Heterogeneous construction of scientific knowledge and practice: III. Six Themes Drawn from the Reconstruction of the Kerang Study

The description of the building of the KFM [in the previous posts], although brief and partial, is sufficient to introduce six themes or propositions about the processes of making science and interpretation of those processes.  These themes are put forward in the spirit of theoretical exploration, that is, to highlight important issues and orient our thinking about them.  I begin with the observation that the different agents involved in the KFM draw on diverse components from a range of domains of social action (theme 1). Themes about scientists as imaginative agents who represent-engage (themes 2 and 3) extend the idea of social-personal-scientific correlations used in in my analysis of H.T. Odum’s pioneering contributions to systems ecology (Taylor 2005, chapter 3).  The other themes take further steps towards a more general framework for analyzing the mutual relationships between scientists’ representations and actions.

Theme 1.  Science-in-the-making depends on heterogeneous webs, not unitary correspondences.  From the description above, it is clear that diverse components were involved in building the KFM—from soil quality data to the terms of reference set by the Ministry.  Moreover, they were interconnected in practice, forming what I will call heterogeneous webs (see figure).  The KFM’s assumption that farmers were subordinate to economic rationality made it easier to concentrate only on options that took the form of government policy.  The power of the government to enact its decisions rendered investigation of how farmers change less relevant, which shaped the data that needed to be collected—generalized agronomic data, rather than sociological interviews, would suffice.  This, in turn, conditioned the relationships that could appear in the model.  Similarly, the modeler’s mediated relationship with the modeled situation and his geographical separation from the region rendered it less relevant to model the novel long-term options.  Their omission from the modeling, in turn, helped ensure that such aspects of the future reality would be less realizable, and the model’s account more real.

Figure.  A schematic picture of the web of diverse, interconnected components involved in the KFM socio-environmental modeling (from Taylor 1995a).  See text for discussion.

“Technical” considerations, such as the assumption of income optimization, and “social” considerations, such as the separation of the modeler from the farmers, had implications for each other in practice.  “Local” interactions were connected with activities at a distance.  For example, the modeler and the principal investigator decided not to pursue sociological inquiry into how farmers change, which meant that the content of and conduct of the survey of farms and farmers could remain unchanged.  No one component in the web stood alone in supporting the KFM as a representation of reality; in the actual process of building the model, technical components could not be detached from social ones, nor local ones from those that spanned levels.

In this sense I would say that science is constructed; science-in-the-making is an on-going process of building from diverse components, just as a house is built over time using plans and measurements, laborers and contracts, concrete and concrete mixers, wood and saws.  This is social construction, but not merely social construction—my interpretation is not that scientific knowledge is determined by or reflects the society in which it is made.  Although it is possible to say that the model reflected all the different social components, that would be stretching the metaphor of reflection.  Given the heterogeneity of components and their inter-linkage in an ongoing process, it is difficult and uninformative to collapse science-in-the-making to a unitary idea that scientific knowledge reflects society.  Likewise for the unitary idea that scientific knowledge reflects natural reality.  Science, in practice, is heterogeneously constructed.

Theme 2.  Scientists represent-engage.  In the process of building the model, the modeler, principal investigator, and other agents linked together technical and social components in order to make a model that worked for them.  These scientific agents tended to make the different components reinforce, not undermine, each other, rendering both the model and the ongoing scientific activity more difficult for others to oppose or modify in practice (see theme 1).  This insight goes beyond observing that representations of natural reality support engagements or interventions in different domains of society (Keller 1992, 74ff), or the claim that interventions provide the basis for scientific representations (Hacking 1983).  Through the model’s heterogeneous construction, representations and engagements were formed simultaneously, and, moreover, jointly.  Interaction between “technical” and “social” considerations fails to capture this relationship in which causes are inseparable (see theme 5).  Let me instead speak of scientists representing-engaging.

Theme 3.  Scientists are practically imaginative agents.  The idea of representing-engaging implies that scientific agents are mindful both of nature and of the social realms in which they act—in which they are situated—and that they project continuously between these realms.  In focusing on scientists’ social situatedness, I am not saying that they are corrupt, fallible, lazy, or taking the path of least resistance.  On the contrary, I am affirming that all human activity is imaginative, that is, the result of a labor process that grows out of the laborer’s imagination.  Agents assess, not necessarily explicitly, the practical constraints and facilitations of possible actions in advance of their acting (Robinson 1984).

In fantasy, people envisage worlds and mentally inhabit them, escaping the practical difficulties of action.  Imagination is not like that.  Achieving some result in the material world requires human agents to be engaged with materials, tools, and other people.  The KFM modeler had to engage with pasture growth, government sponsorship, an agricultural extension system, and so on.  Moreover, materials, tools, and other people confront scientists with their recalcitrance.  So scientists project themselves into possible engagements out in the world in order to imagine what will work easily for them and what will not.  These constant projected confrontations with the components that personal and collective histories make available lie behind all the actions people take, including scientists’ representing-engaging.  Through their imagined engagements people build up knowledge about their changing capabilities for acting in relation to the conditions in which they operate (though this knowledge may not be consciously articulated).

Theme 4.  The agency of heterogeneous constructors is distributed. One consequence of focusing on agents’ contingent and ongoing mobilizing of webs of materials, tools, and other people is that the character of their agency can be interpreted as distributed over those webs, not concentrated mentally inside socially autonomous units whose ideas or beliefs are key to the order they impart on the world.  That is, although agents work with mental representations of their worlds and can speak about motivations, the malleability of those representations and motivations is not a matter simply of changing beliefs or rationality.  Instead, a heterogeneous web of materials, tools, and other people help agents act as if the world were like their representations of it.  During the Kerang study, the principal investigator may well have believed deeply that economic decision-making was of primary importance in people’s lives.  However, he was able to sustain this belief against possible challenges through many practical measures (as discussed under themes 1-3).  For example, although he knew about the sociological study of how farmers change, he did not secure access to it, and he concentrated on econometric investigations rather than developing skills in multi-objective analysis.

Theme 5.  Resources are causes.  Up to this point in my description of how the KFM was constructed, I have used the neutral term component.  But there may be little explanatory significance to some of the diverse things that scientists link into the webs as they establish support their theories and ongoing scientific activity.  Let me apply the term resource for components that make a claim or a course of action more difficult for others to modify.  By extension, a resource for one person is a constraint for another person trying to modify the first’s claim or action.  Resources make a difference; that is, when resources are deployed they function as causes.  In this light, any descriptive use of the term resource also implies a claim about causes, and such claims invite analysis.

Theme 6.  Counterfactuals are valuable for exposing causes.  The components of the construction process I have chosen to mention were significant resources in the building of the KFM—or so my account of the KFM would imply.  But how can I support the causal claims that I have thus structured into my account of the KFM?  For a start, let me note that, to support the causal claim that something made a difference logically requires an idea of what else could have been if the resource in question had been absent.  That is, causal claims involve consideration of counterfactuals—things that might have occurred but did not.

The sources for ideas about what else could have been are varied.  Sociologists and historians of science undertake conceptual analysis or historical and cross-cultural comparisons (Harwood 2000), and give rein to their sociological imagination (Hughes 1971).  They also listen to opposing parties in controversies (Collins 1981a,b) and campaigns for social change (Nelkin 1984).  Indeed, controversies and campaigns provide the clearest, most concrete evidence of alternatives, because the agents themselves identify the resources they consider important.

There is no logical reason, however, why the resources explicitly exposed during a controversy constitute the full set used by a scientist.  There are resources taken for granted and shared by opposing parties and, moreover, resources that must be mobilized even when there is no apparent controversy.  In short, ideas of what else could have been should not be limited by whether anyone actually attempted to construct the alternative situation.  For all these reasons, explicit use of counterfactuals may be needed in order to analyze a more inclusive array of resources used in the construction of science.

I began my account of the building of the KFM as a fairly neutral description.  Notice, however, that I began introducing counterfactuals once I started to draw connections among the heterogeneous components.  For example, in contrast to a single objective of maximizing income in the modeled farms, I mentioned the counterfactual possibility of multi-objective techniques.  In explaining why this was not incorporated in the KFM, I mentioned that the principal investigator’s training, his status relative to the modeler, the Institute’s specialization, and the availability of software.  These were constraints for anyone who might want to construct a multi-objective model.  By identifying them I was implying that the principal investigator’s training and so on were resources for constructing a model with a single objective function.  In this general fashion, exploring the practical constraints on counterfactual possibilities—what did not happen—can, by a logic of inversion, expose the resources that helped those who constructed what did happen (figure 2).

Figure 2 The method of exposing resources by exploring the practical constraints on counterfactual possibilities.

The emphasis on multiple, heterogeneous resources means that the relevant counterfactuals are multiple and particular.  In principle, we could formulate some all-encompassing counterfactual.  For example, an alternative to the Kerang study would be a project that was not conducive to top-down government policymaking.  However, if we were to consider the practical implications of such a counterfactual, we would be challenged to identify specific sites for possible modification of the research.  This would be all the more the case if we focused on the practical implications for the specific scientific agents involved.  In the case of the Kerang study the modeler wanted to consider sociologically realistic processes of farmers changing.   But his ability to produce results that paid attention to such processes was constrained by his distance—geographically, organizationally, and conceptually—from the farmers’ domain of social action.  The geographical and organizational distance was, in turn, related to the centralized character of government and intellectual activities in the one major city of each Australian state, something given by the previous 200 years of development.  Towards the end of the project the modeler contemplated a move counter to that centralization, namely, to live and work in the Kerang region as an agricultural consultant.  He was aware that this would raise practical issues such as purchase and maintenance of a car, long-distance access to computer facilities and libraries, ways to keep abreast of discussions about the wider state of the rural economy, and other considerations of a more personal nature.  The modeler decided not to move which meant that the representation of the Kerang region he was able to produce facilitated the making of policy based on simple economic grounds.  This outcome did not flow from a political or intellectual commitment to the economically based technocratic rationality; many practical, not only intellectual or ideological, considerations would have been entailed in producing a different result.

Extracted from Taylor, P.J. (2005) Unruly Complexity: Ecology, Interpretation, Engagement (U. Chicago Press), chapter 4.


Collins, H. M. (Ed.) (1981a). “Knowledge and contingency.” Social Studies of Science 11: 3-158.

—— (1981b). “Stages in the empirical programme of relativism.” Social Studies of Science 11: 3-10.

Hacking, I. (1983). Representing and Intervening. Cambridge: Cambridge University Press.

Harwood, J. (2000). “National differences in academic cultures: science  in Germany and the United States between the world wars,” in C. Charle, J. Schriewer and P. Wagner (Eds.), Transnational Intellectual Networks and the Cultural Logics of Nations.  Oxford: Berghahn Books.

Hughes, E. C. (1971). The Sociological Eye. Chicago: Aldine Atherton.

Keller, E. F.  (1992). “Critical Silences in Scientific Discourse: Problems of Form and Re-form,” in Secrets of Life, Secrets of Death: Essays on Language, Gender and Science.  New York: Routledge, 73-92

Nelkin, D. (Ed. (1984). Controversy:  Politics of Technical Decisions. Beverly Hills, CA: Sage.

Robinson, S. (1984). “The Art of the Possible.” Radical Science Journal 15: 122-148.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s