European Journal of Education, Vol. 45, No. 3, 2010, Part I

A Web of Learning Opportunities

ejed_1440

481..493

BRITT ANDERSEN, GUNNAR REE & INGUNN SANDAKER Learning Education is learning, but not all learning is education. Learning situations are designed to promote behaviour that serves the designer. Learning organisations are commonly discussed in a way that suggests that learning is a good thing, and we ought to have more of it (Senge, 1990, Argyris & Schön, 1996). Used this way, the concept acquires normative value. In behaviour analysis, learning is seen as a relatively permanent behaviour change that results from reinforced practice, which strips the term of a normative value. Behaviour repertoires change continuously, and some change may be beneficial in the light of one set of goals, for instance the individuals’, but detrimental to those of the organisation which employs them. Treating learning as a selection process means that one must recognise the difficulty of controlling complexity and analyse learning situations through the lens of results, not intentions. Selection processes involve variation, replication, interaction with environmental factors and differential survival as the result of such interaction. In nature, this means that those best adapted to their environmental demands survive (Hull et al., 2001). In a behavioural repertoire, selection means that behaviour that is successful in a given set of circumstances is more likely to reoccur than unsuccessful behaviour — the consequences of the behaviour in interaction with its environment determine whether it is repeated (survives) or not (becomes extinct) (Pierce & Cheney, 2004). As pointed out by D.T. Campbell and his colleagues (Heylighen & Campbell, 1996; Campbell, 1994; Baum, 1999) it is unrealistic to view selection processes as automatically improving the substratum they work on, be it selection of genes or of behaviour. Selection processes are, by their nature, blind (random, not goal-oriented), and occur whether we like it or not. Some outcomes are desirable or undesirable, depending on the perspective of the observer. A third level of selection is observed when we interpret processes of organisational and social change as selectionist or evolutionary; a further detailed explanation is offered by Aldrich and Ruef (2006), Baum and McKelvey (1999), Bowles (2004), Hull et al. (2001) and Mesoudi et al. (2005), while, a shorter and more general analysis is provided by Axelrod and Cohen (2001). When planning and implementing strategies to design learning regions and promote lifelong learning, one may do well to adopt a selectionist, network perspective and use the robust body of knowledge represented by behaviour analysis. The ideal for a knowledge-based modern society would be a seamless, interactive web of learning opportunities where knowledge development and dissemination are integrated to provide optimal learning outcomes for students (or just learners), as well as research and development possibilities for researchers and teachers, and business opportunities for the private sector. The present discussion adopts the concept of the learn unit proposed by Greer & McDonough (1999): ‘The theoretical, educational research, and applied behaviour analysis literatures all converge © 2010 The Authors. Journal compilation © 2010 Blackwell Publishing Ltd., 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA.

482

European Journal of Education, Part I

on the learn unit as a fundamental measure of teaching. The theoretical literature proposes the construct of the interlocking operant and embraces verbal behavior, social interaction, and translations of psychological constructs into complex theoretical respondent-operant interactions and behavior-behavior relations’ (p. 5), regarding the interchanges between members of networks as such learn units. The learn unit is the successful, i.e. reinforced, interaction between the agents. Fundamental Changes in an Ever More Complex Society Formal education has traditionally provided learners at each level with skills and competences thought to be necessary for them to function as productive members of society. According to Lars Quortrup (2003), in a less complex society, it was not unreasonable to expect this to be for life; career tracks were mainly along a path that depended heavily on previous education and experience. However, the complex globalised society has changed this and workers are expected to adapt to changes, be flexible and competent in their use of technology, and expand on their basic skills as needed by the job. The primary determinant here may lie in the amount of change and the number of possible connections between agents, but a large quantitative increase may signal qualitative changes (Qvortrup, 2003). Qvortrup works from the assumption that the information society is ‘. . . another mechanism of structuration than in earlier societies: It is based on decisions, rather than on ex ante given principles’ (op. cit., p. 11). The information society makes traditional, front- end loaded education strategies partly obsolete and requires strategies for learning on demand and learning just in time. Learning on demand is ‘the application and deployment of just the right amount of training at just the right time to those who need to possess the knowledge or learn the skill (Cummings, 2001)’. Just-In-Time principles include swift reactions to market or technology variations and process monitoring to improve the end product (Hall, 1999). Modern society places a premium on skills and competences. Working life is increasingly complex and traditional career tracks are less common. Many job titles advertised in the average newspaper did not exist 10 or 15 years ago, while others have disappeared. The average person must now expect to change jobs several times. More and more, people expect to work in totally new fields and in very different tasks from those for which they trained (Longworth, 2003).This labour market demands flexibility and rewards those who are willing to change. Opportunities for learning include in-firm training and in-firm formal schooling and courses and programmes run by independent educational institutions. Selectionist Perspectives Sandaker (2009) discusses a selectionist perspective on change, describing a model to understand the processes in the strategies of design, harnessing complexity1 and evolutionary change. While the strategies of strict design and selection from random variation are rarely found in pure form, they are a useful model. A general view of selection is that it is consists of entities (genes, behaviour) and activities (replication and variation; interaction). It is the collective label for cyclical processes in which one can identify variability in the replication of some entity, a way for the entity to interact with critical factors in the environment, and differential survival of varieties of the entity as the result of this interaction (Darden & Cain, 1989; Hull et al., 2001; Machamer et al., 2000). © 2010 The Authors. Journal compilation © 2010 Blackwell Publishing Ltd.

Britt Andersen, Gunnar Ree & Ingunn Sandaker

483

The general mechanism of selection by consequences (Skinner, 1981; Donahoe, 2003) explains evolved complexity by a few basic processes that are substrate neutral and yield guaranteed results, or what Daniel Dennett refers to as algorithms (Dennett, 1995). This can explain the evolution of species by natural selection (Darwin, 1859/2004; Futuyma, 2005), of complex individual behavioural repertoires (Skinner, 1981; Pierce & Cheney, 2004) and of complex adaptive systems (Axelrod & Cohen, 2001; Campbell, 1994). Core behavioural concepts are reinforcement, schedules of reinforcement, extinction and selection. Organising for Complexity Organisational design is seen as a strategy to control behavioural variation. Limiting the variation constrains the possible repertoires that can be selected and controls deviation. Evolutionary change — blind variation and selective retention of variants favoured by the environment — is not goal-oriented and may lead anywhere. The compromise — acceptance of bounded rationality (Simon, 1978, Dec 8) as a phenomenon and complexity judo (Axelrod & Cohen, 2001) to harness the complexity of the organisation instead of controlling it — has implications for the design of Lifelong Learning Regions, trading on the scientific base of complexity theory, behaviour analysis, and network theory. An initial comment on the concept of learning may be appropriate here. A selectionist perspective on human social organisation will profit from regarding an organisation as a system, studying the lines of communication and points of change rather than the structures described in the organisational chart. This perspective must include the randomness of behavioural variations, the relative unpredictability of the relevant factors in the system’s environment, and the fact that humans make decisions based on experience with similar situations. A strong emphasis on organisational planning or design is the recipe for controlling as much of the members of the organisation’s behaviour as possible. Strict hierarchies and chains of command and communication are prominent features of this kind of organisation, which has distinct advantages in stable environments, but which may prove cumbersome when the world around changes. Burns and Stalker (1961) describe the mechanistic and organic organisation. Heylighen & Campbell (1996) and Baum & McKelvey (1999) provide research on evolutionary perspectives on organisational change, as do Aldrich and Ruef (2006). Sandaker (2009) analyse the elements in the selection of organisational behaviour with a matrix of outcomes of various processes based on the chosen strategies (Table I). A combination of selection and design would allow for the self-organisation of networks of agents (nodes) based on preferences for interaction. The agents would be learners and providers of knowledge, and it is not of particular interest whether they are individuals or institutions. The position taken in this article is that the actual interchanges in the network, rather than the intended channels of communication denoted by the institutional agreements or the organisational chart constitute the real web of learning opportunities. Networks — The Selection of Structures Both designed and non-designed networks are relevant for lifelong learning. Designed networks in this context are strategic alliances between organisations to achieve various goals. Khanna et al. (1998) define them as ‘voluntary arrangements between firms involving exchange, sharing or co-development of products, © 2010 The Authors. Journal compilation © 2010 Blackwell Publishing Ltd.

Complex adaptive system; “purpose” judged by the functional relation to environment

Hybrid organisations, like matrix organisations

Organisations, institutions; defined by intended purpose & goals?

Harnessing complexity

Design

Function

Selection

Process

Formal structures based on division of labour and chain of command

Communication patterns as emergent phenomena; scale-free (not random) networks selected by preferential attachment Mixture of “chain of command” and “web of influence”

Structure

Production lines, standard operations; process control by correcting deviation

Exploration of new solutions combined with the exploitation of old ones

Self-organizing; emergence; process control by selection

Characteristics

Combinations of formal and informal networks established and maintained by businesses, government, education and voluntary initiatives. Education offered by formal institutions, but adjusted to defined groups.

Voluntary “learning circles”, Red Cross

Examples — regions

Table I. Illustrates different processes in systems, characteristic features and examples (adapted from Sandaker, 2009)

484 European Journal of Education, Part I

© 2010 The Authors. Journal compilation © 2010 Blackwell Publishing Ltd.

Britt Andersen, Gunnar Ree & Ingunn Sandaker

485

technologies, or services’ (p. 293). Examples are business firms, educational institutions or public service institutions that form networks to approach the ideal of lifelong learning for the members of their organisations and thus by-pass the institutional barriers. Khanna et al. (1998) introduce the concept of learning alliances when the primary goal for the partners is to learn from each other. While formal institutions have innate interests in regulating interaction and communication between their members, social networks develop based on evolutionary or selectionist dynamics (Sandaker, 2009). Non-designed social networks develop from another starting point, though sharing of knowledge and experience is still central. They are formed between individuals independently of their organisational affiliation. Relations are based on any combination of common interests or points of contact. These networks are informal and exist side by side with the formal intra-organisational and/or interorganisational structures. Relations in these networks also occasion learning and may contribute greatly to lifelong learning outcomes. While recognising that strategic alliances are an important part of lifelong learning strategies, we focus on these informal social networks in the last part of the article. The importance of the structures, dynamics and robustness of networks for lifelong learning has not been sufficiently investigated. Which network member is directly or indirectly connected to which other network member constitutes its structure and the structure of any network has defining properties which may be expressed in generic measures (Barabási, 2003). Network structure (or topology) is one important determinant of how information is shared in a network and also determines its robustness. This means that they may maintain the same information flow even after perturbances or ‘errors’, such as members leaving the network. When social networks change, so does their structure — the changes happen when connections between members are added or severed. Social networks are Complex Adaptive Systems (CAS). This means that relations are numerous and the network (system) adapts to changes in its environment and to internal changes (Axelrod & Cohen, 2001). There is no general agreement on a definition of a network. One definition of social networks is ‘patterns or regularities in relationships among interacting units’ (Wasserman & Faust, 1994). The relationships (or links) represent transfer of resources, for example information, knowledge (Barabási, 2003) or consequences (positive reinforcement) on our behaviour, between the individuals (nodes) in the network. Network analysis then focuses on the relations and patterns of these relations between individuals. The patterns that emerge from the relations are the structure of the network, and hence the flow of resources (e.g. information and knowledge) in the network (Rogers, 2003; Rogers & Van Den Ban, 1963). An analysis of a social network can also include compositional variables, i. e. actors’ attributes (gender, position, salary, etc.) (Wasserman & Faust, 1994). The unit of analysis is determined by the aim of the study, and the entities can be a single individual (ego-centred), dyads (or triads) of nodes, a subgroup in the network, or the whole network (Wasserman & Faust, 1994; Scott & Davis, 2007). The mutual perceptions of reward in exchanges are the real links between the pairs of individuals in the network. How frequently such rewarding interchanges occur is an attribute of the event (Wasserman & Faust, 1994). The specific link cannot tell us anything qualitative about the link, just that there is a relation, and as such is dichotomous (present or not) and can be directional (three possibilities) © 2010 The Authors. Journal compilation © 2010 Blackwell Publishing Ltd.

486

European Journal of Education, Part I

or non-directional (op. cit). But the relation can also be valued and weighted (Gross & Yellen, 2006; Wasserman & Faust, 1994). The value indicates frequency, intensity, strength, or other values in the relations (Wasserman & Faust, 1994; Gross &Yellen, 2006) between each pair of nodes, i.e. frequency of presentation of reinforcement between each pair of individuals — learn units (Greer & McDonough, 1999), or how often a person A and a person B exchange information about a topic. A valued network (Wasserman & Faust, 1994) gives information about the nodes in the network, the links in the network, and the values associated with each link. It can have a non-directional valued relationship (e.g. the number of presentations of positive reinforcement between each pair of nodes) or a valued directed relation (op.cit) (which gives information about the direction of the reinforced presentation and its frequency). The line and corresponding value represent the flow (the line) and the capacity (e.g. frequency of positive reinforcement) (Gross & Yellen, 2006). Hence, there are two options when describing a network: (a) a network with just the nodes and the links which model its basic structure, and (b) a network with the nodes, links, and valued links which, in addition to the basic structure, also gives information about frequency, duration or intensity. Social networks change over time, and some more rapidly than others (Newman et al., 2006). The mechanisms that are basic to changes in social networks are not fully discovered (Barabàsi, 2005; Barabási, 2003). They have two sources: nodes are added or subtracted, links are established or severed. Changes in network structure conform to the general principles (Barabási, 2003) of a selectionist perspective. Understanding network structure requires knowledge of more basic processes in the formation of social units such as alliances, learning organisations, learning regions and learning teams. One such basic process of structural change described by Barabàsi (Barabási, 2003; Barabàsi et al., 2002; Barabàsi, 2005) is preferential attachment which explains how some nodes in a network acquire more links than others, turning them into network hubs. Investigations of such networks include work on co-authorship in different scientific fields (Guimerà et al., 2005), and more humorously, webs of actors, where a Bacon number is assigned to an actor according to his or her degree of separation from Kevin Bacon (http://en.wikipedia.org/wiki/ Bacon_number). Preferential attachment is an example of the rich-get-richer phenomenon, which means that a node with a large number of links is more likely to acquire an even larger number (Barabási, 2003). The selection mechanisms underlying this phenomenon are not well understood. It is reasonable to think that a number of processes may be involved in the formation and change in a network. Networks are important in entrepreneurship (Anderson et al., 2010), or ‘. . . for instituting change, developing growth and thus creating the future’. They provide access to resources such as information and knowledge, which are central to promoting change (Burt, 1992). This is relevant for the concept of learning organisations and learning regions. Information and knowledge flow are essential in a learning organisation. Features of the informal network structure in an organisation or between members of different organisations in a region can promote or prevent (obstruct) information and/or knowledge sharing. For example, in a highly dense network, the information will be rapidly disseminated (Scott, 2000), but it will not be very varied since its members are connected to each other and have access to much the same information (Burt, 2003). © 2010 The Authors. Journal compilation © 2010 Blackwell Publishing Ltd.

Britt Andersen, Gunnar Ree & Ingunn Sandaker

487

Structural properties of the exchange of ideas must be analysed in terms of networks, and the topological features of any network are closely linked to its function. Bearing in mind the reservations regarding the normative use of the concept of learning and the difficulty in controlling complexity, let us examine network properties that are conducive to the promotion of useful learning and education. These networks must be adaptive and robust and must be allowed to evolve and wither according to the usefulness perceived by the participants which can be viewed in the light of the behaviour analytic reinforcement theory, a selectionist discipline. Selection also informs theory building and research on Complex Adaptive Systems (Axelrod & Cohen, 2001). Reinforcement — The Selection of Behaviour A reinforcer is a stimulus which affects the probability of the kind of behaviour that produces it. There can be positive reinforcers: behaviour that produces stimuli that reinforce consequences will increase. There can also be negative reinforcers: behaviour that produces these stimuli decreases, or behaviour that removes them or postpones them increases. If behaviour does not generate important consequences, it is abandoned. Many reinforcing stimuli are unconditional and function as reinforcers without prior learning. This is commonly thought of as a product of natural selection and is closely related to the survival of members of our species (Catania, 2007). However, man is a social animal and thus extremely susceptible to social reinforcement. This kind of reinforcement is called conditional and has acquired its reinforcing properties through previous experience, after having been paired with unconditional reinforcers. Our biological properties are generally recognised as evolutionary consequences of our ancestral environment, but the response to social stimuli is also understood as a product of natural selection. Any stimulus or activity may be a reinforcer, and the individual variations in what will increase or decrease the behaviour which produces it are numerous (Skinner, 1953; Premack, 1959). Consider the prospect of jumping from the top of skyscraper just as it gets light with a small parachute on your back. Most people would do anything to avoid this, but a certain number will say that they live for such moments. The range of important social reinforcing stimuli which are effective with most people includes behavioural consequences such as praise, social recognition and attention from your superior or from other colleagues, or contacts whose opinions you value. Other important reinforcers can be success at work or in social situations (often accompanied by social reinforcement and primary or unconditional reinforcers) and achieving personal goals. The behaviour of others may be the most important source of reinforcement for modern humans by sheer quantity. Behaviour analysis is the science of quantifying relations between individual behaviour and events in the environment, stressing biological conditions, learning history and current conditions as the necessary variables.The major premise is that the environment is a crucial force in organising individual behavioural repertoires. One way of quantifying these relations is the analysis of reinforcement schedules, which are descriptions of the probability that a given behaviour is followed by reinforcement. These schedules may vary according to time and response frequency and whether they are fixed or variable They also generate different characteristic behavioural patterns, and behaviour maintained by variable response patterns are very resistant to extinction (Ferster & Skinner, 1957). © 2010 The Authors. Journal compilation © 2010 Blackwell Publishing Ltd.

488

European Journal of Education, Part I

Integrating Network Science and the Reinforcement Theory Integrating network science and a complex systems perspective with reinforcement theory, this section considers reinforcement frequency as a variable in establishing and maintaining social networks in formal organisations and informal social groups. Making contact with members of a social network may produce strong reinforcing consequences, and if this is the case, such contact is more likely to reoccur. If the contact is reinforcing for both members, bi-directional ties may be established. Discussing a practical problem with colleague A in another position in your organisation (or in another organisation) may be more effective to solve your professional problem than discussing it with colleague B; this may be the case even when the organisational chart (rules) suggests that you should discuss it with B. Generalisations regarding one observed lawfulness in the relations between behaviour and environmental events were expressed by Richard Herrnstein in a 1961 paper The Matching Law. The generalised matching law (Herrnstein, 1961), with modifications suggested by McDowell (2005), may be stated thus: in a performance involving concurrent operants (the technical term for the fact that the individual may make choices between simultaneous opportunities to perform types of behaviour that both (or all) have reinforcing consequences), individuals tend to distribute behaviour (choices) in such a way that the relative response rate matches its probability of being reinforced. A typical experimental situation involves schedules of reinforcement with different simultaneous properties, with the experimental subject free to distribute his responding between the experimental manipulanda in the chamber, and thus exposing his behaviour to the two reinforcement contingencies in operation. In a typical real-life situation, the couch potato watches a boxing and a football match on two different channels and zaps according to the pleasure derived from what happens in each match. The ratio between effort and payoff is optimal with perfect matching. Being part of a social network can be seen as analogous to being exposed to concurrent operants. Depending on with whom you connect, the probability of your behaviour being reinforced varies. Basic behavioural motivation theory states that behaviour must be reinforced to keep occurring, and so the connections in your network that are most likely to provide reinforcement are those which you seek out most frequently. Quantification of real life situations in these terms may be problematic, but it is easily achieved in a laboratory. If your behaviour in contacting A is reinforced and you do not experience reinforced contacts with B, your behaviour will be selected by its consequences and the ties to A will be strengthened. Maybe you learn that for one kind of assistance B is the right person, whilst for other kinds of help you should turn to A or C; you will distribute your behaviour accordingly, optimising the returns on your efforts (Høyer, 2005). Social networks, like other networks, are functional entities and directionality of contacts may be seen as indicative of histories of reinforcement for the members in the network. If contact is unidirectional, it is not unreasonable to assume that one member (the initiator of the contact) has a history of reinforcement for making contact. We are not free to make assumptions about the reinforcement history of the other node. Not initiating contact does not necessarily indicate that contact is not reinforcing; maybe the number of contacts initiated by the first member is © 2010 The Authors. Journal compilation © 2010 Blackwell Publishing Ltd.

Britt Andersen, Gunnar Ree & Ingunn Sandaker

489

optimal for the mutual reinforcement of contact. On the other hand, bi-directional contacts would suggest strong mutual reinforcement, especially if these occur outside formal organisational channels or chains of command. Networks exist as a matter of function: you contribute or you are excluded. There are several ways in which this may be important. The value of ties between members, measured by the number of contacts between them, is a measure that is easily quantified. A more sophisticated analysis would consider the quality of the contacts and attach importance to the question of uni-directionality or bi-directionality. Whether the contacts occur within a relationship that has been formally defined or within a slef-organised system within the formal institution is another aspect. Robust and independent sub-cultures may self-organise, selecting behaviour that promotes the goals of the members, while undermining those of the larger system or the formal organisation (Campbell, 1994). An important characteristic of a selection process is its essential mindlessness, as Dennett (1995) put it. Campbell (1994) emphasises this, reminding one that selection is not goal-directed but blind and that not all evolution by selection is for the better (however one decides what better is). Selection means adapting to yesterday’s environment. A radical change in the environment of today may render a species unfit for survival (no more dinosaurs), a larger part of an operant repertoire useless (extinction through the removal of reinforcing consequences that previously maintained the behaviour), or a sub-culture mainstream (punk rock as a commercial commodity certainly took the fun out of the safety pin). The following example may serve to illustrate the point: a few years ago, large organisations had their own typing pools. These have vanished, like many clerical functions that provided practical assistance to department managers. The number of such functions that remain is drastically reduced because managers do more of this work themselves and the tasks have become easier to perform thanks to better information technology. But in higher education institutions bureaucracy is growing rapidly because of the increasing demand of governing bodies for result reports and control. However, in addition to these tasks that are imposed by external forces, bureaucracy seems to evolve according to its own logic. Personnel departments grow and appropriate for themselves tasks that go beyond dealing with their original legal and fiscal responsibilities. A department of administration, which was initially designated to deal with practical matters concerning student affairs, performs functions that go beyond this. Originally, in a line organisation, staff functions were defined by management needs. Lately, auxiliary departments seem to define the needs of the primary departments — those concerned with education, research and knowledge dissemination, commonly considered the core activities of a university — independently of the faculty that perform these core activities. This shows how some functions do not survive when they must interact with changing contingencies, while other functions survive and evolve independently of the core activities if the contingencies allow it. In complex networks, topological features may be important for how the network works — structure affects function in various complex ways. Structural/ topological properties may give rise to emergent properties of the network. When new members make contact, they may generate new kinds of reinforcing consequences for the behaviour of others in the network and the distribution of contact may change accordingly, with new patterns replacing the old. Members may become cut off (contact with them is not reinforcing for others) or they may © 2010 The Authors. Journal compilation © 2010 Blackwell Publishing Ltd.

490

European Journal of Education, Part I

redistribute their efforts and keep up contact with only some members of their former network. The structure of the network may determine what it may achieve as a system, as well as what the individual members may achieve. Networks (systems) that are useful for their members survive as entities for longer periods than those that are not. New members may be substituted for older members over time, but defining characteristics of the network may remain, making it recognisable as a functional entity. Selection and Learning Regions In a description of the learning region, Richard Florida (1995) states that ‘Regions are becoming focal points for knowledge creation and learning in the new age of global, knowledge-intensive capitalism, as they in effect become learning regions. These learning regions function as collectors and repositories of knowledge and ideas, and provide the underlying environment or infrastructure which facilitates the flow of knowledge, ideas and learning’ (p. 527).This description has a different emphasis from the concept used in the EU literature on the subject. The EU has put great effort into the development of learning regions. The 6th Framework programme financed a number of projects dedicated to lifelong learning and learning regions and it is clear that the EU has aspirations of designing learning regions, rather than waiting for them to evolve. This is illustrated by Hassink (2005) who describes a learning region as ‘a new theory-led regional development concept which aims at achieving and/or supporting collective learning processes’. This adds another important factor. The term theory driven approaches means that a coherent framework of general principles is initiated when planning and implementing a specific intervention. This, in turn, requires benchmarking in order to measure change along some previously decided scale of measurement. The evolutionary approach of selectionism is such a framework, and CAS theory, network theory and behaviour analysis are all selectionist approaches. This means that they enable the integration of data across levels of analysis. Coherent conceptual frameworks make translating general principles into specific interventions easier. Among the intervention tools of behaviour analysis, instruments for promoting change include those for keeping track of change. This knowledge is important when designing interventions of a certain scale, and the insights from network theory and behaviour analysis integrated with a complex adaptive systems approach have a clear advantage over non-behavioural models.The functional analysis of behaviour has a unique strength when designing interventions. If the topology of networks provides information on their functioning, it should be put to good use. Conclusion In a globalised, ever more porous society, one can benefit from knowledge about how people connect independently of formal boundaries. An institutional approach to lifelong learning and regional development may fruitfully be supplemented by knowledge from functional behaviour analysis and the science of networks. NOTE 1. Sandaker (2009) treats this as bounded rationality. For the sake of clarity, it is changed here to harnessing complexity. © 2010 The Authors. Journal compilation © 2010 Blackwell Publishing Ltd.

Britt Andersen, Gunnar Ree & Ingunn Sandaker

491

REFERENCES Aldrich, H. D. & Ruef, M. (2006) Organizations Evolving (Thousand Oaks, Sage). Anderson, A. R., Dodd, S. D. & Jack, S. (2010) Network practices and entrepreneurial growth. Scandinavian Journal of Management (In Press). Argyris, C. & Scho¨ n, D. (1996) Organizational Learning II: theory, method and practice (Reading, Addison Wesley). Axelrod, R. & Cohen, M. D. (2001) Harnessing Complexity (New York, Basic Books). Baraba´ si, A.-L. (2003) Linked. How Everything is Connected to Everything Else and What it Means for Business, Science, and Everyday Life (New York, Plume). Baraba` si, A.-L. (2005) Network theory — the emergence of the creative enterprise, Science, 308, pp. 639–641. Baraba` si, A.-L., Jeong, H., Neda, Z., Ravasz, E., Schubert, A. & Vicsek, T. (2002) Evolution of the social network of scientific collaborations, Physica A, 311, pp. 590–614. Baum, J. A. C. (1999) Whole-Part coevolutionary competition in organizations, in: J. A. C. Baum & B. McKelvey (Eds) Variations in Organizational Science. In Honor of Donald T. Campbell (Thousand Oakes, Sage). Baum, J. A. C. & McKelvey, B. (Eds) (1999) Variations in Organizational Science. In Honor of Donald T. Campbell (Thousand Oaks, Sage). Bowles, S. (2004) Microeconomics, Behavior, Institutions and Evolution (Princeton, Princeton University Press). Burns, T. & Stalker, G. M. (1961) The Management of Innovation (London, Tavistock). Burt, R. (2003) The social structure of competition, in: R. Cross, A. Parker & L. Sasson (Eds) Networks in the Knowledge Economy (New York, Oxford University Press). Burt, R. S. (1992) Structural Holes.The Social Structure of Competition (Cambridge, MA, Harvard University Press). Campbell, D. T. (1994) How individual and face-to-face selection undermines firm selection in organizational dynamics, in: J. A. C. Baum & J.V. Singh (Eds) Evolutionary Dynamics in Organizations (New York, Oxford University Press). Catania, A. C. (2007) Learning (Cornwall-on-Hudson, Sloan Publishing). Cummings, E. (2001) Relearning E-Learning, Darwin Magazine. Darden, L. & Cain, J. A. (1989) Selection type theories, Philosophy of Science, 56, pp. 106–129. Darwin, C. (1859/2004) The Origin of Species (London, Collector’s Library). Dennett, D. (1995) Darwin’s Dangerous Idea (London, Allen Lane, The Penguin Press). Donahoe, J. W. (2003) Selectionism, in: K. A. Lattal & P. N. Chase (Eds) Behavior Theory and Philosophy (New York, Kluver Academic/Plenum Publishers). Ferster, C. B. & Skinner, B. F. (1957) Schedules of Reinforcement (New York, Appleton-Century-Crofts). Florida, R. (1995) Toward the learning region, Futures, 27, pp. 527–536. Futuyma, D. J. (2005) Evolution (Sunderland, Sinauer Associates). © 2010 The Authors. Journal compilation © 2010 Blackwell Publishing Ltd.

492

European Journal of Education, Part I

Greer, R. D. & Mcdonough, S. H. (1999) Is the learning unit a fundamental measure of pedagogy? Behavior Analyst, 22, pp. 5–16. Gross, J. L. & Yellen, J. (2006) Graph Theory and its Applications, (Boca Raton, FL., Chapman & Hall/CRC). Guimera` , R., Uzzi, B., Spiro, J. & Amaral, L. A. N. (2005) Team assembly mechanisms determine colloboration network structure and team performance. Science, 308, pp. 697–702. Hall, L. (1999) Just-in-time learning: Web-based/Internet delivered instruction. American Conference on Information Systems (AMCIS 1999 Proceedings, Paper 319). Hassink, R. (2005) How to unlock regional economies from path dependency? From learning region to learning cluster, European Planning Studies, 13, pp. 521–535. Herrnstein, R. J. (1961) Relative and absolute strength of response as a function of frequency of reinforcement, Journal of the Experimental Analysis of Behavior, 4, pp. 267–272. Heylighen, F. & Campbell, D. T. (1996) Selection of organization at the social level: Obstacles and facilitators of metasystem transitions, World Futures, pp. 181–212. Høyer, H. C. (2005) Den daglige forvaltningspolitikken — en studie av mindre organisasjonsendringer på regionalt forvaltningsnivå i Norge. Institutionen for samhällsvetenskap — avdelingen för statsvetenskap (Karlstad, Karlstad Universitet). Hull, D. L., Langman, R. E. & Glenn, S. S. (2001) A general account of selection: biology, immunology and behavior, Behavioral and Brain Sciences, 24, pp. 511–527. Khanna, T., Gulati, R. & Nohria, N. (1998) The dynamics of learning alliances: Competition, cooperation, and relative scope, Strategic Management Journal, 19, pp. 193–210. Longworth, N. (2003) Lifelong Learning in Action — Transforming Education for the 21st Century (London, Taylor and Francis). Machamer, P., Darden, L. & Craver, C. F. (2000) Thinking about mechanisms, Philosophy of Science, 67, pp. 1–25. McDowell, J. J. (2005) On the classic and modern theories of matching, Journal of the Experimental Analysis of Behavior, 84, pp. 111–27. Mesoudi, A., Whiten, A. & Laland, K. N. (2005) Towards a unified science of cultural evolution, Behavioral and Brain Science, 29, pp. 329–347. Newman, M. E. J., Baraba´ si, A.-L. & Watts, D. J. (Eds) (2006) The Structure and Dynamics of Networks (Princeton, Princeton University Press). Pierce, W. D. & Cheney, C. D. (2004) Behavior Analysis and Learning (Mawah, Lawrence Erlbaum Associates). Premack, D. (1959) Toward empirical behavior laws: I. Positive reinforcement, Psychological Review, 66, pp. 219–233. Qvortrup, L. (2003) The Hypercomplex Society (New York, Peter Lang). Rogers, E. M. (2003) Diffusion of Innovations (New York, Free Press). Rogers, E. M. & Van Den Ban, A. W. (1963) Research on the diffusion of agricultural innovations in the United States and the Netherlands, Sociolocica Ruralis, 3, pp. 38–51.

© 2010 The Authors. Journal compilation © 2010 Blackwell Publishing Ltd.

Britt Andersen, Gunnar Ree & Ingunn Sandaker

493

Sandaker, I. (2009) A selectionist perspective on systemic and behavioral change in organizations, Journal of Organizational Behavior Management, 29, pp. 276– 293. Scott, J. (2000) Social Network Analysis. A Handbook (London, Sage Publications). Scott, W. R. & Davis, G. F. (2007) Organizations and organizing. Rational, natural, and open system perspectives (Upper Saddle River, NJ, Pearson Education). Senge, P. M. (1990) The Fifth Discipline. The Art and Practice of the Learning Organization (London, Random House). Simon, H. (1978) Rational decision-making in business organizations, Nobel Prize Lecture, Stockholm. Skinner, B. F. (1953) Science and Human Behavior (New York, The Free Press). Skinner, B. F. (1981) Selection by consequences, Science, 213, pp. 501–504. Wasserman, S. & Faust, K. (1994) Social Network Analysis. Methods and Applications (Cambridge, Cambridge University Press).

© 2010 The Authors. Journal compilation © 2010 Blackwell Publishing Ltd.