Social Research Glossary

Social Research Glossary

A B C D E F G H I J K L M N O P Q R S T U V W X Y Z Home

Citation reference: Harvey, L., 2012-24, Social Research Glossary, Quality Research International, http://www.qualityresearchinternational.com/socialresearch/

This is a dynamic glossary and the author would welcome any e-mail suggestions for additions or amendments. Page updated 8 January, 2024 , © Lee Harvey 2012–2024.

_________________________________________________________________

Network

core definition

A network is an interrelated grouping of entities.

explanatory context

Introduction

The friendship patterns of a group of children, the information flow in a large organisation, the communication structure between researchers in a discipline are all examples of networks.

Networks have particular referents in a number of spheres, these include sociometry, network analysis and metascience.

Sociometric networks

See sociometry

Network analysis

Network analysis refers to a process that views interacting groups of people (families, freindship groups, work groups) as networks. Network analysis focuses systematically on interrelationships, interactions and interdependencies within subject groups thus linking interpresonal relations to institutions.

The kinds of things network analysts do is to look at the interpersonal links to see whether they are single or multiple; if there is any transactional content in the relationship; and if so whther it is uni-directional or bi-directional; the frequency and duration of interaction. Network analysis is also concerned with what it calls structural criteria; the size, density, centrality and clustering of networks.

Network analysis grew out of, and partly in reaction to, structural functionalism, which posited the self-regulatory nature of institutions. Network analysis allowed for a systematic analysis of the ways that interacting groups could effect institutions of which they were a part.

Network analysis was popular in anthropology in the 1940s, was used in sociology in the 1950s, had a brief resurgence in the 1970s.

Networks as metascientific units

Network as a metascientific concept refers to the explicit attempt to provide a model of the development and structure of researchers operating with a Kuhnian paradigm.

Networks are, in effect, an extension of the invisible college concept and attempt to broaden the base of the idea of research schools.

A network approach to the development of research communities usually suggests a four-stage development of a research group/ discipline. The stages are from a rather nebulous grouping, through the generation of a network of communicating researchers, to more specific clusters of co-operating researchers, to a speciality or proto-discipline.

Central to the stages of development is the communication structure as critical appraisal of any work is primarily through verbal, direct and immediate discussion in a circle of understanding colleagues. These communicating structures progress from serious discussion, through co-authorship to apprenticeship and colleagueship.

This model operates as follows. Paradigm development occurs when a group of persons experience a 'gestalt shift'. The subsequent research of the persons involved utilises this new perspective and puzzle-solving activity, characteristic of normal science, begins. The emergence of a paradigm is located within the nebulous communicating group.

The second stage, the emergence of a network, involves communication between pairs and triads of scientists engaged in regular communication over a period of time. These patterns, however, lack constancy. Changes in the network and its functioning are dependent upon personal contacts made by scientists. The initial network is augmented by the adoption of students and a general 'thickening' of the research group. The later stages of a network come to resemble a 'school' when successful 'network' groups make an explicit agreement regarding the style and content of work to be done, which appears in a programme statement.

The third stage, the development of a cluster is a more reflective stage than the previous ones. A cluster is formed when scientists become self conscious about the communication patterns and begin to delineate 'problem groups'. Such groups are scientists who are working on common problems and have developed from the earlier pairs and triads through recombinations. Clusters start to generate a distinct culture and draw support and recruit students.

The transistion from cluster to speciality stage is a logistic progression. Students become successful in their own right, the new orientation, as it blossoms, cannot be contained as a cluster at one institution. The successful cluster becomes dispersed.

In short, then, success depends upon whether the coherent group has a research programme. Most groups are not successful.

analytical review

associated issues

Analysis of networks as a metascientific units

Introduction

Mullins (1973) has elaborated an approach to analysing the development of scientific communities that he hopes will lead to an understanding of the way in which different kinds of theory come to be written. This he expects will be achieved through a reassessment of Price's notion of invisible college within a Kuhnian paradigmatic framework. His work in the area began with his analysis of the phage group of microbiologists (Mullins, 1968) and on the basis of this enquiry he posited a network approach involving a multi-stage model. In his later work Mullins, (1973) he attempted to apply this model to American sociology, asking why social theory came to be so variable and how some social theories became established and why all, eventually die out. Mullins approach is an extension of invisible college analysis; the invisible college (under a different name) is, in effect, but one stage in the growth and development of research programmes into disciplines.

Applied to sociology it constitutes, besides a purported basis for understanding the process of the production of sociological knowledge, a critique of the way the history of sociology has been presented. Mullins suggests that history of sociology has been written in one of four ways, as history of core concepts (Nisbet, 1966), as an intellectual history (Bendix 1966, Mitzman, 1970), as biography (Demerath & Peterson, 1967: Martindale, 1960) and as schools of sociological thought (Faris, 1967). The first two reflect an emphasis on ideas the last two on persons. The biography and core concepts approaches being directed at a unit level and the intellectual history and schools approaches at a system level.

Mullins is critical of all. Core concepts approach fails, he maintains because it emphasises current ideas of concepts rather than historically specific meanings and attempts to recreate the cumulative development of theories. The biographical approach, like the 'great man' theory of political history, entirely ignores the social milieu, presenting history as an individual endeavour. The intellectual history is, for Mullins, nothing more than a collection of biographies that fail to adequately establish 'traditions' nor assess their origins or developmental processes. This latter is what Mullins attempts to do by linking intellectual history with the fourth and potentially most potent approach from his point of view, that of schools. However, as it stands, the writing of the history of sociology as a history of schools is inadequate because 'it is largely the product of a socially and psychologically informed philosophy of science which emphasises the importance of early training and the acquisition of paradigms.' (Mullins, 1973, p.11) Mullins is not criticising this approach for its use of a paradigmatic model nor for its concentration on early training, rather he is concerned that the 'paradigm' is taken-for-granted and not examined and that the processes by which theories are generated, indeed the very content of theories, are ignored in favour of interpersonal relations.

‘The school approach is widely used by geneologists of science who wish to discover a disciplines roots. Its strength is on the social context of theory. Its weakness is that the actual product of science, the theories themselves, are not discussed. Moreover, this approach usually fails to explain how the same training (for example, by Talcott Parsons at Harvard during the early 1950s) can produce such different persons as Harold Garfinkel (an ethnomethodologist) and Robert Bellah (a structural functionalist).' (Mullins, 1973, p.11).

While critical of such 'histories', Mullins does not provide a critique of historism. His is still an explanatory approach and not an interpretive one, as will become apparent below. Mullins criticism of 'history' of social science writing provides the introduction for the presentation of networks approach to sociology. His criticism may be seen less as a critique of nomothetic metascience and more as a legitimating context for his network scheme, already developed in his earlier work on the phage group. Mullins notes,

'Each of these four approaches to analysing social theory offers a means to order knowledge, but each has particular weaknesses. As we have noted, the exhaustive, open quality of these systems indicates that each is only a classification system rather than a theory itself. Being open, they can fit all possible future theories into their framework: being exhaustive, they include all theories and thus have no adequate explanation for (no way to predict) the rise and fall of different theoretical systems.' (Mullins, 1973, p.11).

Mullin's model

In his article of 1972 Mullins draws together his work on the phage group of molecular biologists and offers a model of the growth of this speciality (Mullins, 1972). The model is clearly designed as a generalisable model, although the extent of its generalisability is merely inferred from a few unsystematic supportive illustrations. Mullins is concerned to show how the phage group were an integral part of the development of molecular biology and that the speciality itself emerged within a paradigmatic framework as a result of the development of social interaction of scientists.

Kuhn's model of the development of science involving the cycle of normal science through crisis and revolution to a new paradigm is purported to underpin Mullins' approach. Mullins simply elaborates the social organisation that permits this mechanism to operate. However it is not always clear that the two models are compatible, as will be explained below.

There is no doubting Mullins' development of the invisible college thesis. Mullins offers a four-stage model, each stage being marked by empirically demonstrable social and intellectual characteristics. The model involves a progression from a 'paradigm group' (later relabelled the 'normal stage') through a network stage, then a cluster stage to a speciality (or discipline). Central to the stages of development is the communication structure, for, in accord with Medvedev, Mullins posits that critical appraisal of any work is primarily through 'verbal, direct and immediate discussion in a circle of understanding colleagues' (Medvedev, 1971, pp 133–34) and that such appraisal is at the root of the development of knowledge.

The stages outlined by Mullins are each characterised by various types of communication structures and the pattern of such types is indicative of the stages of the social interaction.

These component communication structures are:

a) communication (i.e. serious discussion about current research, unrestricted by institutional, status or disciplinary boundaries)

b) co-authorship (i.e. two scientists jointly reporting research results)

c) apprenticeship (students trained and sponsored by a teacher)

d) colleagueship ( two scientists working together in the same laboratory).

This model, Mullins expands as follows. Paradigm development for Mullins (interpreting Kuhn, 1962b, 1969) occurs when a group of persons experience a 'gestalt shift'. The subsequent research of the persons involved utilises this new perspective, and such a shift is justifiable only if success in problem solving can be established and new problems pointed out. Further, success, Mullins argues, may eventually enable the new approach to 'establish itself as normal science in Kuhn's sense'. Thus the puzzle solving activity, characteristic of normal science, may begin.

The emergence of a paradigm (albeit somewhat loosely defined) is located within the paradigm group that he describes as follows.

'A paradigm group is the minimal form, of scientific group. Its members have no necessary social connections. Kuhn indicates that any useful paradigm must, by definition, be the possession of some social group which is using it.... The minimal requirement of such an entity is two or more established scientists who have shifted from one viewpoint to another (Gestalt switch), and who might or might not be in communication with one another. A paradigm group is thus a set of individuals, all of whom have moved into a similar cognitive situation with respect to the same, or similar, problems.' (Mullins, 1972, pp. 54–55)

This normal stage is, in Mullin's view, somewhat directionless, almost anomic.

'The normal stage is characterized by a low degree of organization both within the literature and within social relationships... no groups of students are being formed. The commitments of persons involved in normal areas are generally short in duration and constitute only one of several commitments, each to different areas. Hence little co-ordinated effort is made to solve any particular problem.' (Mullins, 1973, p. 21)

The normal stage then, is clearly not a puzzle-solving stage and quite different from the 'normal' process of science that Kuhn talks about. In fact, Mullins sees the normal stage as an almost dormant period just waiting to be shaken up by the emergence of a coherent group. The major shift, the gestalt switch, which leads researchers to a new paradigm and new puzzles, operates immediately for Kuhn and while it may be augmented by group interaction, normal science is not a period waiting for the onslaught of a coherent group. Mullins’ normal stage is thus closer to Kuhn's pre-paradigmatic stage.

The second stage, the emergence of a network reflects the emergence of an invisible college. At this stage Mullins sees a tendency for small group co-operative research sustained by the publication of successful findings. This stage in his model comes close to the invisible college idea. However, Mullins' network stage is less extensive than the Price-Crane construct.

The communication network consists of pairs and triads of scientists engaged in regular communication over a period of time. These patterns, however, lack constancy and changes have imperceptible effect on the science (i.e. the broad area within which the network is operating).

Changes in the network and its functioning are dependent upon personal contacts made by scientists (there being little conscious idea that a scientist is part of a specific network and less that it may be a nascent speciality).

Mullins argues that the network stage emerges from the normal stage when a group of 'likeminded' researchers gather round a particular intellectual product and then begin to evolve 'overlapping' invisible colleges, work together and recruit one another. Without some research breakthrough at this stage no further development is likely. Alluding to the invisible college model, Mullins suggests that:

'In a group of scientists writing on the same very specific problem area, some of them might have all their contacts within the group, others might have their contacts within and without the group, while others who are clearly working on the same problem as these scientists already mentioned, might not be connected with any of the other groups. These contacts might include any of the activities from communication through co-authorship and colleagueship to apprenticeship. We should note that the communication network structure shows two changes from the paradigm group structure (1) increased connection among scientists who are working in the area, and (2) a corresponding decrease in disconnected or independent persons.' (Mullins, 1973, pp 58–59).

The initial network (invisible college) is augmented by the adoption of students and a general 'thickening' of the research group. The later stages of a network come to resemble a 'school' when successful 'network' groups make an explicit agreement regarding the style and content of work to be done, which appears in a programme statement.

The third stage, the development of a cluster is a more reflective stage than the previous ones. A cluster is formed when scientists become self-conscious about the communication patterns and begin to delineate 'problem groups'. Such groups are scientists who are working on common problems and have developed from the earlier pairs and triads through recombinations. This development is encumbent upon the existence of favourable conditions that Mullins identifies as, for example, leadership, supporting institutions, luck and substantial research problems. These clusters are often identified by name by the members and by scientists outside their boundaries. This is indicative of the more stable nature of the clusters than of the pairs and triads that constituted them. Clusters start to generate a distinct culture and draw support and recruit students.

'Communication becomes even more rigorous. Clusters of students and colleagues form around the key figures in a group in one or a few institutions. Students are important because only a few scientists in a field ever have any graduate students, and those who do usually perform most of the research and publish frequently as well. (See Price, 1963).' (Mullins, 1973, p. 23).

Mullins goes on to suggest that a cluster normally includes three or more Ph.Ds who reinforce each other's interests along with some graduate students and that this stage constitutes the first real institutionalisation of the research area. It is more stable and less informal than the network stage, and 'outside' contacts tend to be narrower, limited to people with similar research interests, especially amongst students. Co-authorship becomes important and a large amount of research is carried out and the group gradually reveals its detachment from the prevailing pseudo-paradigm. Even at this cluster stage, though, personnel changes rapidly and relations are seldom long lived.

The final stage, the emergence of a speciality is in effect the institutionalisation of the cluster, its work and ideas. Prior to becoming a speciality, the cluster is still vulnerable for it has not established formal structure and procedures and relies on the informal connections of co-authorship and communication. These are dependent on individual contacts, rather than institutional organisations, even though they are self-conscious. The speciality emerges when the cluster develops, through an institutional base (or bases), a 'regular process for training and recruitment into roles which are institutionally defined as belonging to that speciality'.

The transition from cluster to speciality stage is a logistic progression. Students become successful in their own right, the new orientation, as it blossoms, cannot be contained as a cluster at one institution. The successful cluster becomes dispersed.

'We might hypothesize that the foremost factor in determining whether a coherent group develops at some location in the general science structure is: Are the scientists involved proceeding only empirically (simply moving from one research problem to another, without benefit of broad, theoretical guidance) or, alternatively are they carrying out a new theoretical orientation and being guided by it ?' (Mullins, 1973, p. 27).

In short, then, success depends upon whether the coherent group has a research programme. Most groups are not successful, according to Mullins, they do not have any effect on the course of science.

While a programme seems to be essential in Mullins analysis it is not apparently enough. There are necessarily other external factors that must exist in order to provide a context for the working through of the programme. These factors include a centre for the training of students and the carrying out of a research programme that necessitates the creation of an environment for close work equipped with 'intellectual materials'. A social organisational leader is also important, someone to orchestrate the advance of the programme. The availability of these factors, which includes the access to cash, is external to the research programme and thus to the intellectual development of science. Mullins does not bother to attempt to analyse how such resources become distributed to practicing scientists.

The speciality leads to inpersonalisation; the penalty for success is lack of community spirit. Members of a speciality will be aware of (at least some) of the work done by other members.

'They may share a paradigm and a set of judgements about what general work should be done in the field, although the details of those ideas might differ .... The speciality's problems might be described by Kuhn's concept of puzzle-solving which is the normal activity of science. Kuhn describes puzzle-solving as having the following characteristics: an assumed solution, rules which limit the acceptable solutions, and rules which limit the means for arriving at those solutions'. (Mullins, 1973, pp 74–75).

Mullins and Kuhnian Paradigms

Despite Mullins' avowed intention to elaborate a Kuhnian paradigm, he nowhere discusses the fundamentals of Kuhn's mechanism of 'progress', namely anomalies, crises and revolutions. Only at the speciality stage is there any clear isomorphism between Mullins model and the Kuhnian mechanism. Here it seems the normal or puzzle-solving phase of the Kuhnian cycle resides. However, Mullins' model has it that normal science resides in a speciality, that social networks have operated to the ultimate establishment of a speciality. The implication is that scientific knowledge grows through a programme of work that becomes increasingly articulated until it finally becomes institutionalised. Futhermore, that this process has evolved a speciality, the epitome of normal science supposedly initially formulated as a result of a common 'Gestalt switch', which is characterised by a paradigm that may be shared by all the members of the discipline or speciality but which is unlikely to be similar in all details. In addition, the progress from the 'Gestalt switch' which informs the paradigm group, to the speciality (which itself does not totally encapsulate the ideas of the scientist) is one that may run parallel to other areas of involvement for the scientist. These alternative areas of research may go beyond the boundaries of the 'new' paradigm characteristic of stage one.

The Kuhnian mechanism has been substantially amended by Mullins in the construction of his model. Mullins suggested that the speciality stage is the residence of normal puzzle solving activity yet it is not to be supposed that the speciality embodies a paradigm. On the contrary, a speciality is located within a paradigm. Given that scientists may operate within more than one speciality, according to Mullins model, one is left with the unanswered question as to whether these specialities may be from within different paradigms. If so, of course, Kuhn's incommensurability postulate is breached. In effect, specialities appear to be equivalent to the usual notion of discipline (or major sub-discipline) and there appears to be no mechanism in the Mullins model for the progress of science. Community practice can be quite distinct from the state of scientific knowledge and its revolutionary moments. Specialities emerge and die out within a paradigm.

The speciality may contribute to anomalies and crises and may die out following a revolution in science because of loss of new recruits. On the other hand the speciality may be quite independent of revolutions by transcending paradigmatic shifts or even becoming established in a pre-paradigmatic stage. After all, the only requirement in this model is social cohesion, communication and an agreed programme. Mullins is not concerned with the way that programme evolves theoretically, only socially. The content of the science is peripheral, of interest only as an agent of cohesion. It is the social organisation that mediates content not any necessarily inadequate conceptualisation which leads to crises and revolutions.

Mullin's model effectively does away with the 'gestalt switch' and the mechanistic development of scientific knowledge. The espousing of Kuhnian perspectives is irrelevent to the model except in as much as it provides a general context for all workers in an area. The paradigm in Mullin's model is at best a pseudo-paradigm, and indeed the model is better suited to a Lakatosian research programme theoretical base, propounding as it does the description of the social organisation of research programmes. Mullin's model traces an ideal typified process of the development of research programmes from individual endeavours to institutionalised speciality. The implication is that scientific knowledge grows through the success and institutionalisation of a programme. This is only possible for Lakatos providing the positive heuristic is able to sustain progressive problemshifts. However, Mullins does not take up Lakatos' model and sticks with paradigms. Instead of developing Kuhn's thesis through the incorporation and elaboration of the invisible college structure, Mullin's offers an alternative community structure approach only loosely based on a developmental thesis of the growth of scientific knowledge.

The implications of this are important. Instead of providing a development he offers, in embryonic form, a model that argues for the primacy of social organisation within the constraints of a general theoretical orientation (which he inaccurately labels a paradigm and assumes is the result of a 'Gestalt type switch'). The development of scientific knowledge is still the result of an autonomous process in the Mullins model, but one that is no longer mechanistic, resulting from an inevitable pile up of anomalies and an eventual resolution of crises, but rather one that is self-consciously produced. The development of scientific knowledge is at core a reflective process, and one legitimated by a group or interactive structure.

This is clarified by reviewing his work on the phage group, which led him to this model

The Phage Group Study Reviewed

The intentionality that is crucial to Mullin's construct may be illustrated by reviewing the phage group analysis that he developed. Such a review will however also expose other problems, namely those of doing historiography, even given the limited horizons of Mullins historiographical critique. Mullins does emphasise internal history and is 'guilty' of post hoc rational reconstruction. The problem of historical contextualisation also arises.

Mullins' first problem lies in the the designation of the phage group as central concept of investigation. This labelling, and the use of this so labelled grouping as the focus of attention is the result of post hoc rational reconstruction. While the group, (named for the bacteriophage) was important for the development of molecular biology it came to be named and identified only retrospectively (as Kuhn,1970, points out). Who is to be identified as being part of the group, its self identity and the publication of its existence and intent are features Mullins fails to explore. In short it is a recreated group whose intellectual forebears, for example, are still debatable and have been discovered only after the event

Mullins also has a problem isolating the paradigm. He offers at one point a wide paradigm embodying two approaches:

'Phage work was not the first or only attempt to study this problem. Biochemists, geneticists, structural chemists, X-ray crystallographers and others were also trying to determine the structure and function of large, biologically interesting molecules. These scientists can be regarded as constituting a paradigm group. All were studying the same basic problem, but they had no particular contact with each other. Nonetheless, as Stent (1968) points out, their approaches could be generally classified as structural or informational. The phage group's approach was basically informational....' (Mullins, 1972, p. 55).

However, later Mullins emphasises the informationally informed 'paradigm' as provider of research programmatics.

'The group's paradigm, formally stated, became the study of phages to solve the problem of genetic information transmission with methods as precise as could be developed. This paradigm, like most real paradigms, was not initially very precise, but it did become more precise as time and work passed. A very important event of this paradigm occurred when norms were established to govern the kind of research to be done and the manner in which it was to be presented.' (Mullins, 1972, p. 56).

A more consistent account may be provided by the use of 'orientation' or 'general area of study' to delimit interest in the 'structure and function of large biologically interesting molecules', of 'pseudo paradigm' to refer to the 'informational approach', and research programme to refer to the statement of intent. Substituting 'research programme' for 'paradigm' in the above quote does not distort the essence of Mullins comment, but increases the specificity.

This problem is further highlighted by the nature of this paradigm stage. If it is taken to imply an established paradigm, then normal puzzle solving should occur. This however seems to be the prerogative of the speciality stage. What then is the paradigm group stage ? While Mullins is not clear in his theoretical elaboration, he suggests that the 'Gestalt Switch' has been accomplished, thus by implication puzzle solving should be able to commence. In his case study he is clearer. The paradigm group stage is the period of the formation of a paradigm. It is, in Kuhnian terms, the period of crisis and revolution. This period in the phage history is characterised by (retrospectively reconstituted as the period of) debates between Schrodinger and Bohr about the extent of the physical, as opposed to possible non-physical elements, in the transmission of genetic codes (see Mullins 1972, p 56). In short, it was a period of fundamental epistemological struggle.

The struggle resolved itself through a self-conscious praxis. A diffuse group of scientists emerged within the general orientation with an intent to get at the 'secret of life'.

At the paradigm stage, rather than the loose, unconnected groupings that Mullins hinted existed, there was, on the contrary, in the centre, a determined and planned appropriation of molecular biology by a group of fundamental physicists for their own ends. The eventual emergence of a 'new speciality' was the result of conscious social-political decision which has far from an autonomous development and bore little relation to the ideal-typification implicit in the 'natural' development of science through social interaction, behind which Mullins conceals (deliberately or otherwise) the reflective element. Mullins seems to miss or dismiss the significance of his own evidence. He ignores the view of Anderson that the attempt by Delbruck (later in conjunction with Luria) to encourage phage work (in the US, where he emigrated) constituted a 'plot'. (See Anderson, 1966 [no reference]) Delbruck, Szilard and other fundamental physicists decided (in 1933-34) at the paradigm stage, that the future of that area of physics lay in appropriating biology. They decided to enter biology (Delbruck was the agent) and accept that new laws of physics may be necessary to deal with this appropriation. To proceed with their plan involved the recruitment of two 'quantitative' biologists, this they did, and there followed extensive and concentrated discussions between the 'alliance'. Emigration and the Second World War hampered progress but the publication of Schrodinger's 'What is Life' in 1944 provided a 'shot in the arm' for recruitment.

In 1946 Delbruck, Luria and Hershey began taking systematic steps to recruit new scientists for phage work, principally through the initiation on the summer phage course at Cold Spring Harbor. The evolution into a cluster, the most problematic of Mullin's stages, is principally characterised by group motivation. Not just a self conscious awareness but a concerted motivation to unify and train new graduates.

Motivation, for Delbruck, involved recruitment particularly recruitment of students (a possibility once he moved to the California Institue of Technology in 1948).

'Delbruck's recruiting technique was extremely vigorous Delbruck's advance over similar scientists in other positions was his recognition, during the early forties, that under traditional conditions he would not have sufficient students soon enough to maintain the momentum created by his and Luria's initial discoveries. Initiating a summer course was his solution to this dilemma.' (Mullins, 1972, p 64).

Cold Spring Harbor provided the focus for the development of a cluster because of its summer school and continual seminars, the availability of post doctoral and research funds as well as flexible short term funds.

A further problematic relates to the evident lack of paradigm clarity at any one stage. Mullins' description of the period 1935-1945 reflects Lakatosian elaborations of the Methodology of Scientific Research Programmes rather than Kuhnian concepts. There is little evidence of isomorphism with a Kuhnian analysis -or to put it another way, a Kuhnian model is of little use in explaining the procedures, practices and outcome of molecular (phage) biology in this period.

Mullins slips into a discussion of 'first' and 'second' generation workers and notes that while a number of biologists were doing work on phages or related subjects they had little systematic communication with each other until 1946. Even Delbruck and Luria were at different universities and many of the other major workers were separated from communication by the war. Moreover,

'their simultaneous involvement in other activities might equally well have separated them. Some of them knew other scientists involved in phage work but in any case they lacked contact with them.' (Mullins, 1972, p. 58).

Workers operating in a research programme reflect this situation but rather than being a negative aspect, as Mullins implies, Lakatos would see this 'cross fertilization' as essential for innovation.

Furthermore, Mullins shows that the phage group had a high turnover of personell, (that recruitment was highest in the network stage) that a high proportion of the group (over forty per cent) were physicists, that there was little stability in the group, even the earliest members failed (for the most part) to stay with the phage group. In short, recruitment and turnover represent something more like a research programme within an 'informational' framework orientation or 'pseudo-paradigm'. 'The high turnover and the short career...indicate that this network was not formed around a very few, continuously productive, founders. Indeed, at any given point in time, the network's predominant social structure consisted largely of interacting pairs and triads of collaborating and collegial scientists with few, if any, larger groups.' (Mullins, 1972, page 69. Mullins footnotes Price (1963) here, presumably to indicate the similarities with invisible colleges.).

Even at the cluster stage the paradigm specificity is not clear. Mullins points out that this specificity could not be connected to a central core of constant members as they didn't exist. The other factors symptomatic of a cluster, which did sustain the group, might have provided some last hope of insight to the nature of the supposed paradigm. These factors were the recognition by others of a group of scientists studying phages, the cluster's belief system, a common life style and a high rate of interaction among its members. Clues to paradigm specificity might have been the core of Mullins' account of the phage group's role in the development of molecular biology. On the contrary, every evidence suggests that, despite the terminology, Mullins is referring to pseudo-paradigm 'containing' varied research programmes.

Mullins' failure to elaborate the beleif system or the recognition of the group is not just unfortunate for its failure to provide a clue to an articulated Kuhnian paradigm (indeed his whole model effectively shelves the core Kuhnian construct) it is more unfortunate in failing to provide a criterion for examining the emergence of myth.

Standard American Sociology

In relating his model to sociology Mullins provides, when dealing with 'Chicago sociology', an example of its post hoc nature and its propensity to distortion. The exposition of the model in relation to American sociology as a whole is less precise than his phage work and is plagued by innumerable 'grey' areas.

When outlining the emergence of standard American sociology, Mullins appears to be offering a very different view of a speciality than that underlining his phage group analysis of 1972. The latter involved a struggle for survival of a new sub-discipline at the margin of two major disciplines. The emergence of standard American sociology appears to be the evolution of a particular orientation within a paradigm-bounded discipline. Mullins initially refers to standard American sociology in general terms implying that Parsonian structural functionalism emerged at the cluster stage to become dominant. He later seems to ignore this reconstruction which casts the 'Chicago School' as part of the anomalous (and 'pre-paradigmatic') normal stage and redefines standard American sociology as almost exclusively what is commonly referred to as structural functionalism.

Thus what emerges as Mullins discusses standard American sociology and contrasts it with symbolic interactionism is a view of Chicago sociology temporally and intellectually distinct from standard American sociology. The normal stage propounded by Mullins as his analysis unfurls is of standard American sociology rooted in the 'disorganised' work of sociologists outside Chicago and who came together as a cluster after 1935. Mullin's argues that this date effectively marked the end of the Chicago era.

Mullin's account of Chicago sociology, is, he admits, in need of further research to see how it fits the network model. His use of populist accounts, notably Faris (1967) in an uncritical manner along with his naive assumption that early American sociology was characterised by a straightforward conflict between qualitative and quantitative practitioners, has led to his presenting a partial and distorted account of the history of standard American sociology. He is 'guilty' of a rational reconstruction leading up to a grand finale in which structural functionalism engages in battle with symbolic interactionism.

Mullins schematic account of American sociology is typical of the kinds of presentation which distort the history of 'Chicago sociology'. It is based on a popular conception which reflects the myths surrounding the 'Chicago school'. In particular, Mullins argues for the centrality of Mead, that Mead must have been a marvellous lecturer, that his students were dedicated, that most Chicago graduates attended sociology in his lifetime, and so on. Lewis and Smith (1981) and Fisher and Strauss (1978) have both questioned this assumption of Mead's importance. Mullin's goes further and suggests that Mead's death was instrumental in the breakup of Chicago sociology and the college's failure to recruit large numbers of graduate students in the 1930s. That there was an economic recession, that Park retired in 1934 are not mentioned by Mullins. Mullin's also argues that the Polish Peasant was an empirical study of Mead's (and James') theories, despite the fact that Mead, Cooley, James and Thomas had quite different views and Thomas was closer to Cooley than Mead. Thomas' key concept 'definition of the situation' owed nothing specifically to Mead. Mullins' reconstruction builds towards a reification of the role of Blumer. In order for his thesis to hang together, Blumer must be seen as central to Chicago sociology and not a peripheral character. Blumer is afforded the status of perpetrator of the Median tradition by Mullins, (Elsworth Faris is ignored), and placed at the centre as the organisational and intellectual leader of Chicago sociology. Mullins refers to the emergence of a cohesive symbolic interactionist group including Hughes and Becker, and ignores the evidence which clearly indicates that Hughes was not a Blumerian symbolic interactionist and that Becker was a student of Hughes. Further, Mullins argues that the tenets of symbolic interactionism were published by Blumer in 1938 having been previously unavailable to the rest of American sociology. The tenets Mullin's lists are primarily interactionist notions, as opposed to symbolic interactionist per se and were 'available' in the 'Polish Peasant' (Thomas and Znaniecki, 1918) and 'Introduction to the Science of Sociology', (Park and Burgess, 1921), both widely read.

Mullins whole account is based on an unproblematic linking of figures together into convenient groupings, creating artificial labels, (e.g. calling Franklin Frazier a symbolic interactionist) and assuming that Hughes, Becker and so on carried the same anti-standard American sociology banner as Blumer. He sees the early Chicago school as homogeneous and ignores the singular ambition and peculiar development of Mead encapsulated in Blumer's work. Furthermore, the description of the 'Chicago school' is highly selective and ignores any reference to race studies, the quantitatively based cultural analysis of Ogburn and so on. The major theoretical ideas at Chicago identified by Mullin's are Park's interpretation of Simmel's formalism, ecologism, and Meadian social interaction. Thomas is ignored, no mention is made of social disorganisation, and Dewey's contribution is overlooked. In conclusion, Mullin's dismisses all of the Chicago influences on standard American sociology which, he maintains was originated by Parsons.

He writes, 'Following the fate of all good intellectual groups, the Chicago cluster's students were hired elsewhere, its best teachers and researchers aged and retired or died and Chicago sociology became routinized as a speciality (in this case as the majority of the discipline), both intellectually and sociology. Clearly the data on the 1892–1935 period of Chicago sociology look very much like a group progressing through the four stage model.... However, this period has only an indirect effect on contemporary sociology filtering through standard American sociology and symbolic interactionism.... The group should be studied further to see whether it fits the four stage model'. (Mullins, 1973, pp 69–70, footnote 7).

Harvey (1985) concluded, 'Mullins' account of Chicago [sociology] is seriousy inaccurate and, as such, cannot provide the basis for a sound assessment of his theoretical model'.

Conclusion
Mullins' model is one that emphasises the networks of practicing scientists. It concentrates on the 'working environment' and relationships among scientists, assessing the potential institutionalisation of research programmes. The process of the development of scientific knowledge is subsumed under the rubric of the social process of theory articulation, even to the extent that Mullins suggests that an elitist take-over of a discipline (a socially organised 'coup d'état' rather than a concept revolution) leads to a new normal science being established. (Mullins, 1973, p. 25). The process of discovery is ignored.

Mullins' model of scientific community development is a mechanistic model (albeit likely to fail at any stage). While Mullin's suggests factors which are likely to lead to the success of research areas he makes no attempt (despite an espoused Kuhnianism) to investigate the way paradigms change and the relationships of the social relations to crises and revolutions.

Mullins demonstrates his model inductively through selected cases, but does not broach questions about the genesis and form of networks, what relation they have with the production of scientific knowledge, the process of innovation and change nor, even, the relationships between institutionalisation and legitimation. There is nothing intrinsic to his model that allows for elaboration of these areas of concern. His is, as Truzzi (1974) has commented, a thesis about associations of scientific workers rather than an analysis of theory groups.

Indeed, despite emphasising the importance of the social communication network for critical appraisal, Mullins fails to analyse the epistemological foundation of the process of critique. He takes for granted the Medvedevian description of the critique, thereby accentuating the role of internal history. Medvedev notes that critical appraisal involves the determination of weak points in a piece of research, including methodological inadequacies and the assessment of its position among other investigations in the field. This emphasis is non-transcendent and is unlikely to inaugurate innovation. Mullins seizes on this internalistic construction. The assembling of an innovatory group to provide self-legitimation for innovation then becomes essential. The role of the wider scientific and social milieus is ignored.

Mullins excuses the internal orientation of his model by suggesting that science is changed by academics, although ignoring that they have social as well as academic contexts from which to draw ideas. Nonethless, Mullins claims that transformations from one stage of his model to the next are self- regualting and do not require explanations based on individual personalities, general characteristics of society, or the intellectual history of sociology. For Mullins, such elements are neither necessary nor sufficient, they simply clarify how one of the properties of a given stage works itself out in a given context.

Mullins, then, relies entirely on an internal history, one that concerns itself with an organiational transformation devoid of content or analysis of the process of conceptual change.

Mullins claims to have derived his model from an empirical case study in natural science and superimposed it on sociology. Neither the original model nor the transposition are unproblematic. His empirical work is primarily based on secondary sources, and his comparison of theories is based on the following rules of thumb. He considers two theories similar if authors cite similar sources, if authors are colleagues in some respect, (including student teacher relationships), if authors consider their work to be similar or others consider it to be so. His empiricism is shaky and his model relies on post hoc analysis. The result is a pseudo-conventionalist structure. The history of a discipline is rationally reconstructed to create interactive networks that fit the schema, leading to the artificial designation of particular 'traditions', whose evolution are traced. The exercise is infinitely flexible and the identification of specialities is as unconvincing as any other conventionalist designation. Mullins scheme is nothing more than a method of rationally reconstructing the history of disciplines.

Mullin's approach thus has strong Lakatosian elements. Although admitting 'external' elements such as luck, institutional affiliation, etc., as factors in the success of a programme, Mullins treats them as rational factors, in the same way that Lakatos absorbs components, of an ostensive external genesis, into his internal rational reconstruction. However, Mullins is not a Lakatosian in disguise. Fundamentally, he adopts, as has been shown, a community-based approach, he is interested in the structures of different scientific units and relates the progress of science to them in what emerges as an inconsistent manner, there being little nomothetic unity in his model, far too much is arbitrarily determined. While Lakatos' research programme inheres in communities and the nature of problemshifts is (at least normatively) decided communally, it is the content of the endeavour that informs the progress of science. Ultimately, for Lakatos, this content is rationally progressive. For Mullins, progress in science, especially social science resides in fashion and persuasion, aided and abetted by determination, institutional facilities and so on.

Despite his avowed intention to go beyond a 'schools' approach by examining the development of the theories themselves, he merely offers an account of theoretical developments within a four-stage community structure and sidesteps the dialectical problem of linking theoretical change to social structural change.

Mullins seeks to understand how different types of theory come to be written, why some theories are similar and others differ greatly and why any given theory begins and eventually dies out. His answer lies in the structure of research communities whose success depends upon a variety of 'irrational' factors. He provides a context but no analysis of the relationship between the context and the theory other than to ascribe it to intentionality and determination.

The important feature of Mullins' model is that it emphasises reflexivity within an autonomous model, but his analytic scheme, concentrating on the social interaction elements ignores the mythology which is central to historical reconstruction of the process of knowledge production. Not only is Mullins' model based on an internal history, plagued by an unanswerable charge of post hoc rational reconstructionism its utopian historicist base proscribes it from an analysis of mythical elements. it also fails to analyse the mythology surrounding the group. Further the uncritical assimilation of categories from the philosophy of science has provided a framework which gives credence to this myth encrusted 'history' of the phage group.

Mullins thus reduces the role of 'paradigm' to that of 'theoretical focus', and the metascientific enterprise to the self-conscious esotericism of a small consensual group. The 'success' of such a group being dependent, in effect, upon random assortment of other factors operative at various of the 'levels' defined in part two above. Mullins provides no basis for a critical analysis of the interrelationship between the specific group action and the 'external' levels at which these other factors operate. In effect, his development of a model from a peculiar case (the phage group) and its application, uncritically to American sociology, has led him away from a reflective metascientific analysis and the resulting concentration on community structure has failed to bridge the gap between internalist theories (in this case Kuhn's paradigm notions) and externalist theories.

Mullins provides some insights into research as a career but his model is inadequate as a theory of the production of knowledge as it fails to explore the relationship between research careerism and knowledge production processes, intimating that knowledge resides in organisational structure.

It is not clear, then, either whether Mullins dispenses with the construct of progress based on paradigm shifts, because, possibly it reflects only a realm of ideas and is not integral to an analysis centring on social organisation, or whether his system of stages can accommodate the Kuhnian mechanism. If it cannot, then the use of paradigm as a core thesis is unconvincing and Mullins seems to be using some form of 'pseudo-paradigm' structure to underpin his approach.

related areas