Towards Learning Audit In Viable Learning Systems

Towards Learning Audit in Viable Learning Systems Presented at the International Conference in Computer Based Learning in Science (CBLIS), 1995 M.I.Yolles, Liverpool John Moores University Abstract: Many learning situations offer what, to novice learners, seems to be complex learning material. With the development of Web based courseware, we are seeing more interest in learning approaches that are remote from course providers. It therefore seems a good time to consider the need for non-retrospective automatic auditing systems that can enable learners to recognise whether their learning capability may be diminished. Unfortunately, such systems are not currently feasible because existing learning theory is not sufficiently mature to permit us to model them. A possible approach based on viable systems theory is explained. It comes from the field of management system, and has application to the notion of learning organisations, and links with complexity theory. Keywords: complexity theory, remote learning, organisational learning, viable systems, management systems. 1.

Introduction

Viable systems are those that are able to survive in complex situations even though they may experience perturbing chaos. They maintain their structures and processes through selfregulation, and adapt to changes through self-organisation. They experience tension, which is seen to represent the source of adaptation. When tension becomes exessive, it is a stress that interferes with the adaptability of the system. While these ideas are applied to organisations, they derive from explorations in biology. It can be argued that people too are viable systems. This relates to all forms of their behaviour, including learning. Thus, for instance, stress can adversely affect the learning rates of novice learners7. New learning domains can, for many novice learners, often be seen as complex, while for those who know the domain, it is not. Complexity is a subjective phenomenon. In complex learning situations, novice learners are unable to understand all of the details that surround them. Chaos occurs here with bounded learning instability, and this occurs when learners are exposed to unexpected entities that are manifested within, or impact on, the situation. It can perturb continued learning. This can be serious for remote learners – those who are geographically (as in distance learning), or socially (as in very large classes) remote from a course provider, where tutor led remedial action is not feasible. Learners may deal with this by eventually seeing “emergent” concepts that can simplify the situation for them. These concepts do what Cohen and Stewart10 call collapsing chaos. An automated computer based system that is able to recognise and warn remote learners of changes in their learning capability can therefore be useful. However, learning theory is not yet sufficiently mature to enable such a system. This paper will propose some underlying theory for this that comes from viable systems theory. The subject is part of management systems, a field that applies systems principles to organisational management. It has interests in organisational learning, and links some of its developments to connectionism, artificial intelligence, cognitive psychology, chaos theory, and cybernetics. Our interest in learning theory derives from Yolles29. It centres on three ideas that have developed since the concepts of programmed learning in the 1960s. These are: the notion that learning behaviour exists, and can be associated with a style; that learners have learning strategy; and that learners can learn to learn (metalearn). These concepts can be expressed in terms of

Viable Systems Theory (when we can refer to them viable learning systems), and explored in terms of complex situations. 2. Learning Behaviour, Strategy, and Metalearning Learning behaviour is a pattern or schedule of learning actions. Kolb14 proposed that learning behaviour can be described as a continuous learning cycle (Figure 1) that explains the experiential stages that learners pass through during learning. Learning cycle approaches can be a useful simplification of learning behaviour. However, Kolb’s work was deficient in a number of ways11: (i) (ii) (iii)

as a reaction against theory based learning, there can be an over-value of experience based learning; learning can be perceived to be based entirely on what others hand down and by selfreasoning; research evidence does not support learning cycle theory: e.g., the cycle does not indicate how learners gain ability in learning.

E x pe rie n cial le arning qu adrant A ctive exper im en tation

C on cr ete experien ce

R e fle c tiv e le arnin g quad rant Reflective obser va tion

A c tiv e le arning quad ra nt

A bstrac t le arn ing qu adran t A bstra ct con cep tua lisa tion

Figure 1: The Kolb Learning Cycle

Now, Kolb’s phases of behaviour are not defined in terms of more detailed subsidiary behaviour. Perhaps this is because his interest did not lie in learning behaviour as such, but rather in the associated concept of learning style. In learning cycles, learning behaviour may be seen as a constrained set of activity steps. Let us consider this a little further by illustration. In many traditional paper-based open learning texts, a programme of work is structured into units. Each might begin with the objectives of the learning materials, have summaries, embed examples and have a glossary. If a tutor adheres to a rigid learning strategy, then s/he will devise a set of very tight constraints on learning strategy. This will define a unique behavioural schedule permitting only one possible way for a learner to pass through this material. Thus for example, a sequential ordering of objectives, summaries, and so on may occur that defines only one learning path for a learner. This may not be consistent with the personal strategy of a given learner who may wish, for instance, to sample parts of the learning material or consult the glossary before embarking on the learning material. This is what Crampes34 would refer to as a phase of discovery that is personal to the learner. Perhaps a more flexible approach is for a tutor to offer a menu that enables the learner to define his or her own strategy of learning within less tightly tutor defined constraints, for example presenting materials on a Website. Therefore, the degree of flexibility on learning strategy offered to a learner is tutor determined through the constraints imposed on the learning material.

While learning can occur through circumstance rather than intention, more often than not it occurs through the creation of learning purpose, especially in formal pedagogic settings like universities. Learning purpose is associated with learning strategy. It is often expressed in terms of goals that are to be achieved, and as such, it can be seen that learning strategy is to do with the controls and constraints that enable these goals to be achieved. Examples of control features are a content index and a content map. Laurillard18 argues that constraint minimisation provides better results in learning than its maximisation. Constraints are apparent when learning goals are predefined for a learner, rather than allowing them to be learner-defined. Consistent with the ideas of Laurillard and others18, 20, 28, 29, an argument will be put that learning strategy has associated with it not only control and constraint processes, but in addition both cybernetic and rational processes. As part of this, learning strategy is also connected to metalearning and learning behaviour21, 22, 26. These ideas are not prevalent in learning theory, but rather derive from viable systems theory30, that can be used to construct a theory of viable learning as we shall explore later. From this, it is propositional that change in learning strategy will likely result in learning behaviour change. It is axiomatic that learners have cognitive processes that are affected through metalearning (or cognitive learning about learning). Indeed, a base notion in management systems is the notion that organisational learn through “double loop learning”1, when cognitive processes are changed in some way. According to the principles of viable systems theory, cognitive learning is connected to the worldview of a learner, as discussed in the next section. Worldviews are ultimately responsible for the nature of the manifest strategy that a learner will adopt, and changes in that learning strategy. It also ultimately defines the behavioural schedule that a learner adopts within the learning process. When a learner is faced with new learning material, s/he may redefine what constitute the elements of a learning domain by creating new (perhaps personal) conceptualisations, thus explicitly influencing the nature of a learning strategy. It represents another feature of cognitive learning that is capable of dealing with complex learning situations. This in turn is likely to affect the behavioural schedule or sequencing of learning material in the learning domain. It can enable new maps of learning material to be created, based on the new conceptualisations. It can result in new logical strategic processes that sequence learning material, and can result in the definition of new rules of learning to provide a way of overcoming highly constrained learning domains. Once learning purposes have been expressed in terms of explicit or implicit goals, part of strategy is to achieve them. If learners are to achieve their learning goals, then it may not be sufficient to be motivated and have a good learning strategy. They also need to learn how to learn. The nature of how this can be achieved will be determined by the paradigm to which learners attach themselves. Learning about learning (metalearning) enables learners to: (1) (2) (3) (4) (5)

have knowledge to reason about its own operation; have a structure which simplifies the reasoning process; infer conclusions from a chain of inference rules; determine accuracy, consistency, and plausibility of its conclusions; explain reasoning behaviour.

Metalearning induces learners to assess the patterns by which they learn11. 3.

Worldviews

It is of interest to realise that pedagogic perspectives involve different worldviews. The notion of worldview is a literal translation from the German weltanschauung that derives from 24. Worldviews are relative to the institutions that one is attached to in a given society, and they

change as the institutional realities change3. More recently28, it has taken on the meaning of a view or perspective of the behavioural world. This is determined by cultural attributes (that includes language, and cognitive organisation composed of attitudes, values, and beliefs), norms, and a cognitive space that defines concepts and knowledge. Through a process of socialisation, a view is formed within the institutions one is attached to, and they change as the institutional realities change. Worldviews may be shared by a group of people, though when this occurs the individuals each retain their own realities while using common models to share meaning. Further, worldviews have boundaries that are generated within the belief system and cognitive space of their viewholders, and as a result we can explore worldviews in terms of their knowledge attributes. We can distinguish two forms of worldview for the system domain: weltanschauung and paradigm. The term weltanschauung was originally used by Churchman6 within the systems context. It has been given a distinct definition by Checkland4 and Checkland and Davis5 when used as one of the cornerstones of his own systems methodology directed at solving problem situations that involve purposeful human activity. The use of the word by Checkland has been variously defined as the worldview:   

that determines model building of relevant systems and conceptual models; for which in a particular situation certain notional systems are seen as relevant; the perceived social reality of the situation in which the study is made.

It has also been defined27 as the worldview of an individual or the shared worldview of a group that is more or less visible to the viewholders, but not more generally to others who are not viewholders. We can relate the notion of a shared weltanschauung to that of paradigm. A paradigm is more than shared weltanschauung. It is shared weltanschauung together with the explicitly defined propositions that contribute to understanding. When weltanschauungen are formalised they become paradigms, and are more or less transparent to others who are not viewholders. A formalisation is a language that enables a set of explicit statements (propositions and their corollaries) to be made about the beliefs and other attributes that enable (more or less) everything that must be expressed to be expressed in a self-consistent way. This does not of course mean that the paradigm is able to express ideas for which it has no concepts. Formal propositions define a logic that establishes a framework of thought and conceptualisation that enables organised action to occur, and problem situations to be addressed. Formal logic provides a standard of validity and a means of assessing validity17. While groups may offer behaviour in ways that are consistent with their shared weltanschauung, paradigms emerge when the groups become coherent through formalisation. Informal worldviews are more or less composed of a set of undeclared assumptions and propositions, while formal ones are more or less declared. More, since paradigms are forms of worldviews they are by their very nature bounded, and thus constrain the way in which situations can be described. Now paradigms can change16 so that the nature of the constraint is subject to a degree of change - however bounded it might be. It is easy to provide examples of paradigms. In pedagogic learning environments like universities, it is possible to distinguish between at least three types. These are shown in terms of a set of common attributes in Table 129. 4.

Viable Learning Theory

Viable learning theory derives from viable systems theory as developed by Yolles31, 32, 33, that itself develops from a base of work in cybernetics2 and systems theory25. The viable system is an organisation that survives. In doing so, it can respond to changes (whether or not they have been foreseen) that can generate sufficient variety through self-organisation to deal with that variety impacting on it from its environment (called requisite variety). A measure of variety can also be referred as the complexity of a situation. As such the organisation will be able to support adaptability and change while maintaining stability in its behaviour. In particular, a system is viable if it can maintain stable states of behaviour as it adapts to perturbations from its environment. Such systems are normally considered to be at least semi-autonomous in that they are taken to be analytically and empirically independent. In the learning context, a learner will undertake viable learning if s/he can maintain stable learning behaviour while being able to adapt to changes in a given learning environment that changes the learning situation. We refer to the model that describes such learning capabilities as the viable learning model (VLM), first introduced by Yolles28. If a learning situation is to be thought of as a viable system, then the question must be raised: what is its systemic nature? A system can be seen as a non-separable entity that is composed of objects of attention that are defined in mutual relation to each other, and which is not reducible into a sum of its objects. If each object is thought of as a component of the system, then commonly it is a set of components that interrelate. A system is bounded through a frame of reference. This boundary will change according to the learning purpose and weltanschauung of the learner. A system may also be seen in terms of the degree of interaction between the parts that define it. The parts may be richly or poorly interactive. In modelling a situation systemically, an inquirer will make a judgement about what constitutes a rich set of interactions. It will differentiate this group by creating a boundary around it that distinguishes the set of rich interactions from the poor ones. To see learning systemically, we can define a set of analytically and empirically independent objects of attention that together form a rich learning situation. Other objects of attention that are less rich in terms of their mutual learning interactions are not part of the system, even though they may influence it. Three objects of attention are a: (a) learner; (b) learning source (such as predesigned courseware or a natural learning situation); (c) target situation (that may exist in either a “real” behavioural world or an artificial one).

Thus, within the context of a learning system, consideration of a learner also implies consideration of (b) and (c). The learner is part of a purposeful “system” defined by a structured relationship between the learner, learning source, and the target situation. This effectively drives the learning behaviour of a learner. Clearly, this is a different context from non-purposeful learning, that can occur when learners learn “retrospectively” through the influence of events that become “life” teachers. Each of the objects of attention (a)-(c) can have associated with them a worldview27, an idea that will be explored shortly. 5.

The Cybernetics of Learning

The relationship between the worldviews and the perceived real world is illustrated in a paradigm cycle given in Figure 2 that we shall return to in a moment. If the paradigm cycle provides a satisfactory vehicle to describe learning processes and behaviour, then we should be able to link it with the ideas of learning behaviour, strategy, and metalearning. This is done through a systemic model of Viable Learning30. To explain this we need to adopt the foundational principles in cybernetics, distinguishing between the system and the metasystem - the former being structural and behavioural, and the latter defining the “cognitive” domain that

drives the system. In particular, in any purposeful activity system it is possible to distinguish between cognitive, strategic, and behavioural aspects of a situation. Table 1: Three Paradigms Associated with Educational Institutions Attributes Worldview

Programmed Learning Adopts the tutor’s paradigm.

Action Learning Developing weltanschauung

Knowledge view

Knowledge seen as a valuable commodity.

Learning acquisition

Knowledge can be stored externally and disseminated for learner acquisition in solitary environments.

Assumption about learning

The acquisition and addition of facts, concepts, skills.

Structure of learning

Highly structured learning processes are those which are predetermined or preprogrammed. To transfer facts, skills, and knowledge to those who lack it.

Knowledge is seen in humanistic terms of “knowing as process” to develop meaning and is centred on self. Knowledge is seen as a condition, a knowing, and learning is an elaboration of change of the process of relating meaning to doing and of the enhancement of personal competence. The elaboration and change of meaning-making process, personal competence enhancement. Unstructured learning processes within which no constraints have been imposed in a totally free learning domain. To develop the individual as a whole, especially in the relationship between the self and the environment. Self-selecting and organising.

Learning process

Basis of learner choice Learning mode Course structure

Concerns for relevance

Tightly guided through predefined menu. Individualisation. Based on syllabus, involving the organisation and sequencing of course materials. Perceptual - the knowledge of precepts and propositions of the definitively known; descriptions about the world are normally transmitted from a tutor to a learner as predefined models. Course design criteria involve application and transfer.

Learning purpose Metalearning

Improve study skills. Provided by tutors.

Learning motivation The role of others

Provided by tutors, course materials, and learner purposes. Seen as source for moral support, encouragement, comparison for learning task.

Tutor's role

Subject expert, guardian of knowledge. Responsible for teaching and instructing. May delegate to course media and materials. Measure of proficiency against externally recognised standard. Tutor as subject expert is best to judge quality of work.

Knowledge type

Assessment

Autonomy. Based on process of planning, deciding and experimenting.

Cooperative Learning Developing group paradigm within which individual weltanschauungen develop. Knowledge is seen in humanistic terms of “knowing as process” to develop meaning, centres on self, but works through the group. As in action learning, but through group research and practice. It is also a way of learning about groups and develops group working competencies.

Proves for enhancing academic learning and the development of abilities in cooperative work for later life. Unstructured learning processes within which constraints are imposed by the group to develop the learning domain. To develop the individual as a whole, especially in the relationship between the self and the group that develops through impacts from the environment. Involves exploration through group talking to consolidate individual knowledge Group. Based on process of planning, deciding, and group modification.

Experiential - learners build their own models according to their own perceptions and developed conceptions, and are learning about descriptions of the world. Experience from working life of learner a prime source of learning material. Learn to learn. Self-development enables learning to learn. Provided by stimulation of learning situation. Seen as inherent part of learning process, providing challenge in construction of personal meaning. Facilitator, resource person and co-learner, whose meaning attributed to events is as valid as learner's.

Discovery – learners build their own models according to their developing group perceptions and conceptions, and are learning about descriptions of the world.

Part of learning process. Based on collaborative assessment against mutually agreed criteria.

Essentially a group process, that may be independent of the facilitator.

Group learning process provides a prime learning source. Independent learning. As in action learning, but through group processes. Provided through group dynamics. Essential to the learning process. Without it, the learning process is poorer.

Preferably, invisible facilitator, resource person and co-learner, whose meaning attributed to events is as valid as group’s and learner's.

This lies at the basis of viable systems that, through transformational self-organising processes, are able to support adaptability and change while maintaining stability in their behaviour. In the learning context, an individual will undertake viable learning if s/he can maintain stable learning behaviour while being able to adapt to changes in a given learning environment and different learning situations. We refer to the model that describes such learning capabilities as the viable learning model (VLM).

B e h a vio u r a l d o m a in

O r g a n i s i n g d o m a in B e h a v i o u r a l d o m a in r ep r ese n ta tion

B eh a vio u r a l w o r ld

o r g a n isa tio n o f a ctio n /lea r n in g

P a r a d ig m ( fo r m a l w o r l d v i e w )

d ev elo p m en t/ lea r n in g

fo r m a tio n / C o g n itiv e = > c o n s o l i d a ti o n d o m a i n

O r g a n is in g d o m a in

C o g n it i v e d o m a in

in ter p r eta tio n W elta n sch a u u n g ( in fo r m a l w o r ld v ie w ) r e f l e c ti o n / c r e a t i o n

Figure 2: The Paradigm Cycle, showing the homeomorphic potential of the organising domain

In the paradigm cycle for the VLM, it may be noted that the organising process is a transformation that manifests worldview constructs (seated in the cognitive domain) into the behavioural world (seated in the behavioural domain). This conceptualisation enables us to collapse the worldview domain into a single cognitive domain that defines what may be called a learning metasystem. Whatever outcomes derive from this domain, they are transformed. The nature of the transformation is determined by who makes up the groups that define the cognitive basis of the learning system. This may change according to the perceived world that it is to be delivered to. In this way, the transformation has a morphology, and the transformational process is potentially homeomorphic - having the possibility of any number of behavioural manifestations in a “one to many process”. An example of such a process is translating meaning into any number of languages - meaning is manifested into the different language structures (e.g., Chinese and English) that may have little to do one with the other. This homeomorphic potentiality is also illustrated to the right of Figure 2 through the organising process that takes place during transformation. It is a representation reminiscent of the model developed by Chomsky8, in his structural model of the relationship between syntax and semantics in language theory. It is also structurally similar to that used in the field of artificial intelligence by Clancy and Letsinger9 that distinguishes between deep knowledge and surface knowledge. Here. “deep” knowledge (e.g., principles) can potentially be applied to a variety of application domains. This “variety” is strictly represented as a homeomorphic potential that has a status similar to that of a quantum field31. Thus, it can be argued to be expressible in the terms of quantum theory that Pessa23 has for instance applied to systems and the notion of emergence. It is now possible to assign cognitive properties to each of the domains shown in Figure 2. To do this we develop on work by Habermas13 within the context of the complementary use of systems methodologies. As shown in Table 2, the behavioural domain has cognitive interests12, the organising domain has cognitive purposes, and the cognitive domain cognitive influences31. These cognitive properties exist, and are projected from the cognitive domain.

Table 2: Relationship between human cognitive interests, purpose, and influences COGNITIVE LEARNING INTERESTS OF THE BEHAVIOURAL DOMAIN Technical Practical Critical Deconstraining Work as learning. This enables Learning interaction. This requires that Emancipation. This enables people to (i) people to achieve goals and people as individuals and groups in a liberate themselves from the learning generate material consesocial system learn to gain and develop constraints imposed by power structures quences. It involves technical the possibilities of an understanding of (ii) learn through precipitation in social ability to undertake learning each others subjective views. It is conand political processes to control their action in the environment, and sistent with a practical interest in mutual own destinies. the ability to make prediction understanding that can address disand establish control. agreements, which can be a threat to the social form of life COGNITIVE PURPOSES OF LEARNING WITHIN THE ORGANISING DOMAIN Cybernetical Rational Ideological Learning intention. This is Logico-relational. Enables learning Manner of thinking about learning. An through the creation and strasource missions, goals, and aims to be intellectual framework through which tegic pursuit of learning goals defined, and approached through planpolicy makers observe and interpret realand aims that may change over ning. It involves logical, relational, and ity that has a politically correct ethical time, enables people through rational abilities to organise thought and and moral orientation, provides an image learning control and commulearning action and thus to define sets of of the future that enables action through nications processes to redirect possible systemic and learning behavpolitically correct strategic policy, and their futures. iour possibilities. gives a politically correct view of stages of historical development in respect of interaction with the external environment and related learning processes. COGNITIVE INFLUENCES ON LEARNING WITHIN THE COGNITIVE DOMAIN Social Cultural Political Formation. Enables individuThinking. Influences occur from knowl- Freedom. Influences occur from knowlals/groups to be influenced by edges about learning that derive from the edges about learning that affect our polknowledges about learning cognitive organisation (beliefs, attitudes, ity determined in part, by how we think that relate to our social envivalues) other worldviews. It ultimately about the constraints on group and indironment. This has a consedetermines how we learn to interact and vidual freedoms to learn and in connecquence for our learning about defines our logico-relational understand- tion with this to organise and behave. It social structures and processes ings. ultimately has impact on our learning that define our social forms related ideology and our degree of emanthat are related to our intencipation. tions and behaviours.

The Habermasian classification of cognitive interest has technical and practical attributes. The former of these relates to work. In our context this can also be directly connected to Kolb’s behavioural phases of the learning process, so that, the notion of work refers to the work of learning. An interpretation of the practical attribute is concerned with learners interaction within a learning process that can also be related to Kolb’s ideas. From Habermas’ perspective, this should ideally be accompanied by emancipation in the way learning occurs to enable learners the opportunity to maximise their learning potential. Learning also occurs according to cognitive purpose, and the cybernetic attributes of this relate to intention by the strategic pursuit of learning goals and control processes. Rational and ideological attributes also affect cognitive learning purposes. Finally, cognitive influences can occur with respect to social, cultural or political attributes that guide learners to learn according to their embedded knowledges about learning (i.e. metalearning). The social attribute is connected to structures, processes (including facilitating technologies), and roles, and learning will have attributes that relate to this. Similarly one is able to identify cultural and political attributes, the meaning of these being relative to the nature of the learning situation that one is concerned with. Thus for instance, in the political nature of learning, polity is seen as being concerned with engineering the enablement of group form, condition of order, and related processes. It

relates to the learning situation that a learner wishes to be in. Its commodity is power that determines who directs the learning process, and how it is directed. It also relates, for instance, to Table 1. It is clear that learning processes can occur at three levels: cognitive, strategic, and behavioural. The cognitive domain is knowledge related, some of this being knowledge about the learning process itself (metalearning). Part of this process also concerns planning as considered by Lester and Kroll19. It is through metalearning that an ability to develop learning strategy is developed that itself determines the nature of learning behaviour and associated style. Metalearning operates from a learner paradigm which has within it a learning system as its metaphorical “cognitive consciousness” that provides the system with direction and guidance. Learning behaviour can therefore be seen as a manifestation of this. Thus, for instance, a learner (who may be an individual, a group of individuals, or an organisation), enters a learning situation. Learning strategy is determined from the learner’s metasystem, but is an organising process that ultimately determines behaviour. It is a result of cognitive purpose that Kroll15 refers to as metacognitition. It involves cybernetic and rational process that operate as transformations that may be subject to surprises, and that will constrain learning behaviour. As an example of this, learning strategy may break down, when the learning process become destabilised, and learning becomes an ad hoc affair. This can occur in situations of learning “chaos”. We are aware that according to VLM, metalearning, learning strategy, and learning behaviour are all directly linked, and any change in one domain can directly affect the others. Thus for instance, suppose that a perturbation caused by the environment occurs in the domain of a learner’s learning strategy. Then under the appropriate conditions we should see an effect in either learning behaviour and/or metalearning. In the event that these perturbations are very severe, then chaos can be induced and the learning system becomes divorced from its metasystem. This may appear to occur as an arbitrary learning behaviour from the perspective of an observer. However, according to propositions by Yolles31 that conform to chaos theory, a new metasystem may spontaneously arise that provides a metalearning basis for the new learning behaviour. 6.

Learning is a Viable Systems Framework

Earlier we considered three objects of attention (a)-(c). For the VLM this suggests that there are also a triad of worldviews that in some way complement one another. Worldview complementarism has been criticised by those that we refer to as the “fundamentalists” of paradigm incommensurability31 that suggests that worldviews cannot be made complementary since they are conceptually or qualitatively different. Approaches to counter this have occurred using table 2. The nature of complementarism is of interest in this section, arguing that from it a worldview results that provides a learning model that the learner pursues. We can consider that each of these objects of attention is itself a viable system in its own right. In the case of (a), this means that the learner is able to be purposeful and adaptable in his or her learning behaviour, and is able to survive as a learner. In the case of (b) and (c), a learning source and target situation seen to form the basis of a viable system will have associated with them groups of people who themselves form a viable learning organisation. As a consequence, each of the three objects of attention can be thought of as having cognitive, organising, and behavioural domains. Concentrating on the former of these for a moment, we can represent the embedded worldviews associated with each as in Figure 3. The intention of this diagram is to illustrate that the three objects of attention can be kept analytically apart the relationship between the target situation, the inquirer, and the targeting methodology clearly indicate that

the processes of learning, and learning sources influences, are distinct. The diagram is also intended to highlight the idea that in learning situations:   

the worldviews of the actors in a situation are seen as an abstract property learning involves a learner with a worldview (or worldviews for a plural inquirer) there is an interaction and influence between the formalised worldview(s) of a learning source, the worldviews of a learner, and the target situation through their cultures and “truths”.

In ter a ctiv e sp a c e o f w o r ld v ie w s d e fin in g th e m e ta syste m o f lea r n in g P a ra d ig m (s) o f le a r n in g so u r c e

P a r a d ig m (s)/w o r ld v iew (s) o f ta r g e t situ a tio n

W o r ld v iew (s) o f lea r n e r

Figure 3: Learning triad defining an interactive space of worldviews and their knowledges The context of the situation can be defined within a framework of thought that constructs these three objects of attention as its orthogonalities against which changes in one can be related to the other. This visualisation is not unlike traditional approaches where independent but related variables (e.g. space and time) are graphed one against another within a frame of reference to enable changes (e.g., velocity) to be represented. Part of our interest here is to argue that learning can be represented as a viable system, and we do this through a direct translation from systems theory to the learning domain. Thus, the explanations that we propose will be subject to experimental verification, yet to come. The precepts that define a viable learning system will involve the argument that a metasystem can be formed that is manifested as a purposeful adaptive learning system, that can evolve according to the principles of viable systems theory as proposed by Schwarz25. The formation of a cognitive domain as represented in Figure 4 requires the creation of shared paradigms through the establishment of a common cognitive model. This enables at least some of the knowledge of a learner to be used to apply at least some of the knowledge of a learning source to at least some of the knowledge of the worldviews that make up a target situation. The selection of knowledge comes from the frame of reference that enables a learner to learn within a given context. If we think in terms of systems having learning structure and related behaviour, then the cognitive domain can be thought of as forming a metasystem that directs learning into complex target situations. We can refer to this as metalearning. When the learning metasystem couples with a system we refer to it as a purposive in its learning behaviour. If the three apexes of worldview do not relate to each other in a common model, then the triad of Figure 3 is not seen to form the basis of a systemic learning process. As a result the learning process will be seen to be composed of an arbitrary selection and application of learning to the target situation. Some might refer to it as illustrating a misunderstanding of a learning material, of a target situation that manifests as behaviour what has been learned, of source learning, or of the application of

the latter to the former. Any learning behaviour that results from a strategic learning process will likely be meaningless and of little value. When a learning process falls into chaos, it loses its connection with the learning metasystem, and in response to stimulus from a target situation the learning process behaves spontaneously in a way that is structure determined. To understand this, we can say that if a learner’s structural approach is defined as an ordered set of procedural steps of learning, then only these steps are available for selection in that order by a learner. If learning behaviour is stable then its scheduling process of these steps is controlled. Contrary to this, under chaos the scheduling process will be arbitrary. Let us suppose that the learning system passes from a condition of stability to one of chaos, and then back to one of stability. As stability is regained, a new metasystem arises so that the system has passed through a metamorphosis. To understand the nature of the new metasystem, we must explore the learning process a little further. In the formation of a viable learning system, we have proposed three objects of attention to be required. Each has its worldview, but together they do not define a whole metasystem. What functions as a metasystem is a disconnected and disjointed set of worldviews that simply contributes to the confusion of chaos. It is only when the objects of attention come together by forming a whole metasystem (that defines purpose), that the system will be able to achieve and maintain that purpose. The emerging metasystem occurs through the formation of a virtual paradigm that may endure for the duration of a learning period into the situation. Three functions of the metasystem are that it will: 1. Define a shared worldview that enables a structured learning approach to be applied to a situation by a learner meaningfully, 2. Constrain the learner by use of a structured learning approach suitable for complex learning situations 3. Establish the basis of control on the selection or use of structured learning to make it appropriate for the target situation, having care in how learning is applied to the situation.

The first of these results come through the creation of a virtual paradigm that formalises a learner’s approach. Part of this process is to define the purposes of a learning. This comes from the learner’s understanding of the actors’ view(s) of the target situation, the mission and goals of the structured learning, and his own purposes in applying the learned knowledge and skills to the target situation. Within the learning process, a learner will adopt a set of propositions from the different worldviews that represents its “truths” and enables knowledges to be recognised. In structured learning situations this is guided by the selection of a learning approach such as illustrated in Table 1, or a circumstantial one that comes through action learning. In the former case the learning content is directed by the institution, while in action learning it is directed by the circumstances of the target situation. Sometimes it can also be a result of both of these as in cooperative learning, where we see the target situation as being the cooperative group. The idea that there are three autonomous worldviews in interaction is complexified, because the worldviews attached to each autonomous apex of Figure 4 are themselves likely to be plural. This means that:   

there may be many worldviews in a target situation that must be addressed; the learner may be a group that involves a number of worldviews; the paradigm that defines learning content may also be pluralistic.

The nature of cooperative learning is of particular interest since it consists of a plurality of learner worldviews. The group assembles, and the individuals more or less together form a set of common cognitive purposes within a single frame of reference defined within a virtual (or working) paradigm that enables them to work together (more or less) as a team. Group behaviour is possible because of the formation of a shared weltanschauungen. This occurs through a common cognitive model that enables meaning to be shared. Its boundaries are defined as a frame of reference for group behaviour. The individual weltanschauungen are maintained, though through association there may well be a learning process in which weltanschauungen are changed in some way. If failure of a group process occurs, then one explanation is through weltanschauung incommensurability. In a situation worldview plurality can occur with respect to both the informal and formal worldviews of an organisation. There is always an interaction between weltanschauungen and paradigms in the same way as there is between different weltanschauungen and different paradigms. The plurality of informal worldviews is often ignored by supposing that there is a consensus in a situation. Often, little is done to determine what the consensus actually means in a given context, and whether it has any value in respect of a learning strategy. With respect to formal worldviews, it is normally the dominant paradigm that is referred to by a novice learner during learning. This can also be seen as a supraparadigm of the organisation. 7.

Conclusion

In this paper we have attempted to show that learning situations can be seen as viable learning systems that are subject to chaos. As a result, it is possible to see learners within a given definable context having properties of viable systems, including purposefulness, adaptability, and learning survivability. It also provides the opportunity of discussing such problems as complexity and evolutionary learning within this context. More, it offers the possibility of encompassing and developing existing learning theory, subject to experimental evidence, that the propositions of viable system do indeed operate. The potential of such theory is to enable a learning auditing system to be developed that can assess whether learning may become subject to chaotic interference that differentiates learning intention from learning strategy and behaviour. This would operate by establishing tests that would evaluate learners at the cognitive, strategic, and behavioural levels of learning, and evaluate any changes that might have occurred. For this to work, we must be sure about the nature of the relationships between these three domains. 8. References 1. Argyris, C., 1990, Overcoming Organisational Defences. Allyn and Bacon 2. Beer, S., 1959, Cybernetics and Management. English Universities Press; Beer, S., 1975, Platform for Change. Wiley; Beer, S., 1979. The Heart of Enterprise. Wiley; Beer, S., 1981, Brain of the Firm. 2nd ed. Wiley, New York; Beer,S., 1985. Diagnosing the System for Organisations. Wiley 3. Berger, P., Luckman, T., 1966. The Social Construction of Reality. Penguin 4. Checkland, P.B., 1981, Systems Thinking, Systems Practice, Wiley, Chichester 5. Checkland, P.B., Davies, L., 1986, The Use of the Term Weltanschauung in Soft Systems Methodology. J. Applied Systems Analysis, vol.13, pp109-115. 6. Churchman, C.W. 1979, The Systems Approach, 2nd ed. Dell, New York 7. Chi, M.T.H., 1992, Conceptual change within and across ontological catagories: Examples from learning and discoverery in science. In Giere, R., (Ed.), Cognitive Models of Science: Minisota studies in the philosophy of science (pp.129-186). University of Minnesota Press, Minneapolis, MN. 8. Chomsky, N., 1975, Problems of Knowledge and Freedom. Pantheon, New York

9. Clancey, W.J., Letsinger, R., 1981, Neomycin: Reconfiguring a Rule Based Expert System for Application to Teaching. IJCAI 7(2). 10. Cohen, J., Stewart, I., 1994, The Collapse of Chaos: discovering simplicity in a complex world. Viking, London 11. Cunningham, I., (1990), Openness and Learning to Learn. Contained in Computers and Learning, ed. Boyd-Barret, O., Scanlon, E. Addison Wesley 12. Habermas, J., 1970, Knowledge and interest in: Sociological Theory and Philosophical Analysis, pp36-54, (Emmet, D., MacIntyre, A., eds), MacMillan, London 13. Jackson, M.C., 1992, Systems Methodologies for the Management Sciences. Plenum, New York 14. Kolb, D.A., et al, 1974. Organisational Psychology: An Experiential Approach. Englewood Cliffs, New Jersey. Prentice-Hall. 15. Kroll, D.L., 1988, Cooperative mathematical proble-solving and metacognition: A case study of three pairs of women. Dissertation Absracts International, 49, 2958A (University Microfilms No. 8902580) 16. Kuhn, S.T., 1970, The Structure of Scientific Revolutions. University of Chicago Press, Chicago 17. Kyberg Jr., H.E. (1968). Philosophy of Science: a Formal Approach. Collier-Macmillan, London. p.20 18. Laurillard, D., 1987, Computers and the Emancipation of Students. Contained in Computers and Learning, ed. Boyd-Barret, O., Scanlon, E. Addison Wesley. 19. Lester, F.K., Kroll, D.L., 1990, Assesing student growth in mathematical problem solving. In Kulm, G. (Ed.), Assessing Higher Order Thinking in Mathematics. American Association for the Advancement of Science, Washington DC. 20. McClelland, R.J., Yolles, M.I., 8-9 April 1997,The Cybernetics of Learning Environments. CTI-AFM 8th annual conference, Bristol. 21. Marton, F., 1983. Beyond individual differences. Educational Psychology. 3, 289-303. 22. Marton, F., Säljö, R., 1976. On quantitative differences in learning: outcome and process. British Journal of Educational Psychology, 46, 4-11. 23. Pessa, E., 1998, Emergence, Self-Organisation, and Quantun Theory. In Minati, G., First Italian Conference on Systemics: Systems Science to design a sustainable development. Apogeo srl, Viale Papiniano 38-20123 Milano, Italy. www.apogeonline.com. ISBN 887303-432-2. 24. Scheler, M., 1947, Die Stellung des Menschen im Kosmos.: Nymphenburger Verlagshandlung, Munich 25. Schwarz, E., 1994 (April), A Metamodel to Interpret the Emergence, Evolution and Functioning of Viable Natural Systems. Presented at the European Meeting on Cybernetics and Systems Research, Vienna, and in Trappl, R., (ed.), 1994, Cybernetics and Systems ‘94, World Scientific, Singapore, pp1579-1586; Schwarz, E., 1995, Where is the Paradigm? In the Peoples Mind or in the Social System? Seperata de Revista International de Systemas, vol. 7, No. 1-3, pp5-54. 26. Strang, A., 1987. The hidden barriers. Contained in Beyond Distance Teaching - Towards Open Learning, ed. Hodgson, V.E et al, Open University Press. 27. Yolles, M.I., 1996 (Oct), Critical Systems Thinking, Paradigms, and the Modelling Space. System Practice, 9(3). 28. Yolles, M.I., 1997a, An Introduction to the Theory of Viable Learning, In Pintellas, P., Potari, D., (Eds.), Didactics of Mathematics and Informatatics in Education. Pneumatikos Publishing Co., Athens, Hellas.

29. Yolles, M.I., 1997b, Learning Style and Strategy, and the Theory of Viable Learning. In Chapman, G.M., (Ed.), Computer Based Learning in Science, CBLIS’97. (ISBN 80-7040217-2). 30. Yolles, M.I., 1997c (Aug.). From Viable Systems to Viable Inquiry Systems. Systemist, 19(3)154,173 31. Yolles, M.I. 1999, Management Systems:A Viable Approach. Financial Times Pitman, London. 32. Yolles, M.I., 1999a (July), Towards a Viable Systems Theory of Joint Ventures, Systemist. 33. Yolles, M.I., McClelland, 8-9 April 1997, Developing Measures for Learning Strategy. CTI-AFM 8th annual conference on Educational Innovation in Economics and Business Administration, Bristol University. 34. Yolles, M.I, Pirani, M., 1991, Computer Aided Learning and the Pedagogic Paradigm. Contained within: COMETT/VOLET Cb: Final report strand Cb-Contract year 1990/1991. Reference number 90/2/5081/Cb, Brussels; Yolles, M.I., Pirani, M. 1994, Towards Expanded Open Learning for Special Needs. J.Computer Information Technology. 2(4)251-263. Maurice Yolles, Liverpool Business School, John Moores University, 98 Mount Pleasant, UK Liverpool L3 5UZ. [email protected]

Short title: Viable Learning Systems