AustLII Home | Databases | WorldLII | Search | Feedback

University of Melbourne Law School Research Series

Melbourne Law School
You are here:  AustLII >> Databases >> University of Melbourne Law School Research Series >> 2009 >> [2009] UMelbLRS 39

Database Search | Name Search | Recent Articles | Noteup | LawCite | Author Info | Download | Help

Dent, Chris --- "Not All Practices Are Equal: An Exploration of Discourses, Governmentality and Scale-Free Networks" [2009] UMelbLRS 39

Last Updated: 29 August 2011



This is an electronic version of an article published in (2009) Social Semiotics Vol. 19 pp. 345-361. Social Semiotics is available online at:


Foucault’s ideas surrounding the notion of governmentality are built upon the intersection of multiple discourses and discursive practices – a ‘complex topography of rule’ (Dean 2007). The notion of disciplinarity is well accepted in the literature, however, there are few attempts to conceive how practices, from a range of discourses, relate to each other. Everyday observations indicate that not all learnt practices are equally important to a given subject. To say that all these practices are inculcated in order for the subject to be able to conduct themselves appropriately does not provide insight into the relative importance of particular practices to a specific subject’s constitution.

The central contribution of this article is that the sum of practices acquired by a subject may be conceived of as a network; more specifically, as a “scale free network”. The adoption of this perspective adds texture to the Foucaultian tradition in a way that emphasises the connections between acquired practices – the perspective facilitates the conceptualisation of the sum of practices in terms of a topography of practices. This approach is explored through an examination of the background, training and work of patent examiners. Examiners make a useful example as they operate within both the scientific and legal discourses and, therefore, learn and express practices associated with both discourses. This network-based approach offers a new perspective for understanding practices that occur in multiple disciplines – that is, it offers a useful conceptualisation of the processes by which discursive practices are ordered (thereby facilitating acquisition).


I. Introduction

Foucault’s ideas surrounding the notion of governmentality are built upon the intersection of multiple discourses and discursive practices. Governmentality, as a concept, has been well embraced in the literature. There are, however, few attempts to conceive how practices, from a range of discourses, relate to each other. Such a lack does not detract from the value of Foucault’s work, though, it does limit more detailed understandings of the practicalities of governmentalist subjectification. This article introduces a concept that provides for a more nuanced perspective on the operation of multiple discourses on the life of modern subjects.

Observations of everyday life indicate that not all learnt practices are equally important to a given subject. Some practices are, for example, required as preconditions for the taking on board of other practices; further, some practices are learnt for a specific task that may only be carried out once or twice. To simply say that all these practices are inculcated in order for the subject to be able to repeat them does not provide insight into the relative importance of particular practices to a specific subject’s constitution. The introduction of a new perspective, then, to conceptualise the relationships between diverse discursive practices has the potential to add depth to understandings of the day-to-day conduct of conduct of modern, multi-tasking citizens.

The central argument of this article is that the sum of practices acquired by a given subject may be conceived of as a network. The argument here is that the understanding, and analysis, of discursive practices from multiple discourses may be achieved through the use of a concept imported from the sciences – that of a particular type of network, a “scale free network” (SFN). This approach is explored, in a practical setting, through an examination of the background, training and work of patent examiners. Examiners make a useful example as they operate within both the scientific and legal discourses and, therefore, learn and express practices associated with both discourses. They also act as functioning members of a government agency. Work has already commenced on the discursive analysis of legal practices (for example, Dent 2005; Dent and Cook 2007) and the interaction of multiple subjects constituted by different discourses (Dent 2007). This piece is among the first, however, to focus on the conduct of subjects constituted by multiple discourses, who are trained to operate, on a daily basis, across distinct disciplines.

The adoption of the perspective allowed by SFNs – a topography of practices – adds texture to Foucault’s theories of governmentality and discursive practices in a way that emphasises the relationship between acquired practices. The perspective enables a focus on the individual training, knowledge and experience of modern, multiply constituted individuals. This article will, in order to explore this understanding, first, engage with Foucault’s ideas of subjectification and discursive practices, and second, introduce the concept of SFNs. Finally, the article will draw together the practices and the perspective in an exploration of the training and experience of patent examiners. The use of the practical example of the diverse training of patent examiners will allow the demonstration of the application of SFNs to the understanding of the operation of multiple discourses within the rubric of governmentality.

II. Discursive Understandings

Foucault’s work on governmentality encourages a focus on the role of discursive practices in the analysis of everyday conduct. He, for example, spoke of ‘procedures, analyses and reflections, ... calculations and tactics’ as being central to the modern form of Western governance (1991a, 102). This ‘ensemble’ of practices may, at the very least, be understood to cover three broad areas: (1) those associated with life as a citizen in a modern democracy; (2) skills associated with the particular profession of the subject; and (3) practices that may be seen to arise from more personal interactions – those relating, in part, to the “care of the self” (for example, Foucault 1990a, 1990b and 1992).[1] Each of these areas (though the areas are by no means discrete) relate to different practices acquired by individual subjects in order for them to conduct themselves in a manner in accordance with different bodies of knowledge.

The central point of this article rests on Foucault’s understanding of the processes of subjectification – the inculcation of dispositions and habits (Rose 1999, 46) – inherent in today’s society.[2] At the heart of the processes of subjectification is the contention that ‘the self is not given to us’ (Foucault 1983a, 237). However, ‘it is not enough to say that the subject is constituted in a symbolic system. It is not just in the play of symbols that the subject is constituted. It is constituted in real practices – historically analysable practices’ (Foucault 1983a, 250).[3] The discursive practices of a given discourse can be understood to produce the permitted actions for the subjects of the discourse. These discursive practices act as norms of behaviour. These practices are not to be understood as commands from a higher authority. Instead they can be understood to operate as accepted and habituated patterns of action. Governmentality, then, may be understood as sets of practices that are learnt by citizens, with their “learning” of the practices coming from other subjects of the discourse.

It is to be emphasised that each citizen, from this perspective, is constructed to behave according to learnt forms of behaviour:

Under the surface of images, one invests bodies in depth; behind the great abstraction of exchange, there continues the meticulous, concrete training of useful forces ... it is not that the beautiful totality of the individual is amputated, repressed, altered by our social order, it is rather that the individual is carefully fabricated in it, according to a whole technique of forces and bodies (Foucault 1977, 217).

For Foucault, no individual exists prior to discursive inscription, for ‘discursive formations produce the object about which they speak’ (Cited in Dreyfus and Rabinow 1983, 61). This is done through the web of power relations that permeate society, through the multiplicity of interpersonal relationships that exist in the rearing of children and adults. From this perspective, the mechanisms of discursive control ‘homogenise populations through knowledge, separation, observation, and experiment, but do so only by “individualising” people in cells and classroom desks, by examination and experimentation’ (Bové 1980, 29).

Those of us who have been trained as students and patients, for example, actively participate in the discourses of education (Foucault 1977) and medicine (Foucault 1973) – through our training as students and patients we behave as we should and in this way perpetuate these discourses. The practices necessary to learn, to control ourselves in the name of our own health (whether in the short term suffering the “indignities” of medical examination, or the longer term, healthy living patterns) are instilled in us for our own good. Western subjects have taken on board the medical and educational discourses through a variety of means, including interpersonal interactions with teachers, doctors and other members of the discourses. Subjects are, then, embedded in a web of power relations – a skein that produces an “individual” complicit in her or his own constitution as a subject. In Foucault’s words, ‘the myriad of bodies which are constituted as ... subjects [are] as a result of the effects of power’ (1980a, 98). The self, and the capacity to conduct the self, is therefore built up habit by habit, practice by practice. More simply, we are governed through the ‘regulated freedom’ (Barron 2005, 976) of apparent choice; no matter how shallow that freedom is (Rose 1999, 97).

It has been suggested that Foucault does not have a fully ‘rounded theory of subjectivity’ (McNay 1994, 102). For the purposes of this article, it is sufficient to accept that subjects are constituted as “individuals” by exposure to, and a motivation to copy, the practices of others. The Foucaultian explanation is that all subjects are embedded in power relations that ‘materially penetrate the body in depth without depending even on the mediation of the subjects’ (Foucault 1980b, 186).[4] This does not require any conscious engagement, on the part of the subject, in the process of learning, and repeating, practices – though, of course, some practices are deliberately learnt by subjects. Certain practices of governance encourage (young) subjects to ‘acquire habits’ (Foucault 1977, 135); at some point, subjects “choose” particular habits to acquire as an exercise of ‘fashioning their freedom’ (Bernauer and Mahon 1994, 143). The purpose here is not to add to the debate on the processes of subjectification, rather it is to accept that practices are learnt and to provide a perspective that considers the relationship between them. In the same manner that Foucault accepted that subjects are governed and sought to understand government as a ‘complex topography of rule’ (Dean 2007, 84), this article simply seeks to consider a concept that enables a new way of understanding the relationship between learnt practices.

What is important here is that members of a particular profession acquire, for example at university, the discursive practices associated with the discipline/s in which they are trained. These learnt practices are then repeated by the professional in the conduct of the work in that profession. This means that the practices of a given profession tend to remain (relatively) constant.[5] There are many subjects, however, who operate on the border between different discourses. Further, the theory of governmentality presupposes that citizens are constituted, and therefore regulated, by multiple discourses. Such a simple description does not assist in the analysis of the impact of multiple discourses on the constitution of a single individual. Recourse to specific categories of subjects is, however, helpful in order to better understand the interaction of discourses and discursive practices. One example of a profession that operates across discourses is that of patent examiners, to be discussed below, who work in both legal and scientific disciplines. The incorporation of the SFN concept into the understanding of governmentality, therefore, provides a tool that may be used to better understand the relationships between acquired discursive practices.

III. Scale Free Networks and Discursive Practices

A. Scale Free Networks

Before the specifics of SFNs are described, it is important to understand what is meant by a network. For the purposes of this article, it is sufficient to describe a network as an open-ended collection of sites – otherwise known as nodes. These sites can be any collection of objects; common examples are web-pages, airports or groups of friends and acquaintances. The number of sites may not be limited and more sites may be added according to the structure of the network. The sites have links with some, but not all, other sites. The links produce the structure that potentially limit the accretion of new sites.

Scale free networks are both a particular type of network and a relatively new way of describing complex systems. The types of systems to which they have been applied include transport systems, academic citations, the internet and metabolic and protein networks (for example, Barabási and Bonabeau 2003). The common feature of these networks is their very high number of nodes. Equally important to the number of sites is the connections between these nodes. What makes SFNs “scale free” is the distribution of connections between the nodes. Not all nodes have the same number of other nodes connecting to them. Diagram 1 shows a simplified version of this. There are some nodes with a high number of connections while there are others with very few. Take for example, the internet – a network comprising of billions of sites. Some nodes such as or have many more connections than the single page of a person who only blogged once.

Diagram 1: Stylised Scale Free Network – showing the inter-nodal connections

The relationship between the number of connections of each of the nodes in a SFN is described by a “power law”. For the mathematically inclined, a “power law” represents a ratio between two factors, in this case the number of nodes and connections, that when mapped on a logarithmic scale, produces a straight line. For the non-mathematically inclined, it is sufficient to say that in ‘contrast to the democratic distribution of links seen in random networks, power laws describe systems in which a few hubs ... dominate’ (Barabási and Bonabeau 2003, 53). Unlike random networks, the “nature” or characteristics of particular nodes means that they have many more connections to other nodes.

This characteristic means that SFNs are ‘highly efficient and relatively robust, as the random failure of any node will have little effect on the system because it is far more likely to be poorly connected that well connected’ (Lake and Wong, forthcoming, 4). The corollary of this, however, is that an event that takes out one of the dominant nodes will have a catastrophic impact on the overall network. This suggests that particular nodes are disproportionately important to the system. Using, again, the internet example, if a number of the root servers were unavailable then the World Wide Web would be severely compromised, whereas, if the server at a small university in Australia was unserviceable, it would have minimal impact on the wider network.

A further, relevant, characteristic of SFNs is that they have a ‘high clustering coefficient’ (Ahmed et al 2005, 240). This means that there is a high level of inter-relatedness between the nodes of an SFN. More specifically, the clustering coefficient ‘represents the probability that two neighbours of a randomly chosen node will themselves be neighbours’ (Steyvers and Tenenbaum 2005, 46). An SFN, therefore, is more than just a collection of minor nodes connected only to a small number of major nodes. The minor nodes themselves are connected to each other, however, the number of connections to a minor node is much smaller than the number to a major node.

The importance of key nodes flows onto the mechanisms of change that occur in these networks. The research of Klemm and Eguiluz suggests that for “growing” SFNs, a new link will “prefer” to connect to a node that has been added to the network more recently (2002, 036123-1). In some networks, this is obvious. If, for example, the network of academic citations is explored, the more recent publications are the ones that tend to get cited more (except for a few key, discipline-defining, texts) (Redner 1998). This process means that the key nodes of an active network can change over time; a feature that is important for the application of the concept of SFNs to discursive practices – in order to produce an understanding of the topography of practices.

The final aspect of SFNs, relevant to this article, is the assertion of Strogatz that ‘structure always affects function’ (2001, 268). This suggests that the relationships between nodes impact on the acquisition (or habituation) and expression of the practices themselves. More importantly, it indicates that the rules of formation of the network – the manner in which the network is “allowed” to grow – are integral to the way in which the network operates. An example of this is the manner in which the network of neural pathways (neurons and synaptic connections) impacts on the functioning of the brain. More specifically, the structure of the brain, and the connections between cells, directly affects the “thinking” done by that organ (see, for example, the discussion of neuroplasticity in Tancredi 2005, 46).

All the features of SFNs described here are important for the exploration that follows in this article. The concept, though taken from the analysis of physical objects and the relationships between them, provides a useful analogy for the relationships between habituated discursive practices. The following section explores this application of the idea in more detail.

B. Distribution of Discursive Practices as a Scale Free Network

It is not difficult to consider, as a network, the number of acquired discursive practices that constitute a subject. It makes sense to understand that practices learnt by a person are linked to other practices already learnt – riding a motorbike, for example, is often based on skills learned from riding a bicycle. From a discursive perspective, each practice that is taken on board by a discursive subject can be seen as an object, a node, connected to other nodes, or discursive practices. To take an example from a specific academic discipline, a doctor or nurse who learns to operate a sphygmomanometer internalises that practice alongside the practices acquired when the operation of a stethoscope was learnt.

The contention here is that the perspective provided by SFNs is an effective one to understand the interaction of practices learnt by a given subject; and, in particular, the perspective enables insight into those subjects who are on the fringes of a discourse or who are trained in multiple discourses. The characteristics of SFNs described in the previous section are useful for this understanding of discursive practices. First, the network of acquired discursive practices is best conceived of as scale free rather than any other form of network because of the fundamental importance of particular practices or nodes. The fundamental nature of some nodes means that specific practices may be understood to dominate the entire network. These central practices include those associated with the learning of new discursive practices, the communication of practices and those associated with the repetition of previously acquired practices – in other words, those associated with learning, language and memory respectively.

A significant benefit from using the SFN concept to understand discursive practices is that there can be a focus on the relationship between the acquired practices. A relationship that is of particular importance is that of the learning of new practices. More specifically, the advantages of a SFN approach include that it enables a different understanding of the processes of subjectification; and it encourages other understandings of the processes of discursive change. From this perspective, in order for a new discursive practice to be learnt, a discursive subject must have already taken on board practices to allow for the acquiring of other practices. To deconstruct and repeat the behaviour of someone else, or to remember the actions of another, is to express the discursive practices of learning and recall. As a result, all newly acquired practices may be seen to be connected to these core practices.

Other practices that may be seen to dominate sections of the network associated with particular disciplines include those that relate to the “scientific method” in the sciences and those that relate to the “doctrine of precedent” in law – where lawyers and judges learn, or subsequently use, a particular decision or a particular court is connected to the fundamental legal discursive practices relating to precedents (see Dent and Cook 2007). Practices expressed by scientists, when they behave as scientists (as opposed to when they behave as parents or aikidoka), are connected to those practices that dictate the norms of behaviour of scientific research.

That there are dominant hubs of practices is also evident in the expression of practices. As long as the practices associated with learning and memory are retained then the subject maintains the ability to relearn a skill that has been forgotten. It matters less, for example, if a lawyer forgets the detail of a particular case than if they lose the skill of reading or of discerning the precedential value of cases based on a knowledge of jurisdictions and the hierarchy of courts. In other words, the network of practices is robust – the forgetting of minor practices is not catastrophic to the network as long as the hubs survive.

Further, when a practice is acquired it is linked with other practices. The use of particular scientific instrument is connected to both the theoretical science and the use of other instruments in the field. As highlighted above, for example, the practices associated with a sphygmomanometer are connected to practices associated with the theory of medicine (including the specifics of blood pressure and the more general ideas of diagnostic procedures) and with practices connected to the use of stethoscopes. These close links reflect the high clustering coefficient feature of SFNs where nodes are grouped into areas of practice. Nodes relating to the use of medical apparatus are connected to other nodes relating to medical training and are not directly connected to practices centring on other activities such as practising aikido or raising children.

A key feature of the SFN conceptualisation for the understanding of discourses is the manner in which the growth of networks is allowed for. This understanding provides a grounding for a more complete perspective on the processes of discursive change. That a new practice is acquired in connection with previously learnt practices accounts for both the change in practice of individual subjects and in the discourse overall. A new practice, for example, is acquired in the context of previously learnt practices. A new practice will only make sense given what has already been acquired.[6] It will be easier, and the process more complete, when a lawyer attempts to comprehend a new judgment from the Court of Appeal (for the purposes of understanding and absorbing a new statement of law) than if practitioner of traditional Chinese medicine attempted the same feat (the reverse is also true – a traditional Chinese medicine practitioner will make more sense of a new chart of points for Tui Na than a lawyer would). It must be emphasised that the new judgment of the Court of Appeal would, nonetheless, be produced through the appropriate expression, by the judges, of internalised legal discursive practices (see further Dent 2003).

This understanding also offers insight into the rate of discursive change. From this perspective, individual subjects, and discourses themselves, change slowly. The number of connections associated with each node, each practice, indicates how embedded the practice is in the experience of the discursive subject. The more embedded it is, the greater the inertia of the practice and, therefore, the harder it is for the practice to change. That is, for a discourse overall, the larger the discourse the greater the inertia of the practices associated with the discourse (see also Dent 2007).[7]

The application of the SFN conceptualisation to a Foucauldian understanding of the operation discourses in a governmentalist world has a number of benefits. These include the privileging of practices, thereby removing the need to use ill-fitting labels for discourses; the provision of a perspective that caters to the importance of particular, dominant, practices; an acknowledgement of the relationship between discursive practices – including the importance of the structure of the network to its function; and the provision of a mechanism that accounts for both the existence of discursive change and the inertia associated with such change.[8] The last part of this article will apply the concept of SFNs to the training and experience of patent examiners. Examiners are discursive subjects who, in order to do their job, have to work across disciplines – the result of training in two distinct academic discourses (science and law) and an administrative one. They are, therefore, an appropriate example to explore the value of the concept to the understanding of modern, governmentalised, subjects.

IV. Patent Examiners, their Practices and Concept of SFNs

Patent examiners are an integral part of the patent grant process. Their role is to read and assess the applications for patents. The analysis, in terms of their training, of the manner in which they examine patent applications is problematic if a disciplinary approach is taken – particularly if their “individuality” is focused upon. One study, for example, has found that ‘there may be as many patent offices as patent examiners’ (Cockburn et al 2002, 2). The utilisation of the SFN concept allows for an understanding of their actions that takes into account the diversity of academic backgrounds and, therefore, provides a theoretically rigorous perspective that accounts for the individualised impacts of examiners.

Most examiners have completed a significant amount of training in a scientific field before their employment at a patent office. The European Patent Office (EPO), for example, requires applicants for patent examiner positions in the area of engineering to have a full university degree in engineering; many of them, in fact, have PhDs. Therefore, by the time examiners commence any training in patents they have already acquired significant sets of specific, and specialised, discursive practices. As patent examiners are employed by patent offices to examine the patent applications filed by inventors or their employers, they have to be trained at the offices to examine applications. This involves specialised training in the legal tests for the patentability of inventions. Also, examiners have to be trained in the administrative and bureaucratic practices associated with (in most cases) public sector employment – these include personnel policies and health and safety matters. Whilst not necessarily as intellectually challenging as their university training, these procedures still need to be learned by the examiners in order to function as public sector employees.

A. The Patent Examination Process

To assist in the application of the perspective to the work of examiners, it is useful to consider, in more detail, what they have to do when examining a patent application. An outline of the steps involved in the Australian patent examination procedure is given in the Australian Patent Office Manual of Practice and Procedure.[9] The steps include:

This list of requirements (the order of which is not strict) illustrates that a range of practices need to be deployed in order to fulfil the task. Some relate to administrative matters (checking a request for examination has been made), others relate to the scientific training of the examiner (reading the “prior art”) and others require the exercise of legal knowledge (reading and construing claims). The latter two sets of practices will be considered in more detail here.

Applying the SFN concept to the training and practices of examiners suggests specific sets of nodes. These would include hubs that accord with their scientific expertise, nodes linked to the skills learned with respect to the tests of patentability, for their employment generally and other skills that allow for the acquisition of discursive practices and for the communication of practices – the nodes are evident in all subjects who learn and who interact with others. it is assumed here that an examiner has acquired discursive practices associated with their scientific training (otherwise they would not have gained their qualification); the focus, therefore, is on the learning of practices associated with the legal aspects of examination. Of particular value is the benefit of the idea to considering the processes of change in acquired practices; processes that are best demonstrated through a consideration of the function, and practices, of a sub-category of examiners: opposition hearing officers.

B. Legal, Scientific, Administrative Practices as Sets of Nodes

Key examples of patent examination practices are the skills required to apply the legal tests of patentability to patent applications in order to assess whether a patent should be granted for the specified invention. One of the significant legal tests is whether the invention that is the subject of the patent application demonstrates a sufficient “inventive step”. Under the European Patent Convention (EPC), the international agreement that underpins patents granted by the EPO, the test for inventive step, in Article 56, is that ‘an invention shall be considered as involving an inventive step if, having regard to the state of the art, it is not obvious to a person skilled in the art’. This test, therefore, is judged against the “state of the art” in the field of the invention; with the art being the published expressions of knowledge in the area – including journal articles and previously granted patents. As such, the training that the examiners have had in their particular scientific discourse is necessary in order to comprehend the intricacies of the prior art.

Another legal test that is acquired by examiners is that of novelty. Under Article 54(1) of the EPC, ‘an invention shall be considered to be new if it does not form part of the state of the art’. Therefore, the examiners need to consider the state of the art associated with a particular patent application in terms of the tests of both novelty and inventive step – with the state of the art base being, to an extent, different for the two tests. The state of the art may indicate, for example, that the invention may be novel but does not incorporate a sufficient level of inventive step to attract patent protection. Therefore, the examiner must read the prior art with these two distinct tests in mind.

The tests for inventive step and novelty require a significant amount of legal understanding. The tests have been argued over, and refined, by lawyers and appeal boards over the course of the life of the EPC (with the tests in individual countries based on case law that goes back over a century). Examiners must do more than just know the words of the convention (or legislation) – patents granted by them have to comply with the tests as understood by legal experts, otherwise, the patents will be challenged in the courts. That is, the tests need to be fully inculcated and not merely “parroted” without intellectual engagement as an examiner’s use of the tests will be assessed by experts in their application (patent lawyers) when patents are subject to revocation proceedings (or to opposition proceedings). That the vast majority of patents are not subject to actions for revocation indicates that the acquisition of the legal practices by the examiners is effective – otherwise, a significant percentage of patents granted would be successfully challenged in the courts.

Further, there are benefits to considering the acquisition of these practices in terms of nodes in a network. If the practices associated with the application of the tests are seen as nodes, then, each time a test is used it is connected (in the examiner’s network of practices) to the specifics of the application to which it is applied. The more times the test is applied, therefore, the more connections there are to the node and the more deeply embedded it is in the network. The more connections there are, then, the harder it is for the practice to be “removed” from the network and the more important the practice is to the overall constitution (sense of self) of the examiner.

A concrete example of “growth” of the network of practices arises from the manner in which the legal tests of patentability may change. Case law, legislation and even international agreements are not set in stone. A judge, a panel of judges or a legislature may amend the wording of one (or all) of the tests in order to account for a change in technology or policy. In order for the new statement of law to take effect, it has to be understood, and acquired as a practice, by patent examiners. This may take the promulgation of a regulation (if the change is legislative) and/or an amendment to the Manual of Practice and Procedure. If the change is brought about through a court decision (see Dent and Cook 2007 for a discussion of the discursive limits of a judge producing a new statement of law) then the judgment will, usefully, explain where the previous law was wrong and how the new test relates to the previous test. If not, the authors of the Manual will have to interpret the new legislation or regulation for the consumption of the patent examiners.

If a specific example is taken, a new test for novelty is not likely to be radically different from earlier incarnations, in part because the policy justifications for patents require that an invention be new (Bently and Sherman 2004, 327-329). A new test may, therefore, only change the categories of knowledge that are included in the definition of prior art. Such a change could be the inclusion of certified translations of patent documents from certain patent offices where those translation, previously, had not been included in the definition. If the understanding of the first test is seen as a node in the network, the practice associated with the new test will extend out from that node as the new practice will both replace and be based upon the previous practice. That is, the idea that there are limits placed on the category “prior art” will remain, however, the new practice will include a wider (or narrower) set of documents to be searched in order for the proper conduct of the test/practice to carried out.

The practices/nodes associated with the legal tests would necessarily be linked to practices associated with their scientific training. The practices connected to the science qualification would depend on the area of science involved – a chemist would have acquired skills and knowledge that are different to those learnt by those trained in applied photonics. The examiners, for example, will apply the tests to the patent application before them and read the claims of the application against previous publications in the relevant area of technology. It is in the comparison of the application to the prior art that the examiner will exercise their mental practices associated with their scientific background. There would also be practices associated with the finding of the relevant prior art related to a particular invention. These practices would include those relating to the searching of databases and those that developed “on the job” in terms of the best search strategies to adopt. Of course, it is through the daily repetition of the practices of examination that they are reinforced as part of the examiner’s professional constitution and the practice becomes more connected with other discursive practices.

The search for, and interpretation of, the prior art relevant to a patent application under examination demonstrates the role of an examiner’s technical expertise; the process also provides an example of the continued development of that expertise. An examiner, after all, loses their value to the patent office if they do not stay current in terms of their technical knowledge. Each granted patent becomes a new item of prior art that represents the latest in technical knowledge with respect to that invention; every journal article that is cited (for example, during a challenge to the validity of the application) includes a current academic perspective of the state of play in the area. The reading of these accepted statements of knowledge[10] adds to the examiner’s understanding of the state of the art. Further, as the description contained in the patent should be sufficient for someone ‘skilled in the art’ to produce the invention (Bently and Sherman 2004, 354); it is possible that the examiner would be able to construct the invention on the basis of the information in the patent documents.

In terms of the nodes in the examiner’s network of examination practices, the knowledge captured in each new piece of prior art may also be seen as a new node. It may represent the intersection of two previously known, but unconnected, expressions of knowledge (where new prior art is a patent that’s inventive step involved the bringing together of two distinct innovations); it may also be the new recognition of therapeutic benefit that arises from a sub-group of a previously category of chemicals (known as a “selection patent” – Bently and Sherman 2004, 466-468). These new expressions of knowledge are not learnt in isolation from practices that had been taken on board by the examiner in the past; these new expressions of knowledge are, necessarily, connected to the examiner’s understanding of the state of the art. Further, once these new practices have been learnt, they may be used and re-used when later patent applications in that area of technology are examined.[11]

The second set of specific examination practices are more administrative and include the decisions that have to be made with respect to the application and the paperwork associated with the processing of the application. The former decisions relate to whether the application is complete and the relevant documents have been filed by the applicant. The paperwork that needs to be completed by the examiner includes examination reports and documents to be sent to the applicants informing them of the success or failure of the application. The skills needed for these actions are linked, for the sake of organisational efficiency, with the other bureaucratic practices acquired in their roles as public sector employees.[12] Underlying many of these practices, and those arising from an examiner’s scientific and legal training, are basic practices such as reading, writing and communicating orally. These practices are, of course, repeated when an examiner considers the application and the prior art. It is further arguable that the nodes relating to reading and communicating would be amongst that have the most connections within the networks of acquired practices of a “modern” educated citizen.[13]

The benefits of the conceptualisation of practices in terms of an SFN are further demonstrated when the gaining of experience, as a patent examiner, is considered. In the context of patent examiners, this may discussed in terms of hearing officers. These officers adjudicate opposition proceedings that accommodate the opposition of patent applications by third parties without the need for those parties to go to court to apply for the revocation of a patent. These officers are senior examiners who, through a demonstration of their understanding of the tests (the embeddedness of the practices), are given the role. Over time, then, the nodes relating to the tests “grow” connections the more the practice is repeated; when an examiner is made a hearings officer, the nodes relating to the prior art cease to grow as many connections (as they are spending less time on examining applications). This slowing of connections, however, is matched by new sets of practices (nodes) that are linked to new acquired practices that relate to the conduct of opposition hearings and the adjudication between the two interpretations of the application of the legal tests as argued by the patent attorneys representing the patentee and the party opposing the patent. With time, the nodes of practices relating to the conduct of hearings grow connections and, with this expertise and seniority, the officer may be given more senior roles in the patent office (at which point the nodes relating to the application of legal tests will diminish in relation to newer, more current, nodes).

The key benefits, then, of using an SFN concept to understand the skills and training of patent examiners mirror the general benefits of the perspective enabled by the idea of the SFN for discursive understandings of society. The benefits are that it removes the need to understand examiners as subjects of a particular discourse – whether it be the scientific or the legal discourse. Further, the use of the concept privileges the role of key practices, such as those relating to learning and communication; acknowledges the connections between particular discursive practices and sets of practices, for example, linking the use of current art with both scientific and legal practices; and allows for the processes of change in terms of discursive practices.

V. Conclusion

This article has introduced the concept of SFNs to Foucault’s understanding of discourses and discursive practices in order to better conceive of the relationship between diverse practices within the broad conception of subjectification in a governmentalist mode. The use of patent examiners as a profession that operates at the point of intersection of the legal and scientific discourses demonstrated the value of an approach that accommodated a diversity of training in order to account for variations that can exist in subjects within a single discourse. Further, the concept allows for the subject’s own role in her or his training. The desire to learn a particular skill may be seen as comprising the interests of the subject (expressed in the form of acquired practices), her or his previous training and practices associated with acquisition. The ability to successfully undergo training can be readily understood in terms of past practices associated with learning. That is, the conceptualisation suggests that no two subjects will acquire the same practice in exactly the same way; again, as any new skill is learned in the context of those practices already acquired.[14] Each of these effects are caused, in part, by the subject’s active involvement in the learning of practices; or, in more Foucauldian terms, their active role in their own subjectification (Foucault 1983b, 208).

An SFN-based approach offers a new perspective for understanding practices that occur in multiple disciplines – that is, it offers a useful conceptualisation of the processes by which practices are ordered (and thereby acquired and expressed). The approach facilitates the viewing of the totality of acquired practices in terms of a topography – some practices stand out as fundamental to the constitution of the subject while others, of lesser importance, operate in the shade of these key practices. In any profession, or even any activity, different practices are going to be more fundamental than others. Further, there may be shared practices across different professions and different activities. The use of the perspective enabled by SFNs allows for the variation of importance in, and for the commonality across discourses of, particular discursive practices. The use of the concept, then, permits a more detailed reading of the practices and the relationships between practices than that allowed by other, more general, understandings of discourses and a more practice-based understanding of governmentality itself.


Ahmed Adel, Tim Dwyer, Seok-Hee Hong, Colin Murray, Le Song and Ying Wu. 2005. Visualisation and Analysis of Large and Complex Scale-Free Networks. In Proceedings of Eurographics-IEEE VGTC Symposium on Visualisation, ed. K. Brodie, D. Duke and K. Joy. DOI: 10.2312/VisSym/EuroVis05/239-246.

Barabási, Albert-Laszlo and Eric Bonabeau. 2003. Scale-free Networks. Scientific American. 288: 50-59.

Barron, Anne. 2005. “Foucault and Law”. In Introduction to Jurisprudence and Legal Theory: Commentary and Materials, ed. James Penner, David Schiff and Richard Nobles, 955-1034. Oxford: Oxford University Press.

Bently, Lionel and Brad Sherman. 2004. Intellectual Property Law. 2nd ed. Oxford: Oxford University Press.

Bernauer, James and Michael Mahon. 1994. “The Ethics of Michel Foucault”. In The Cambridge Companion to Foucault, ed. Gary Gutting, 141-158. Cambridge: Cambridge University Press.

Bové, Paul. 1980. The End of Humanism: Michel Foucault and the Power of Disciplines. Humanities in Society 3: 23-40.

Cockburn Iain, Samuel Kortum and Scott Stern. 2002. Are All Patent Examiners Equal? The Impact of Characteristics on Patent Statistics and Litigation Outcomes. National Bureau of Economic Research Working Paper 8980.

Dean, Mitchell. 2007. Governing Societies. Maidenhead: Open University Press.

Dent, Chris. 2003. Reflecting on Continuity and Discontinuity in “the Law” – An Application of Foucault’s Archaeological Method in a Reading of Judicial Decisions in Negligence. PhD diss., Murdoch University.

———. 2005. The Privileged Few and the Classification of Henwood v Harrison: Foucault, Comment and Qualified Privilege. Griffith Law Review 14: 34-60.

———. 2007. To See Patents as Devices of Uncertain (but Contingent) Quality: A Foucaultian Perspective. Intellectual Property Quarterly 148-163.

———. Forthcoming. Copyright, Governmentality and Problematisation: An Exploration. Griffith Law Review.

Dent, Chris and Ian Cook. 2007. Stare Decisis, Repetition and Understanding Common Law. Griffith Law Review 16: 131-150.

Dreyfus, Hubert and Paul Rabinow. 1983. Michel Foucault: Beyond Structuralism and Hermeneutics. 2nd ed. Chicago: University of Chicago Press.

Florence, Maurice. 1994. “Foucault, Michel, 1926 –”. In The Cambridge Companion to Foucault, ed. Gary Gutting, 314-319. Cambridge: Cambridge University Press.

Foucault, Michel. 1973. Birth of the Clinic. London: Tavistock.

———. 1977. Discipline and Punish. London: Penguin.

———. 1980a. “Two Lectures”. In Power/Knowledge: Selected Interviews and other Writings 1972 – 1977, ed. Colin Gordon, 78-108. New York: Pantheon.

———. 1980b. “The History of Sexuality”. In Power/Knowledge: Selected Interviews and other Writings 1972 – 1977, ed. Colin Gordon, 183-193. New York: Pantheon.

———. 1981. “The Order of Discourse”. In Untying the Text: A Post-Structuralist Reader, ed. Robert Young, 48-78. Boston: Routledge & Kegan Paul.

———. 1983a. “On the Genealogy of Ethics: An Overview of Work in Progress”. In Michel Foucault: Beyond Structuralism and Hermeneutics, ed. Hubert Dreyfus and Paul Rabinow, 229-252. 2nd ed. Chicago: University of Chicago Press.

———. 1983b. “Afterword – the Subject and Power”. In Michel Foucault: Beyond Structuralism and Hermeneutics, ed. Hubert Dreyfus and Paul Rabinow, 208-226. 2nd ed. Chicago: University of Chicago Press.

———. 1990a. History of Sexuality Volume One: An Introduction. London: Penguin.

———. 1990b. History of Sexuality Volume Three: The Care of the Self. London: Penguin.

———. 1991a. “Governmentality”. In The Foucault Effect: Studies in Governmentality, ed. Graham Burchell, Colin Gordon and Peter Miller, 87-104. London: Harvester Wheatsheaf.

———. 1991b. “Preface to the History of Sexuality Volume II”. In The Foucault Reader, ed. Paul Rabinow, 333-339. London: Penguin.

———. 1992. History of Sexuality Volume Two: The Use of Pleasure. London: Penguin.

———. 1994. Archaeology of Knowledge. London: Routledge.

Klemm, Konstantin and Victor Eguiluz. 2002. Highly Clustered Scale-free Networks. Physical Review E 65: 036123-1 – 036123-5.

Lake, David and Wendy Wong. Forthcoming. The Politics of Networks: Interests, Power and Human Rights Norms. In Networked Politics: Agency, Power and Governance, ed. Miles Kahler. Ithaca: Cornell University Press. Available from: (accessed February 16, 2009).

McNay, Lois. 1994. Foucault: A Critical Introduction. Cambridge: Polity Press.

Muckelbauer, John. 2000. On Reading Differently: Through Foucault’s Resistance. College English 63: 71-94.

Redner, Sidney. 1998. How Popular is your Paper? An Empirical Study of the Citation Distribution. European Journal of Physics B 4: 131-134.

Rizzolatti, Giacomo and Corrado Sinigaglia. 2008. Mirrors in the Brain: How Our Minds Share Actions and Emotions. Trans. Frances Anderson. Oxford: Oxford University Press.

Rose, Nikolas. 1999. Powers of Freedom: Reframing Political Thought. Cambridge: Cambridge University Press.

Steyvers, Mark and Joshua Tenenbaum. 2005. The Large-scale Structure of Semantic Networks: Statistical Analyses and a Model of Semantic Growth. Cognitive Science 29: 41-78.

Strogatz, Steven. 2001. Exploring Complex Networks. Nature 410: 268-276.

Tancredi, Laurence. 2005. Hardwired Behaviour: What Neuroscience Reveals about Morality. Cambridge: Cambridge University Press.

[*] Senior Research Fellow, Intellectual Property Research Institute of Australia. I would like to thank Ian Cook, Kieran Tranter, Jason Bosland and the anonymous referees their comments and/or support throughout the long road this piece has taken to publication. The research for this article was, in part, supported by an ARC Discovery grant (DP0666803 ‘...and by opposing, end them: A Comparative Examination of Opposition Processes in Patent Law’).
[1] Foucault, however, was less than clear about what constituted a “discursive practice”. A substantial part of his Archaeology of Knowledge (1994) was devoted to the analysis of “statements” and his ‘Order of Discourse’ (1981) examined discursive controls which may be seen to be discursive practices (the focus of Dent and Cook (2007) was the consideration of stare decisis as a set of legal discursive controls); further, Foucault’s later work (1990a), (1990b) and (1992) focused on the role of practices in the “care of the self”. Whether “statements”, “discursive controls” and “practices” are synonymous with “discursive practices” is a matter capable of significant, but perhaps unnecessary, academic debate. For the purposes of this article, discursive practices are understood to include specific actions (or inactions), thoughts (or non-thoughts) or words (or silence) that a discursive subject acquires and reproduces.
[2] This understanding is a reworked version of the text in Dent (forthcoming).
[3] The central nature of discursive practices to Foucault’s work was also emphasised in a short piece often attributed to Foucault himself. Two of the three methodological ‘principles’ necessary to his project were: a ‘return toward the study of the concrete practices by which the subject is constituted in the immanence of a domain of knowledge’ and ‘of appealing to “practices” as a domain of analysis, of approaching one’s study from the angle of what “was done”’ (Florence 1994, 317-8).
[4] From a more scientific perspective, the motivation to copy may be explained through the recent discovery of a “mirror neuron” that enables babies and people to copy the behaviour of others: see, for example, Rizzolatti and Sinigaglia (2008). That there are discursive practices associated with habituation beyond this neuron is suggested by the complexity of the learning processes adopted by people – learning high energy physics, for example, goes far beyond the simple copying of the behaviour of others.
[5] This is not to suggest that discourses and discursive practices do not change. The structure of a discourse, however, indicates that its internal cohesion emphasises continuity or discontinuity of practices; see Foucault (1981). Further, see Dent (2007) for a discussion of the role of inertia, or stability, in the perpetuation of discourses.
[6] This is different from the standard understanding of the connections of new nodes in an SFN where a new node prefers recently added nodes to a network. The two understandings may be aligned if it is considered that a new node prefers a recently active node. That is, in a citation network a recently added node is recently active (the publication has recently been written) and in a network of discursive practices, the node relating to learning, for example, would be active in the process of acquiring any new practice.
[7] Another view would be that the weight of practices associated with an already learnt one makes it difficult for a new practices to replace it. The difference between these two conceptions of discursive change rests on whether a new discursive practice would replace the previously acquired practice or whether it is another node that connects to the other practices around it. Either conception fits within the perspective enabled by the idea of SFNs. Further, the value of the application of the concept to the understandings of discourses and discursive practices is not adversely affected by choosing one conception of the other.
[8] That the concept of SFNs may be usefully applied to the web of discursive practices is demonstrated by the analysis of semantic networks using the same idea. Steyvers and Tenenbaum found ‘scale free structures in semantic nets’ – they found that the connections between words in a vocabulary reflected a power law distribution (2005, 73-74). As the choice of words by an individual is only one of a vast number of discursive practices that may be expressed by her or him then it is likely that the body of discursive practices overall may be seen in terms of an SFN.
[9] Given the international nature of the patent system, the practices undertaken by patent examiners in different countries are substantially similar. The Australian Manual is cited as it provides a clear, step-by-step description of the steps required to be undertaken – unlike the Manuals of the UK Intellectual Property Office and the European Patent Office.
[10] The prior art may be seen as “accepted” by the disciplines where the invention described in the patent has been categorised as both new and inventive by an examiner or where the journal article has passed a peer review process.
[11] This raises the issue of the relationship between practices that are expressed in the physical world and those that remain as activities of the mind – such as memory. For Foucault, “thought” ‘can and must be analysed in every manner of speaking, doing, or behaving in which the individual appears and acts as subject of learning, as ethical or juridical subject, as subject conscious of himself and others. In this sense, thought is understood as the very form of action – as actions insofar as it implies the play of true and false, the acceptance or refusal of rules, the relation to oneself and other’ (Foucault 1991b, 334-335). For the purposes of this article, then, memory, as thought, may be conceived of as another set of practices that function as nodes within a given subject’s network of acquired practices.
[12] There would also be a further order of nodes. These would be the very particular practices that relate to the minutiae of day-to-day examiner activity and would include the precise sequence of mouse clicks used to navigate around the patent administration system. It is important to recognise that these exist, however, this level of detail is unnecessary for most analyses of behaviour.
[13] The picture does, however, gets more complex when it is recognised that there different discursive practices associated with reading itself – whether for prior art, administrative documents or pleasure (Muckelbauer 2000).
[14] Further, this conceptualisation of discursive practices accords with Muckelbauer’s understanding of Foucauldian “resistance” as ‘simply the convergence of multiple and conflicting powers’: (2000, 79).

AustLII: Copyright Policy | Disclaimers | Privacy Policy | Feedback