Week of July 19, 2020

The Substance of Scientific Research: Reflections on David Bloor and the “Strong Programme”

Scientific theories commonly describe causation. The statement of one such, in consequence, will describe the effects expected to obtain given that a certain empirical observation has been made. Physical mechanics, in the crudest sense, describes the trajectory of an object based on an array of initial parameter values (mass, velocity, direction, etc.) that characterize the particle and the surrounding environment. The array describes the world at the outset, which is then transmuted mathematically into pertinent claims about the state of the world to come at the end of a defined period. In a sense, the formulae represent statements about the nature of reality—the manner in which particles behave in a vacuum. Abstract away the details and calculations, and one will find that most causal theories possess a similar form. Hampered by problems of induction and determination, these forms (lower-case ‘f’) lie buried in the Holy Land of modern scientific endeavor, waiting to be revealed—like the Holy Lance, or the Ark of the Covenant—by the intrepid explorer. One may quibble with such a realist and correspondence-based philosophy of science; nevertheless, that causal theories are claims about the nature and function of reality is broadly and almost tautologically accurate. Given that A has been observed, process B occurs such that outcome C obtains. A leads to C.

David Bloor, once a professor at the University of Edinburgh, insisted that the same logic be applied to theories themselves. The first principle of his “Strong Programme” in the Sociology of Scientific Knowledge, co-developed with Donald MacKenzie, was that it should be “causal, that is, concerned with the conditions that bring about belief or states of knowledge.” Theories about the world explain why various material states emerge; Bloor, in turn, sought to explain why theories themselves emerge as distinct, complex information structures. Such explanations include confirmatory empirical evidence, but also—by stark contrast with the traditionally triumphalist and individualistic narratives of scientific discovery—consist of social factors such as biases, networks, and incentives. The supposed veracity of the claim is here irrelevant, as, per Bloor’s second tenet (of four), the Strong Programme is “impartial” between true and false formulations. Beliefs may possess differing degrees of accuracy, longevity, and importance, but these characteristics have no bearing on the applicability of sociological methods—nor on the potential for insights on the nature of knowledge-formation. The elaboration and acceptance of a true theory is no more or less a valid subject for inquiry than a popular, pseudoscientific delusion. Indeed, both instances may be equally revealing with regard to the nature of evidentiary confirmation.

Crucially, symmetry must be preserved in methodology as well as study; true and false are deemed to arise as a result of “the same types of cause” (note: not the same causes), and can thus be studied from the same “naturalistic” social perspective. Classical historians of science (the poor, naive, nameless Whigs) focus on individual achievement and success, ascribing a vague inevitability to the eventual triumph of accurate, explanatorily-powerful theories and the fall of fallacies. Bloor demurs. Nothing in the nature of “true” belief necessitates success, and false belief is equally capable of sustaining adherence. The proximity of a statement to reality, even if it could be unequivocally detected, has no special or unique allure for the knower. Instead, an array of other factors—conventions, chiefly—make certain kinds of beliefs and their supporting evidence attractive and acceptable. Conventions regulate the language of a sound theory, the evidence (and sources) required for confirmation, and the potential explanatory scope. Agreements tacit and explicit determine the methods by which beliefs are generated and corroborated—and the very practice of corroboration is by nature a social one. “Without this co-ordination,” Bloor writes, “all we should have… would be persons who might be recognized as scientists, but something less than science.”

The view that convention underpins, stabilizes, and enables scientific research has been difficult to accept within the natural sciences themselves. The so-called “Science Wars” that followed the publication of Kuhn’s The Structure of Scientific Revolutions bitterly opposed realists and their cohorts of postmodern critics; the former group sought to defend verification as the time-honored and apparently fundamental anchor of progressive research, while the latter stressed evident epistemological gaps and the role of social activities in theoretical development. By reducing confirmation (and falsification) to mere components of the causal array behind scientific knowledge—defined provocatively in Knowledge and Social Imagery as “what is collectively endorsed” by practitioners—Bloor enraged the realist camp. Interestingly, he also drew criticism from the postmodernists, who deemed his “naturalistic” sociological approach overly ambitious in treating social factors. But most of the ire came from those who perceived the decentralizing and subjectivizing influence of Kuhn’s work, which Bloor approach supports, as a threat to the myth of scientific progress.

The chief implications of Bloor’s explicit relativism are in his “epistemological challenge” to the realist absolutism espoused by many scientists, observers, and commentators. Explicit logical positivism may have disintegrated (thanks to Kuhn, Popper, and Lakatos), but the notion of a research agenda that proceeds by unearthing and verifying discoveries through varied combinations of abstract reasoning and empirical evidence remains attractive—and widely (if tacitly) endorsed. If conventions, not material reality, are responsible for agreement in science, the objectivity of the resulting theories would be dangerously suspect. Given the perceived importance of scientific explanation in anchoring the larger corpus of human knowledge, the kinds of social causes offered by Bloor can appear threatening, if not completely destabilizing. Even if conventions can maintain Mertonian norms or defend theories from invalid critiques, the influence of collective human judgment in the process inserts arbitrary choice. Inevitably, this will lead to the opposite tendency—as Pierre Duhem noted, to mobilize consensus and mutable standards for the protection of inadequate claims from testing.

Whether conventions are inimical to the sciences, however, is beside the point: Bloor’s most convincing argument is that it is a “ubiquitous feature of all systems of belief.” Separating theories from creators and interpretation from theory is, in most common-sense accounts of research practice, impossible in all but a semantic sense. Agreement and choice are universal features at all levels of collaboration and are required to convert the impressions of an individual into the knowledge of a collective. Clearly expressed, many of Bloor’s claims are more common-sensical than radical:

Concept application, and rule following, must be thought of as a move from case to case, where every such move, in principle, calls for an exercise of the analyst’s curiosity… Furthermore, no two cases which fall under the same concept are identical. Or, to express the point more fully, whenever things are treated as identical that identity be seen as (a) a theoretical claim, and hence (b) one which needs to be collectively sustained by a group of concept users. Alternative employments of the concept of identity always lurk in the background.  Even if it is overwhelmingly natural or easy to see certain items as identical, that ease and naturalness is also but one contingent factor in the situation. Other considerations could (and often do) over-ride such impressions.

Bloor does not disregard the role that evidence plays in the process of theory-elaboration; rather, social and material factors coexist—and are practically indistinguishable—in the set of causes. Experimental evidence, for instance, can confirm or reject a proposition. The conditions for confirmation and the content of the evidence, however, are necessarily mediated by conventionalization, even if the conclusions appear to be obvious. As Bloor notes, “inferences cannot be inductive without being conventional.”

Concepts which are meant refer to objects in the material world must be such that they can be used rightly or wrongly… The normative aspects of concept application are consensual. Standards of right and wrong derive from agreement in use. They only exist in and through that use. To exist at all standards must be invoked, cited, employed, referred to, challenged and defended… The correct use of that precedent, that is, applying a concept in a way that is consistent with its meaning, is correct because a group of users agree that h is correct. Without the sociological machinery of interaction there would be no normativity, and without normativity there would be no conceptual content, that is, no meaning. Meaning itself is unintelligible except as a sociological phenomenon.

To deny that normativity originates in some form of consensus (with or without outside influence) is to invite regress. Eventually, the adoption of a certain standard requires explanation without reference to a higher criterion. Norms are not completely flexible; “cognitive adaption” to the material environment will ultimately privilege some sets over others. Modern scientific values arose and persist in large part because they serve the purposes of users more effectively than, say, more dogmatic or less critical systems. Reality subjects norm-complexes to the test of efficacy, whereupon the role of the user is to accept the result. Failures and inaccuracies make these choices all but inevitable.

Two questions remain: whether a convention-based account of scientific research is avoidable and damaging, and whether social factors can be treated as unproblematic in the way that unadulterated evidence (read: pure representation) cannot. Common sense and the history of science suggest human agreement can only be ignored in a semantic sense, rendering concerns about the insertion of subjectivity mere wish-thinking. Science, as a body of knowledge, is perpetually expanded and reified through discussion and interpretation. A “view from nowhere” would be useful, no doubt, but no description of what this would consist of stands up to much scrutiny. How would we, as essentially socialized beings, imagine a perspective denuded of such inculcation? A more interesting objection involves the extent to which similar issues affect the sociological analysis itself. By Bloor’s fourth tenet, reflexivity, the Strong Programme itself is to be viewed in the context of causation, impartiality, and symmetry—so presumably the same issues said to dog the scientific theories under study can be found in any sociological ones compiled about them. Subjectivity, under- and overdetermination, and theory-dependence, for instance, are if anything even more pervasive issues in the social sciences than the natural. The classification of entities within and across civilizations is famously convention-laden and discretionary; witness the futile efforts to agree on “capitalist” and “socialist” states, or to establish historical periodizations (e.g. “long centuries”). Bloor’s approach is “concrete” and “naturalistic” only insofar as social sciences are—a presumption at best problematic, and at worst ludicrous.

Nevertheless, founding traditionally abstract and a priori epistemological debates on a historical and empirical basis holds some promise for exceeding the bounds of long-stagnant philosophical debates. Social realities, though fraught, are more solid grounds for analysis than the mental gymnastics involved in theories of correspondence and representation—and, moreover, better reflections of how science seems to work. Bloor’s version of constructivism does not imply either complete arbitrariness or falsehood, but even if it did, one must accept that convention is ever-present in the scientific story—establishing and safeguarding the very parameters of “correct” work that realists sought to protect. Science is and has always been an endeavor practiced by human collaboration, and cannot help but reflect that fact.

Links

The Dollar and Empire, by Herman Mark Schwartz

Schwartz claims—in response to arguments that the “exorbitant privilege” of owning the global reserve currency actually imposes an “exorbitant burden”—that the negative effects of the dollar-funded current account deficit are counter-balanced by benefits to geopolitical power. Control over the dollar buttresses U.S. military spending, binds together American alliances, and permits domestic financial institutions to favorably manipulate the world’s monetary system.

July Isn’t the Same Without a Tour de France, by Jason Gay

Sport-watching is a profoundly ritualistic activity, a self-revelation made yet more stark by the early-morning absence of cycling’s most storied race from the July television program. Gay’s reflections on the aesthetic accoutrements of the Tour—the views and sounds that make it uniquely meaningful—are worth reading even for non-cycling fans, if only to better understand why life has become so strange.

The $52 Trillion Bubble: China Grapples With Epic Property Boom, by Stella Yifan Xie and Mike Bird

China has been experiencing a real estate bubble of truly staggering proportions over the past few years, one which may be exacerbated by safe asset demand and currency fears amid the global economic malaise. As buyers over-leverage, sinking an incredible 78% of their wealth into residential properties, the Chinese government has now found itself guaranteeing a boom that has become too big to fail. Strongly recommended.

The Best Books on the Reformation, by Peter Marshall

Both the discussion and the selections are excellent in this FiveBooks interview, which compellingly illustrates a far less dogmatic image of Martin Luther than is traditional and stresses the complexity of the Catholicism under siege. Marshall also argues that the Reformation should be seen less as an abrupt rejection of clerical doctrine than an extension of a longer-term evolutionary process underway within the Church.