Scientific knowledge and sociological theory

by Barry Barnes

Originally published in 1974.Scientific Knowledge and Sociological Theory centres on the problem of explaining the manifest variety and contrast in the beliefs about nature held in different groups and societies. It maintains that the sociologist should treat all beliefs symmetrically and must investigate and account for allegedly "correct" or "scientific" beliefs just as he would "incorrect" or "unscientific" ones. From this basic position a study of scientific beliefs is constructed. The sociological interest of such beliefs is illustrated and a sociological perspective upon scientific change is developed.

Genres: Sociology

14 pages, Hardcover

First published June 28, 1974

About the author

S. Barry Barnes was Professor of Sociology at the University of Exeter. He is known for his pioneering work on the sociological study of knowledge generation and evaluation in science, and on the credibility of scientific expertise.

Reviews

[Sharad Pandian · Rating 5 out of 5]

One of the founding texts of the Sociology of Scientific Knowledge, along with David Bloor's Knowledge and Social Imagery. What is striking to me is how modest the methods prescribed here are, and the ways they're mangled when described by others outside the field.

At it's heart is the call to replace the usual asymmetric method of explanation, where contemporary scientific truth doesn't need explanation but deviation from it does, with a symmetric one where all beliefs need causes. In practise, this would first identify a baseline of normality for communities and then identify deviations from it, explaining both the mechanisms of the transmission of "normal" beliefs and the deviations:

The manifest variability in institutionalized natural beliefs is to be made intelligible by being set against an unproblematic baseline of normality, not 'truth' or rationality. It is possible for the sociologist to identify normal patterns of belief by the investigation of human collectivities in the light of existing sociological theory.

The sociologist will associate an actor with certain institutionalized beliefs and actions by virtue of the position he occupies within the social structure. For an explanation of their acquisition he must look to the theory of socialization, and studies of the process of cultural transmission. To account for their persistence, he must look to the stability of the context in which they occur. To account for their change, he must ask what mechanisms, if any, can produce cultural change in a stable context, and he must identify any special influences and causes which disturb that context and modify elements within it which have served to sustain particular beliefs. Where actors or groups hold idiosyncratic beliefs as judged against some background of normality, some special cause or condition must be identified which distinguishes the actor or group from that background. (41-43)

The notion of cause employed here is drawn from MacIntyre, and not Hume's "cause as constant conjunction" (71). This alternative, which is "far more widespread and frequently employed" concerns the explanation of deviations from an accepted baseline:

This is the designation of one of a set of necessary conditions as a cause, particularly when that condition is of special interest to us. The implication is that, had the necessary condition been absent (or different), the caused event would not have occurred (or would have occurred differently), given that there was no change in the other necessary conditions. This is an entirely different conception of causality from the Humean kind. As MacIntyre points out, citation of an ice-patch as 'the cause' of a road accident implies neither that 'whenever ice-patches occur there is an accident', nor that 'there is never an accident unless there is an ice-patch'.

Suppose we cite metal fatigue as the cause of an air crash. The implication is that in other relevant respects the aeroplane was normal, how we would expect it to be. Had the metal been likewise in its normal condition, no crash would have occurred. The deviation from normality explains another deviation from normality. The background of normality forms the set of unchanged necessary conditions against which the causal story stands out and operates as an intelligible communication. The cause is a necessary condition in which we are interested; in labelling it, we define a taken for granted background of normality in terms of which our utterance makes sense. (71-2)

This is particularly helpful, because while even Bloor calls for the use of causal explanations in his book, he never really specifies what that means. Barnes then does a lot of the heavy lifting that gets used in the SSK programme. Note that for Barnes, there's no sharp divide between causes and reasons in explanations - "no logical or a priori incompatibility between reasons and causes as explanations of why actions occur or beliefs are adopted; sometimes reasons may be causes" (73).

His emphasis that the transmission of scientific culture isn't automatic, and instead relies on building credibility, anticipates the eventual sociological-historical work of Steven Shapin:

Science enjoys no advantage because its beliefs are in unique correspondence with reality or uniquely rational, hence its processes of cultural transmission will be in no important respect different to those employed by other knowledge sources. Like the prophet, the astrologer and the witch-doctor, the teacher of science will have to deal with the problem of his own credibility; he is faced with the task of transmitting lore.

...within science itself, devices to maximize credibility, and lubricate the mechanisms of culture transmission, have become institutionalized, and remain of interest; many of them have been discussed by Kuhn. A particularly clear-cut instance, although it is no longer in widespread use, is the historical cameo found introducing the real substance of a chapter in a science textbook. This presents, in a page or two, an account of how alternative theories to those about to be presented were refuted in the past, and occasionally mentions a 'crucial experiment' which established the present position as valid... More generally and Significantly, one can point to the selection of successful areas, and the suppression of difficulties and criticisms, in the transmission of models and theories [particularly to students]. (64-5)

One approach used throughout is to emphasise that science isn't about simply collecting objective, context-independent facts - there isn't a neutral observation language that would allow you to do that. Additionally, theories are essential to science and these do not arise "inductively from experience", they do not arise from experience alone (10). He argues that theories move from metaphors to being institutionalized, and demonstrates this using Dalton's Atomic Theory as an example (49-57) and later Carnot's reliance on the properties of concrete engines (120). This argument about what theories are matters for him, because "To show the metaphorical nature of thought is to show the culture bound nature of thought" (57). For him,

there are two ways in which the initial selection or construction of a model, and its institutionalization, may be related to the social context. First, the stock of available cultural resources is a function of the milieu and the range of actors' experience within it. Second, what counts as a part of accepted knowledge or an accepted standard of judgment will again depend on the milieu, and may depend on actors' social roles, and the concerns and interests of the groups to which they belong. It is impossible to generalize about the importance of these latter specific factors; their influence is mediated by too many features of the particular contexts involved (146-7)

However, he's not simply calling for a broader intellectual history which simply accepts cultural ideas as well as narrower theoretical ones, since this would still be silent about "why certain ideas come into fashion" (115). Instead,

A better approach is to regard ideas as tools with which social groups may seek to achieve their purposes in particular situations. Ideas suit purposes not because of any logical relationship, but because they are naturally suited to particular kinds of use within an existing system of beliefs and norms. (116)

He's quite clear that he's not advocating a mathod of making "a one-to-one correspondence between a belief, or set of beliefs, and a particular type of social class or interest group." Instead "ideas are related to social structure by examining the perceived situation of actors in particular collectivities, and their perceived problems and aims." Note both collectivities and attention to their perceived problems. He illustrates using the example of Forman's analysis of the "opposition to determinism" (116).

Note: As a consequence, recognizing that cultural influences and interests are important to Science means that labelling/criticizing ideology cannot simply consist in pointing to the social origins or functions of certain beliefs - this is likely ubiquitious. (130)

He clarifies that these kinds are explanation he's arguing for are not meant to be replacements of agent's accounts, but suplements to understand broad scientific-cultural patterns:

Since the danger of imposing external perspectives upon the diversity of scientific practice has been stressed, and the importance of proceeding from the actors' own definition of the situation, the present discussion clearly stands in need of justification.

This can be provided if the discussion is thought of as an account of cultural change in science, rather than as a description of every individual move an accredited scientist makes. Despite the existence of positivists and positivistic activity in science, all research traditions develop their beliefs, and culture generally, through the deployment of metaphors; long term cultural changes are metaphorical extensions, or changes of metaphor. Actors who make these extensions may not talk of them in these terms, but the present account does not replace actors' own accounts or necessarily devalue them; rather it draws attention to general features of scientific belief and action, as subjectively defined, which are of particular sociological interest. (53)

He's very sensitive to science's specialized nature, and only moves to talk of "external" factors in Chapter 5. For him,

we [analysts] should not seek to define science ourselves; we must seek to discover it as a segment of culture already defined by actors themselves. (This is not to imply that science can be 'found' as a pure phenomenon; it means merely that someone with a sociological perspective should treat it as an actors' category.) (100)

This deference to actor-categories means that

If the complicity of unacknowledged, illegitimate causes in these processes is to be suggested, it is first of all necessary to establish the insufficiency of the accounts of the actors themselves (140)

[For example] If one considers the theories of human heredity, evolution and eugenics which grew in Britain from the time of Galton until, say, 1914, it is difficult to understand them as nothing but attempts to describe and predict phenomena, or to apply a biological theory in a new area. Actors' accounts of their own 'scientific' activities fail to convince in some respects. The taken for granted assumptions which guided work do not appear natural in the light of accepted knowledge or legitimate aesthetic considerations; comparisons of different social classes are curiously confident and uniform, given how underdetermined they are by the indications of the situation. By attributing some significance to unacknowledged functions, the same theories become more intelligible, naturally appealing, and, one might even loosely say, logical. This is the case, for example, if one holds that the theories were, in part, developed as reassuring fables for members of a declining social class. Similarly, early eugenic calculations and prescriptions may, in many cases, prove more intelligible as responses to the practical problem of eliminating the inconvenience or 'threat' of the London destitute. Since these unacknowledged functions, besides making sense of the theoretical innovations, help to account for the social distribution of their credibility, and reflect the everyday (if formally irrelevant) preoccupations of eugenists and social Darwinists, there is good evidence for their causal role. (144)

What appears natural isn't an infallible guide but a method employed with effort: "The use of empathy is the only way we presently possess of investigating whether the factors cited by an actor with a given background would in fact suffice to generate a natural propensity for him to construct or accept an unprecedented innovation" (141).

Barnes wants to make a firm distinction between sociology and normative epistemology, arguing that

the sociological task of causal explanation is held to be separable from epistemological and evaluative concerns. (131)

If such [social] influences are found, that does not imply that the beliefs involved are erroneous, or 'unscientific' in some context-independent sense. (131)

However, he admits that "theories of belief suggest theories of knowledge, even though they do not logically imply them" (153). For one,

Where normal practice exists, social influences may be identified which cause deviations from that practice… On occasion. causal explanation of this kind may readily be employed in criticism. It may indicate that beliefs are being advocated in bad faith. or that an unacknowledged and illegitimate factor has partially determined scientific work. (139)

In addition, we can "ask with which ontological and epistemological positions the present discussion of belief has the most natural affinity" (153). The adoption, credibility, and fruitfulness of the book's assumptions and methods suggest a certain epistemology in two ways:

[First,] a materialistic or naturalistic ontology has been assumed and advocated. Similarly with epistemology, the implication has been that knowledge of how we know may be acquired by unselfconscious scientific or 'empirical' study, and that there should be no hermetically sealed off, autonomous area of 'conceptual' philosophical inquiry lying 'under' or 'beside' this.

[Second,] In arguing that all belief systems must be treated symmetrically for purposes of sociological explanation, many traditional ways of justifying belief as knowledge were incidentally undermined. It transpired that one perspective can only be shown to be preferable to another in expedient terms, which means that the perspective adopted in this volume is itself a contingent one. Thus, the epistemological message of the work could be said to be sceptical, or relativistic. It is sceptical since it suggests that no arguments will ever be available which could establish a particular epistemology or ontology as ultimately correct. It is relativistic because it suggests that belief systems cannot be objectively ranked in terms of their proximity to reality or their rationality. This is not to say that practical choices between belief systems are at all difficult to make, or that I myself am not clear as to my own preferences. It is merely that the extent to which such preferences can be justified, or made compelling to others, is limited. (154)

[Harry · Rating 4 out of 5]

This was pretty good. All chapters were thought provoking, carefully written, and made use of well-chosen examples. I also liked how modest Barnes' main thesis is: that all belief systems must be treated "symmetrically" for the purposes of sociological explanation.

It's also worth knowing your Kuhn well too as it feels like Barnes references him, normal science, puzzle solving, or the Copernican revolution every four pages.