Home > Book Summary > The Intellectual and Social Organization of the sciences (Book Review)

The Intellectual and Social Organization of the sciences (Book Review)


Richard Whitley gives a broad overview of the similarities and differences in the modern organization of the sciences. Crucially Whitley argues that m0dern science is unique in relying on reputational evaluation by peers. In short Whitley claims that science is built around the production of novel facts in ways which can generate future research.

Scientists operate in a field which Whitley defines as “The social contexts in which scientists develop distinctive competencies and research skills so that they make sense of their own actions in terms of these collective identities, goals and practices as mediated by the leaders of the employment organizations and other major social influences” (8)

The audience of scientific work is other scientists who use it to produce their own work. Therefore the ultimate aim is to produce work cited and read by the largest number possible.

“The modern sciences are distinguished then, by their commitment to producing novelty and innovations, on the one hand and their coordination of research procedures and strategies through collective appropriation and the use of their results, on the other hand.” (13) Novelty creation is spurred by rewarding scholars whose work influences that of others. Obviously then not all sciences have the same mechanisms. Every new result and task outcome in science must be different from preceding ones if it is to be regarded as contributing to knowledge. This means that a scientists training will constantly need updating or risk being out of date(22). Then modern science has high task uncertainty due to novelty but strong coordination of research due to the need for peer valuation for rewards.

Part of the development of this modern construction of science relied on developments in communication, such as an efficient and speedy postal system. Thus Whitley seems to argue that the modern form of science could not have existed in previous times. Additionally there was no labor market for scientific skills as there is in today’s university system. Scientists had money and time to spend and could therefore pursue any research they wanted, their rewards being less linked to peer evaluation. With time however as science was institutionalized the high degree of task uncertainty was ameliorated by setting an administrative hierarchy which sets overarching goals. This hierarchy has higher ranking scientists supervise and guard the rewards from the lower scientists. Finally there is also a peer review process by top scientists through what gets published. Therefore despite the uncertainty scientific work comes out as coherent through institutional mechanisms. There are differences between the sciences with the more Kuhnian ones being strictly controlled and the less mature ones giving greater freedom to individual researchers.

“Science is a professional work organization, it controls how work is carried out, how it is evaluated and its criteria and procedures govern access to material rewards… research is oriented to collective goals and purposes through the pursuit of public scientific reputations among a group of colleague competitors.” (25)

The high degree of work uncertainty cause by novelty allows a large amount of freedom in work methods by individual scientists. But since he or she requires collegiate approval this uncertainty whittles down in specific directions and methods for research. That is why when new approaches emerge it is easier for them to establish sub-fields rather than altering the whole original research program. A scientific field is established when it becomes prestigious and leads to rewards, it controls access to rewards through reputations, reduces uncertainty by controlling competence and performance standards, and it has its own language and forms of communication of goals and problems which reduces public participation and allows coordination and comparison.

“Essentially, modern sciences are systems of jointly controlled novelty production in which researchers have to make new contributions to knowledge in order to acquire reputations from particular groups of colleagues” (85)

Before the emergence of the university and the discipline structure, science had a much more fluid constituency and subject matter. Physicists, philosophers, mathematicians, and naturalists all worked interchangeably. The number of those working in science also was quite low and rarely did work that they could translate into one another’s language. “Knowledge tended to be less objectified in social structures and more connected to individuals than to organizations” Whitley calls this the gentlemen scientist model and puts it in opposition to the French model and the Prussian model. Ultimately it is the Prussian model which we have inherited today as they institutionalized a Fordian mass production of knowledge that outperformed the British model by brute force of numbers. The audiences for science were also more public, in general everyone was the audience, but this meant that the audience did not matter because the researcher could afford to ignore his audience. When someone gained reputational credit it was across the board, as a philosopher/physician/general scientists and it gave him authority over a wide range of subjects. The dominant unit of knowledge production in the 19th and 20th century however was the academic discipline as determined by training, employment and national and international reputations.

So in academic science (as opposed to state or industrial science) reputational control of goals and policies are high. Research has to conform to publicly agreed on goals, Employment follows the reputational hierarchy, and promotions are controlled by an elite who also control reputations. Scientists can work around this high control through specializations and focusing on new topics in which they may escape the strictures of the existing reputational hierarchies. This is even true for highly stratified sciences like Physics and economics. This can be done by transfering to other fields and bringing your own research to bear on it. Whitley describes this as colonizing. Obviously the amount of control differs by field depending on a number of factors.

Universities existed before they were institutions crucial for modern science. The first university in Bologna was more of an open forum, but it did not do research, it taught law and latin and paid little attention to producing new knowledge. But once modern sciences were institutionalized into universities it had a number of effects on the scientific craft. National curriculums began to emerge which ensured that the training of new scientists was uniform and extended to high school periods. Science became an activity requiring state support and gained legitimacy. Because Universities had been dominated by the humanities the incorporation of science pushed them into the direction of theory and “pure” science as opposed to the high degree of practicality that had existed. Universities became legitimate sources of certification. It also meant that pursuing new fields of research was much more expensive as an entire bureaucracy had  to be set up around, therefore science became less flexible. It was German universities which first incorporated science into the purposes of their universities and emphasized the production of novel innovations. By providing resources to science in universities, German science became industrialized, large numbers of new entrants and a large amount of funding became available with only general direction established by the state. “The sheer number of practitioners in the German universities coupled with their use of students on a large scale for research, which was in part subsidized by state support for laboratories, led to European science becoming dominated by  academic professionalized science” (60) Academic establishments are therefore quite powerful for the modern sciences because they ensure employment, but also because they control the production of knowledge by providing underlings.

The consequences of this were that scientists had to establish means to share research results (Leibig’s apparatus for combustion analysis meant he created chemistry) since there were so many new players and they were interconnected by their reputational credit science conformed towards common languages and methods. Hierarchies also emerged in universities and it povided a blueprint of how to control knowledge production. You also needed to use the man power at your disposal and techniques allowing new graduates to produce data quickly became dominant in an field especially since everyone had to contribute novel information.

The typology of modern sciences:

The degree of mutual dependence between scientists depends on. 1) the extent that scientists have to use the research results, methods, procedures, of their peers in order to conduct their own research in ways that are seen as competent and useful. this is called the Degree of Functional Dependence.

2) If researchers have to persuade their peers of the importance of their own research problem and approach to obtain a high reputation. This is called The Degree of Strategic dependence. High strategic dependence leads to intense competition and highly stratified hierarchical fields dominated by a few centers. Research has to prove meaningful to others.

These two are interconnected, usually when one is found in a high degree so is the other.

    Degree of Functional dependence
Degree of Strategic Dependence   Low High
Low Weakly bounded fields. Pursue variety of goals using variety of methods. Little coordination of results. (sociology) Specialist groups pursuing differentiated goals with specific standard procedures. Considerable coordination of results but little hierarchy. (US Maths)
High Strongly bounded research schools pursuing distinct goals with separate procedures. High coordination within schools but little within schools. (German philosophy before 1933) Specialist groups pursuing differentiated goals with specific standardized procedures. Large coordination of results and problems. Competition over which subfield is best in discipline very hierarchichal. (Physics)

Increasing mutual dependence fosters a sense of community and identity. It strengthens the borders between a field and the outside world and the language becomes less familiar to laymen. In academically dominated disciplines, a single organization controls the selection and training of recruits, the production of knowledge, definition of terms and proper procedures, evaluation and setting of goals and output. There is a very high dependence on colleagues. Locality becomes less important as mutual dependence grows, the field becomes more international.  High functional dependence pushes towards specialization and differentiation because of the high amount of competition. Researchers can no longer ignore one another’s work.

The degree of mutual dependence is affected by contextual factors as well. Whether external sources of employment in the private sector exist, How concentrated the resources are (Think CERN) and who the audience of your research is all affect the dependence. Other factors are the control of the communication system (journal editors), training programs, and jobs, if equipment is prohibitively costly. When these are concentrated there is high mutual dependence.

The second component of the typology of modern sciences is the task uncertainty involved in scientific research. This uncertainty comes from the ever changing nature of science in which novel findings not only build but displace old knowledge so that the scientist must be constantly alert. There are two types of uncertainty. Technical Task uncertainty which has to do with the techniques of research, and whether they are well understood and produce consistent outcomes (A thermometer in japan works the same as one in the US). High technical task uncertainty means the same result can be interpreted in many ways, methodology and procedures will vary from individual to individual. For example the development of a periodic table greatly decreased the uncertainty because the basic components were all agreed upon, additionally the amount of problems and ways of dealing with them decreases. The greater the control over the material under study (including definitions) the less the uncertainty. This obviously has an effect on the way novelty can be produced, whether by theoretical proliferation or methodological innovation etc… High technical task uncertainty implies high strategic uncertainty but not vice versa.

Strategic task unertainty has to do with the uncertainty around the hierarchy within the field. It is a more social uncertainty. This means that if a researcher knows what the field’s priorities are, what problems are the most prestigious, which method is the most favored, there will be low strategic task uncertainty. So there is large strategic uncertainty with large numbers of problems whose value shift with respect to the audience over time.

Uncertainty is affected by contextual factors, There is tension between remaining uncertain enough to produce new knowledge and certain enough to make that knowledge communicable. whether states want to exert control and direction,  if reputational control is high uncertainty is diminished by elites setting standards,  if means of intellectual production and dissemination are expensive and concentrated uncertainty will be lowered due to concentration of funds etc… finally if audience is diverse uncertainty is raised because you can get rewards from audience by writing on many different things.

    Degree of technical task uncertainty
Degree of Strategic Task Uncertainty   Low High
Low Considerable predictability, stability, and visibility of task outcomes. Results of experiments easy to understand. Problems and goals ordered. (Physics since 1930) Research can be communicated with unique disciplinary language. Specific audience. Limited control of raw material and phenomena under study. Results of studies hard to interpret in a consistent manner. Problems and goals are clear and known, no future problems will likely arise. (Econ since 1870) Local reputations more important. So problems are clear inside a country but no across borders.
High Task outcomes and studies have clear results which are uncontroversial to interpret. Large number of varied problems and goals with no clear hierarchy of which are more important. (Biology 1950) No consensus on outcome of experiments or procedures. Low degree of control over raw material. Problems also varied and diverse with no clear hierarchy of importance. (US sociology, Political Science?) Limits control of central establishment.

Putting the two together Whitley comes up with a typology covering all possible combinations of dependence and uncertainty which he showed in the first chapter to be the cornerstones of modern science.

          Degree of functional dependence
          Low High
          Degree of strategic dependence Degree of strategic dependence
          Low High Low High
Degree of technical task uncertainty Low Degree of strategic task uncertainty Low   1 2 3 4
High   5 6 7 8
High Degree of strategic task uncertainty Low   9 10 11 12
High   13 14 15 16

Combinations 1,2,5,6 are quite unlikely to occur. Low degrees of task uncertainty create high degrees of functional dependence. Similarly 11,12,15,16 are also unlikely since high degrees of task uncertainty yield lower functional dependence. Though this latter case is more possible to maintain if the social bonds in the group are strong.

Whitley actually argues that the sicences can be exhausted in 7 combinations:

3: technologically integrated bureaucracy, producing empirical specific knowledge. (2oth century chemistry). Standardized rules and methods are used to research different problems. Stable sub-fields provide theoretical common ground and reputational credit within. A high degree of specialization occurs without necessarily coordinating across schools.

4: Conceptually integrated bureaucracy, producing specific theoretically oriented knowledge. (post 1945 physics). Highly rule governed, very strongly bounded. Very formal control of research and high standardization of reporting with a strong hierarchy. Stable sub-units form but it is hard to create new sub-units, their hierarchy is determined by the central core which stops them from drifting too far apart. Reputations are therefore shared across the whole field and this leads to the highest amount of conflict.

7: Professional Adhocracy producing empirical specific knowledge. (bio-medical science) The tasks and methods are very specialized but the goals are not. There is no hierarchy of what the important questions are in the field. There is an abundance of funding and this ensures that conflict is avoided and prevents the formation of strong hierarchies, Some sub0units may emerge but they are weak. There is a strong mechanism to exchange information but it is not tightly controlled by any one group.

8: Polycentric profession, producing specific theoretically coordinated knowledge. (continental mathematics) This field forms some coordination over relevant problem areas and some coordination between specialists but across schools. There is a large degree of competition over setting the hierarchy of important problems and convincing others that these problems are important. Theoretical coordination is very important.

10: Partitioned Bureaucracy, producing analytical specific knowledge and ambiguous empirical knowledge. (Anglo saxon economics?) They seperate the theory which is dominant and uniform from the empirical application which is diverse. Work is controlled through a strong communication system and theory is highly valued, but empirical results are hard to give meaning to. Sub-units develop around empirical trends but they are all subservient to the main theoretical core.

13: Fragmented adhocracy producing diffuse discursive knowledge of commonsense objects. (Political Studies). Structure of the field is unmarked and changing, research is fluid and researchers are not interconnected, the large potential for conflict is mitigated by the low degree of interdependence. The role of the individual researcher is quite important here.

14: Polycentric oligarchy: producing diffuse locally coordinated knowledge. (continental european ecology). There are distinct schools dominated by reputational leaders. Within schools there is great mutual dependance which means that researchers try to coordinate and show that their research is valid inside schools. It is quite local and decentralized.

He argues this typology helps overcome the qualitative distinctions between psychology and economics and physics for example and helps us understand where they differ and why in way that overcome the subject matter they focus on.

Within each type there are differences in degrees of: Specialization and Standardization of task and materials, segmentation, hierarchization of sub-units, impersonality and formality of control procedure, degree of theoretical coordination, scope of conflict, and intensity of conflict. Readers can judge the varying degrees for each discipline but there is a table summarizing the relationship on page 169.

Additionally there are different contextual factors around each type of science. The variations occur in Reputational Autonomy over setting: Performance Standards, Significance standards, problem formulation and descriptive terms. Concentration of control of access to resource: Horizontal (Small number of agencies allocating funds and therefore determining research), Vertical (Decisions and rewards controlled by small group on top determining research for those under them). And Audience structure: Audience variety (more various less control) and audience equivalence (Audience are equals or users of information) encourages diversity of topics.

Relation to Randall Collins:

This is an interesting departure as it goes against Collins’s model in which he traces over 5 thousand years of intellectual production (though he doesn’t call it science) and locates common repetitions. Most importantly in Collins’ work is that he locates the limit of possible rival schools at any time being between 5-7 which he describes as the law of small number. (this coincides nicely with some psychological studies on human memory and attention span) with the minimum amount of intellectual schools being 3 and the maximum being 7 (or is it 5?)  Collins also argues that intellectual schools develop out of antagonism and debate, by creating conflict space an approach can gain attention. He also finds that these approaches proliferate during political pluralism and dynamism. Though collins does comment that the modern organization of the sciences may overcome the limit of 5-7 by specializing into many sub-fields he does not remove his argument that within these subfields conflict space is what generates scholarly attention. So Collins gives us the lesson that in order to grab intellectual credit one must generate conflict space.

Enter Whitley who seems to provide a more limited study (at least bound within the same epistemologically unified period)

The Good:

The Book provides a good summary of the organization and limitation on the production of scientific knowledge  in the contemporary world. It presents a good basic understanding of how science is organized and how different models of science exist in the world today.

It does not go into the pitfalls of other studies in using the term science to refer to everything but divides science into 7 categories.

Whitley provides us with summaries after the end of every chapter and he has some very good tables summarizing all the points he wants to make. And this was a godsend because…

The Bad

This book is boring. This book is beyond boring. I had to hire tiny midgets to punch me in the eyeballs with their tiny fists while I was reading so my brain wouldn’t combust from the sheer boredom that is this text. The author has obviously done much case study research but he rarely shows us or shares any anecdotes with us from the Sciences. This would lead to suspicion had he not been given some authority by his own scientific field. He could have seriously summed up his entire book in those tables I copied above.  But still I understand that books have to be a certain lengths these days to earn credibility, but why did it have to be so dry and with so little relaiton to the real world!

His definition of Science does expand the boundaries to include social sciences, but he does not explain why something like horoscopes could not count as a science in this model. It actually could. So there is a tension between his relying on a definition of science by looking at what exists as science today, and then his models including this which are not defined as science.

Categories: Book Summary
  1. No comments yet.
  1. No trackbacks yet.

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

%d bloggers like this: