged connection between fat, cholesterol, and chronic disease to the level of nutritional-medical wisdom and enshrined the suggestion to reduce fat consumption in favor of carbohydrates as the main “dietary goal” of all Americans, then either controversy would have continued in perpetuity or a normal PCR process would have eventually generated something approaching an unalloyed consensus concerning dietary-scientific knowledge. Without the McGovern committee's stamp of official government approval, Keys’ hypothesis would likely have remained popular only among a small part of the population rather than becoming a nutritional lodestar for the majority of Americans.

This is the critical point. Although private actors were involved, the distortion of the PCR process in this case could not have occurred without the intervention of government actors. Such a combination of private and public opportunism in science is reminiscent of “crony capitalism” and could be labeled “crony science.”Footnote ¹⁶

The report that the McGovern committee issued on January 14, 1977, the first Dietary Goals for the United States, was written “almost single-handedly” by a scientific novitiate, committee staffer Nick Mottern, a “former labor reporter [… and …] researcher for a consumer-products newsletter” (Taubes, Reference Taubes2007: 46). Mottern managed to avoid confronting the refractory evidence concerning Keys’ hypothesis by ignoring it in favor of a near-exclusive emphasis on the research of a single investigator, Harvard's Mark Hegsted, an unreserved believer in the connection between fat consumption, cholesterol, and heart disease.Footnote ¹⁷ Dietary Goals marked “the first time that any government institution (as opposed to private groups like the AHA) had told Americans they could improve their health by eating less fat. In so doing, Dietary Goals sparked a chain reaction of dietary advice from government agencies and the press that reverberates still, and the document itself became gospel. It is hard to overstate its impact. Dietary Goals took a grab bag of ambiguous studies and speculation, acknowledged that the claims were scientifically contentious, and then officially bestowed on one interpretation the aura of established fact” (Taubes, Reference Taubes2007: 44–45; emphasis added).

The resulting controversy over Dietary Goals led McGovern to hold eight subsequent hearings. However, the skeptical experts “were sandwiched between representatives from the dairy, egg, and cattle industries, who also vigorously opposed the guidelines, for obvious reasons. This juxtaposition served to taint the legitimacy of the scientific criticisms” (Taubes, Reference Taubes2007: 47). A revised version of Dietary Goals was soon published that included “a ten-page preface that attempted to justify the committee's dietary recommendations in light of the uproar that had followed” (Taubes, Reference Taubes2007: 48). The revised report acknowledged the presence of disagreement among nutrition scientists and the belief among some that its recommendations could lead to harm, but the committee deemed such destructive consequences unlikely. In effect, Dietary Goals transformed the dietary fat controversy from a scientific into a political issue. The consensus that has marked the last 40 years of dietary science was gerrymandered in large part by the McGovern committee and the subsequent piling-on of other government agencies.

Although the Department of Health, Education, and Welfare and the National Institutes of Health (NIH) considered dietary research to fall under their respective purviews, Congress in 1977 designated the USDA lead agency. In entering the fray, Assistant Secretary of Agriculture Carol Foreman, a former consumer advocate, “believed it was incumbent on the USDA to turn McGovern's recommendations into official government policy” (Taubes, Reference Taubes2007: 48). Once again, like Mottern before her, Foreman would have no truck with scientific controversy, perhaps because of the massive amount of research money at stake. As William Broad (Reference Broad1979: 6) noted, “the fighting began in 1977, and the story of its origin elucidates the politics of an evolving area of research.”

After drafting an agreement with the Food and Nutrition Board of the National Academy of Sciences to appraise the McGovern committee's dietary recommendations, Foreman backed out when it became clear that members of the Academy's brass were skeptical of the McGovern committee's report. Indeed, the Board soon released a report highly critical of Dietary Goals titled Toward Healthful Diets:

Foreman encountered similar skepticism at the NIH and the Food and Drug Administration, where the McGovern committee's report was (at the time) deemed less scientific than political (Taubes, Reference Taubes2007: 50). But she eventually found the Surgeon General's Office sufficiently pliant and representatives of the two agencies set about writing an official set of government-certified dietary guidelines. The USDA was represented by the first head of the Department's Human Nutrition Center, Mark Hegsted. The first USDA Dietary Guidelines for Americans, written by Hegsted and the Surgeon General's J. Michael McGinnis and released in 1980, made Keys’ fat-cholesterol hypothesis the quasi-official scientific position and the recommendation to reduce fat intake in favor of carbohydrates the official dietary guideline of the US federal government. This was the effective end of the policy debate.

It was not surprising that the American public soon accepted the government's low-fat dogma, which was quite harmonious with environmental concerns popular in the 1970s, especially the budding vegetarian movement. The American food industry soon flooded the market with reduced-fat alternatives, such as low-saturated-fat margarines and vegetable oils, and products engineered to conform to the requirements of America's new dietary paradigm.

Meanwhile, the science concerning Keys’ hypothesis of a connection between fat, cholesterol, and CHD was left to catch up. Four smaller trials attempted to establish such a connection within different populations. None succeeded (Taubes, Reference Taubes2007: 53). Further studies aimed less to test any particular hypothesis than to reconcile the evidence with the government's official dietary guidelines. However, much of this research was waylaid by the unexpected discovery of what appeared to be a link between low cholesterol and cancer (Taubes, Reference Taubes2007: 53–54). This link was eventually erased from scientific concern when two NHLBI workshops in the early 1980s concluded it merely a mark of the uniquely cancer-prone; for everyone else, the Institute insisted, the connection between high cholesterol and CHD held. The results of two large trials conducted by the NHLBI in the 1970s were similarly ambiguous for any connection, yet the Institute soon committed to a massive public-marketing campaign designed to convince Americans to lower cholesterol through reduced fat consumption (Taubes, Reference Taubes2007: 58).

Despite the stubborn evidence, the acquiescence of the medical and scientific communities was soon obtained, largely by ignoring anyone who raised a fuss against Keys’ hypothesis. The NIH held a “consensus conference” in December 1985 that ended the scientific debate. Three skeptics of Keys’ hypothesis presented testimony (out of 20 speakers), but their message was effectively silenced in the conference report. As Taubes (Reference Taubes2007: 59) puts the point: “The NIH Consensus Conference officially gave the appearance of unanimity where no unanimity existed.”

This brief history illustrates a case in which an undecided scientific controversy was determined by political means rather than through an unfettered PCR process. Between the McGovern committee's original pronouncement, the USDA's subsequent declaration of the scientific and nutritional wisdom of low-fat diets, and the NIH's gerrymandered consensus conference, the government effectively regulated the contents of dietary-scientific knowledge, such as it was, for several decades to follow.Footnote ¹⁸

7. The consequences of the government's interference in nutrition science

According to our epistemic theory of social order, the government's pronouncement of a nutritional consensus without recourse to the error identification and correction functions of a daylighting PCR process made the resulting scientific knowledge less actionable than would have otherwise been the case. Thus, according to our epistemic theory of social order, any plan of action that presupposed a proposition following from this politically perverted process was more prone to fail than an otherwise identical plan that presupposed no such proposition.

We can imagine this knowledge-by-fiat figuring among the presuppositions of many different kinds of plans, but, given our concern to test the epistemic theory of social order against the evidence as far as it will permit, two kinds of plan would seem most pertinent, namely, plans of scientific action and plans of (for lack of a better word) dietary action.

According to our theory, a plan that presupposed that Keys’ hypothesis would pass a particular experimental test without encountering falsifying evidence was more likely to fail than a plan that did not presuppose the scientific adequacy of the fat hypothesis. Simply put, if you expected the government's quasi-official scientific stance to pass your experimental test, you were likely to be disappointed.

Similarly, a plan that presupposed the nutritional adequacy of low-fat diets, e.g. a plan that presupposed that a low-fat diet would contribute to weight loss and minimize the risks of heart disease and diabetes, was more likely to fail than a plan that did not presuppose the nutritional adequacy of low-fat diets. If you expected the government's official dietary guidelines to make you svelte and lower your risk of heart disease, you were likely to be disappointed.