March 21, 2018

Thinking about theory and practice in online learning

Taking the float plane to Victoria: always a wonderful experience

Taking the float plane to Victoria: always a wonderful experience

I ran a short face-to-face workshop yesterday on ‘Thinking about Theory and Practice’ for about a dozen students taking the Masters of Arts in Learning and Technology at Royal Roads University  My online open textbook, Teaching in a Digital Age, is being used in this program and the instructors asked me to run a workshop on this topic, as students struggle with the relationship between epistemology, theories of learning, and methods of teaching.

The exercise

I’m not surprised that students struggle with this, as the relationships are by no means clear. I started by asking them to define different epistemologies. I then asked them what the connection was between different epistemologies and different learning theories. Then I asked them to choose from about 18 different methods or approaches to teaching (all covered in my book) and try to place them in relationship to theories of learning, as in Figure 1 below.

Figure 1: Thinking about theory and practice

Figure 1: Thinking about theory and practice

I also raised questions about whether constructivism and connectivism are epistemologies, or theories of learning, or both.

This was meant as a heuristic exercise, to get students arguing about and discussing the relationship between epistemology, theory, and practice, and why it is important to think about this in terms of learning design.

I ended my session with the following questions:

  • Constructivism and connectivism: are they epistemologies or learning theories?
  • Is there a direct relationship between epistemology, theory and practice?
  • How well do different teaching methods ‘fit’ with a specific learning theory?
  • Does technology change the nature of knowledge? If so, is connectivism an ‘adequate’ epistemology for a digital age?

Following my workshop, in the afternoon the students were divided into two teams to formally debate the motion (chosen by the instructors):

Connectivism should be adopted as the learning theory for educating students in our digital culture.

Both the workshop and the debate resulted in very thoughtful and forceful, sometimes impassioned, discussion.


It is impossible to capture the richness of the discussions in a short blog (I am hoping that the MALAT team will make an edited recording of the sessions available online). Different participants will have come away from the two sessions with different conclusions. Although I am fairly confident about discussing theories of learning and methods of teaching, I am not a trained or qualified philosopher, so I hesitate to tell students what the truth is in this area (OK, so I’m a relative constructivist).

However, here are some of my conclusions:

  • the most important is that I believe that connectivism is more of an epistemology than a theory of learning. Indeed it is an epistemology that relies on other theories of learning to explain how learning occurs in networks, although it has established conditions that make for ‘effective’ networks (see, for instance, Downes, 2007). In this sense it can be seen as an overall belief system about the importance of networks for sustaining and creating knowledge, but the mechanisms by which learning occurs in networks still need to be identified or worked out, or explained in terms of existing theories, such as constructivism.  This does not mean that over time, particular ways of learning and creating new knowledge through networking will not be identified, but more importantly, it would seem to make sense that we should be making use of networks and social media in education, since we are all becoming increasingly immersed in a connectivist world, and learning how to adapt and thrive in such a world probably requires using connections and networks for teaching and learning;
  • similarly, I am uncomfortable with defining constructivism as an epistemology. It is a strong theory in terms of explaining how learning occurs, but it takes its philosophical roots from other more general epistemologies. I would need to be a philosopher to define accurately what these would be, but constructivism is strongly influenced by philosophers such as John Stuart Mill (free will), Jean Jacques Rousseau (the Natural Human), and Jean Piaget (‘genetic’ epistemology);
  • although there is some relationship between epistemologies and theories of learning, they are not isomorphic, in the sense that a single theory of learning derives solely from one epistemological position. For instance, cognitive theories of learning draw heavily on both objectivist approaches (e.g. brain research) and more subjective or reflective approaches, such as constructivism;
  • there is even less isomorphism between theories of learning and methods of teaching, because methods of teaching are driven primarily by context. For instance, in a digital age, trades apprentices increasingly need both manual and cognitive learning. The learning of manual or mechanical skills through an apprenticeship model may be behaviourist in approach, but cognitive apprenticeship may draw much more heavily on a constructivist approach. Nevertheless some teaching methods, such as lectures or xMOOCs, are generally more towards the objectivist spectrum, while cMOOCs are more towards the connectivist spectrum (even though in practice they may include other approaches, such as more objectivist webinars, and support from teachers or experts through constructivist forms of discussion);
  • different subject areas tend to favour different epistemological positions, such as science favouring more objectivist approaches to teaching, and arts more subjective and interpretive approaches. However, it is still possible to teach science in a constructivist way – for instance through problem or inquiry-based learning – and arts in a more objectivist way (for instance, Mrs. Thatcher wanted British school children to learn the facts about British history, rather than discuss imperialism or racism and their legacies), although purists will argue that students will not become ‘true’ scientists or historians if the teaching does not reflect the ‘core’ epistemological nature of the subject area.

However, I’m a ‘relativist’ on all these points and open to be persuaded.

Does it matter?

Isn’t this all terribly abstract and philosophical? Nothing seems clear and definite, so how does thinking about these things help to teach better?

Well, if you are going to be an instructional designer, you will come across instructors and subject experts who may have a fundamentally different epistemological position from you. It will really help if you understand their position and how to take this into account when designing courses.

Second, there is nothing more practical than a good theory. If you have a theory that is convincing to you in terms of explaining how learners best learn, this should drive your teaching practice. It may not tell you exactly what to do as a teacher, but it should enable you to work out for yourself what to do – and more importantly, what learners need to do. But this theory needs to fit with your overall epistemological position about the nature of knowledge in your subject area.

Third, teaching is a pragmatic profession. It may take several different approaches, depending on the context and above all on the learner. In some contexts, such as safety compliance, employers don’t want workers questioning the process; they need to learn exactly what to do in a particular circumstance (behaviourism rules). In others, where problem-solving is essential, rote learning is not going to help dealing with a new or unanticipated danger.  Having a range of options in terms of teaching approaches for a range of different kinds of learners and contexts is more likely to produce results than slavishly following one particular method.

Lastly, all this uncertainty and choice illustrates why teaching and learning are not well defined activities that can be easily mechanised. Humans are better than machines at dealing with uncertainty and fuzzy or ambiguous circumstances, but only if they have a deep understanding of the options available to them and the circumstances in which each option is likely to succeed. This means thinking carefully about epistemology and theories of learning as well as various methods of teaching.

Galiano Island, on the way to Victoria

Galiano Island, on the way to Victoria. Vancouver Island is in the background.

The nurturing and social reform models of teaching and their relevance to connectivist online learning

Counselling face-to-face 2 A counsellor and student at Empire State College, New York, which has a mentoring approach to adult education

These are the last of the teaching models discussed in Chapter 4 of my open textbook, ‘Teaching in a Digital Age.’ (I thought I had covered all five of Pratt’s perspectives on teaching in my last blog post, but realised I had accidentally left out these two important approaches to teaching. I have also added a scenario to introduce the Chapter.)

What both these models, identified by Pratt (1998), have in common is a focus on the individual rather than on the teacher, the institution, or state. They are both in a sense attempts at liberating learners from the restrictions of formal and institutional types of education.

The nurturing approach

A nurturing approach to teaching can best be understood in terms of the role of a parent. Pratt (1998) states:

We expect ‘successful’ parents to understand and empathize with their child; and that they will provide kind, compassionate, and loving guidance through content areas of utmost difficulty….The nurturing educator works with other issues…in different contexts and different age groups, but the underlying attributes and concerns remain the same. Learners’ efficacy and self-esteem issues become the ultimate criteria against which learning success is measured, rather than performance-related mastery of a content body.

There is a strong emphasis on the teacher focusing on the interests of the learner, on empathizing with how the learner approaches learning, of listening carefully to what the learner is saying and thinking when learning, and providing appropriate, supportive responses in the form of ‘consensual validation of experience‘. This theory is driven partly by the observation that people learn autonomously from a very early age, so the trick is to create an environment for the learner that encourages rather than inhibits their ‘natural’ tendency to learn, and directs it into appropriate learning tasks, decided by an analysis of the learner’s needs.

The social reform model

Pratt (1998, p. 173) states:

Teachers holding a social reform perspective are most interested in creating a better society and view their teaching as contributing to that end. Their perspective is unique in that it is based upon an explicitly stated ideal or set of principles linked to a vision of a better social order. Social reformers do not teach in one single way, nor do they hold distinctive views about knowledge in general…these factors all depend on the particular ideal that inspires their actions.’

This then in some ways is less a theory of teaching and more an epistemological position, that society needs change, and the social reformer knows how to bring about this change.

History, and relevance for connectivism

These approaches to teaching again have a long history, with echoes of

  • Jean-Jacques Rousseau (1762) (‘education should be carried out, so far as possible, in harmony with the development of the child’s natural capacities by a process of apparently autonomous discovery‘ (Stanford Encyclopedia of Philosophy)
  • Malcolm Knowles (1984) ‘As a person matures his self concept moves from one of being a dependent personality toward one of being a self-directed human being.’
  • Paulo Freire (2004) (‘education makes sense because women and men learn that through learning they can make and remake themselves, because women and men are able to take responsibility for themselves as beings capable of knowing—of knowing that they know and knowing that they don’t.’)
  • Ivan Illich (1971) in his criticism of the institutionalization of education (‘The current search for new educational funnels must be reversed into the search for their institutional inverse: educational webs which heighten the opportunity for each one to transform each moment of his living into one of learning, sharing, and caring.’

The reason why the nurturing and social reform approaches to teaching are important is because they reflect many of the assumptions or beliefs around connectivism. Indeed, as early as 1971, Illich made this remarkable statement for the use of advanced technology to support “learning webs.”

The operation of a peer-matching network would be simple. The user would identify himself by name and address and describe the activity for which he sought a peer. A computer would send him back the names and addresses of all those who had inserted the same description. It is amazing that such a simple utility has never been used on a broad scale for publicly valued activity.’

Well, those conditions certainly exist today. Learners do not necessarily need to go through institutional gateways to access information or knowledge, which is increasing available and accessible through the Internet. MOOCs help to identify those common interests and connectivist MOOCs in particular aim to provide the networks of common interests and the environment for self-directed learning. The digital age provides the technology infrastructure and support needed for this kind of learning.

The roles of learners and teachers

Of all the models of teaching these two are the most learner centred. They are based on an overwhelmingly optimistic view of human nature, that people will seek out and learn what they need, and will find the necessary support from caring, dedicated educators and from others with similar interests and concerns, and that individuals have the capacity and ability to identify and follow through with their own educational needs. It is also a more radical view of education, because it seeks to escape the political and controlling aspects of state or private education.

Within each of these two models, there are differences of view about the centrality of teachers for successful learning. For Pratt, the teacher plays a central role in nurturing learning; for others such as Illich or Freire, professionally trained teachers are more likely to be the servant of the state than of the individual learner. Volunteer mentors or social groups organised around certain ideals or social goals provide alternative forms of support for learners.

Strengths and weaknesses of these two approaches

There are, as always, a number of drawbacks to these two approaches to teaching:

  • The teacher in a nurturing approach needs to adopt a highly dedicated and unselfish approach, putting the demands and needs of the learner first. This often means for teachers who are experts in their subject holding back the transmission and sharing of their knowledge until the learner is ‘ready’, thus denying to many subject experts their own identity and needs to a large extent;
  • Pratt argues that ‘although content is apparently neglected, children taught by nurturing educators do continue to master it at much the same rate as children taught by curriculum-driven teaching methodologies‘, but no empirical evidence is offered to support this statement, although it does derive in Pratt’s case from strong personal experience of teaching in this way;
  • like all the other teaching approaches the nurturing method is driven by a very strong belief system, which will not necessarily be shared by other educators (or parents or even students, for that matter);
  • a nurturing approach is probably the most labour-intensive of all the teaching models, requiring a deep understanding on the part of the teacher of each learner and that learner’s needs; every individual learner is different and needs to be treated differently, and teachers need to spend a great deal of time identifying learners’ needs , their readiness to learn, and building or creating supportive environments or contexts for that learning;
  • there is likely to be a conflict between what the learner identifies as their personal learning needs, and the demands of society in a digital age. Dedicated teachers may be able to help a learner negotiate that divide, but in situations where learners are left without professional guidance learners may end up just talking to other individuals with similar views that do not progress their learning (remembering that academic teaching is a rhetorical exercise, changing the way learners view the world.)
  • social reform depends to a large extent on learners and teachers embracing similar belief systems, and can easily descend into dogmatism without challenges from outside the ‘in-community’ established by self-referential groups.

Nevertheless, there are aspects of both models that have significance for a digital age:

  • both nurturing and social reform approaches seems to work well for many adults in particular, and the nurturing approach also works well for younger children.
  • nurturing is an approach that has been adopted as much in advanced corporate training in companies such as Google as in informal adult education.
  • connectivist MOOCs strongly reflect both the nurturing approach and the ability to create webs of connections that enables the development of self-efficacy and attempts at social reform
  • both methods seem to work well when learners are already fairly well educated and already have good prior knowledge and conceptual development.
  • such approaches that focus on the needs of individuals rather than institutions or state bureaucracies can liberate thinking and learning and thus make the difference between ‘good’ and ‘excellent’ in creative thinking, problem-solving, and application of knowledge in complex and variable contexts.

Over to you

Once again, your feedback on this analysis of the two teaching models will be invaluable. In particular:

  • Do you have experience of teaching in one or both of these ways? If so, do you agree with the analysis of the strengths and weaknesses of each component?
  • Do you think that connectivism is a modern reflection of either of these models of teaching – or is connectivism a distinct and unique method of teaching in itself? If so, what distinguishes it as a teaching method from all the other methods I have covered?

It’s always great to hear from readers.

Up next

Although I posted some key takeaways at the end of Chapter 4, I will be doing an extended reflection on all the models described with a particular focus on their suitability for developing the knowledge and skills needed in a digital age.


Freire, P. (2004). Pedagogy of Indignation. Boulder: CO, Paradigm

Illich, I. (1971) Deschooling Society, (accessed 6 August, 2014)

Knowles, M. (1984) Andragogy in Action. Applying modern principles of adult education, San Francisco: Jossey Bass.

Pratt, D. (1998) Five Perspectives on Teaching in Adult and Higher Education Malabar FL: Krieger Publishing Company

Rousseau, J.-J. (1762) Émile, ou de l’Éducation  (Trans. Allan Bloom. New York: Basic Books, 1979)

Learning theories and online learning

Figure 3.3. Adults learning in groups in a constructivist manner - and assisted by technology

Figure 3.3. Adults learning in groups in a constructivist manner – and assisted by technology


Chapter 3 of my open textbook on ‘Teaching in a Digital Age‘ is about theory and practice in teaching for a digital age, which I am still in the process of writing. I have to admit that I approached writing about learning theories with some dread. In particular I was concerned (in order of dread) that:

  • this will appear incredibly boring/lack originality, because it has been done so many times before by other, more qualified authors (but then those that already know this stuff can easily skip it)
  • I’m not sure that theories of learning actually drive teaching (although surely an understanding of how students learn should do so)
  • I would have to deal with connectivism somehow, and I am certainly not an expert on that topic – but maybe that might be an advantage in bringing it to the attention of people who have previously shown no interest in it, and how it differs from previous theories
  • it could be argued that past learning theories are made irrelevant by digital technologies (and I certainly don’t agree with that point of view.)

In the end, I can’t see how a discussion of learning theories can be avoided. Unless readers of the book have this basic understanding of the different views of learning, they will not be in a good position to make choices, especially regarding the use of technology for teaching and learning. In particular, I see a danger of becoming dogmatic and blinkered by unchallenged assumptions about the nature of learning that results from not exploring alternative theories. But lastly, as Kurt Lewin said, there is nothing more practical than a good theory. A good theory helps us make informed decisions in areas of uncertainty. So, I am sharing here my first draft with you. Please note this is just part of the whole chapter, which also includes the following:

  • Teaching and learning styles
  • Deep vs surface learning.
  • Learner-centered teaching, learner engagement.
  • What we know about skills development
  • Competency based learning.
  • Learning design models
  • learner characteristics: digital natives and digital literacy
  • are we right to fear the use of computers for teaching?
  • Summary of research on teaching.

Also, Chapter 2 discusses the nature of knowledge, and in particular different epistemologies that underpin different theories of learning. However, theories of learning are more than enough to chew on for the moment.

Theories of learning

…there is an impressive body of evidence on how teaching methods and curriculum design affect deep, autonomous, and reflective learning. Yet most faculty are largely ignorant of this scholarship, and instructional practices and curriculum planning are dominated by tradition rather than research evidence. As a result, teaching remains largely didactic, assessment of student work is often trivial, and curricula are more likely to emphasize content coverage than acquisition of lifelong and life-wide learning skills.”

Knapper, 2010, p. 229

“There is nothing so practical as a good theory.” Kurt Lewin, 1951, p. 169

Why an understanding of theories of learning is important

Most teachers in the k-12 sector will be familiar with the main theories of learning, but because instructors in post-secondary education are hired primarily for their subject experience, or research or vocational skills, it is essential to introduce and discuss, if only briefly, these main theories. In practice, even without formal training or knowledge of different theories of learning, all teachers and instructors will approach teaching within one of these main theoretical approaches, whether or not they are aware of the educational jargon surrounding these approaches. Also, as online learning, technology-based teaching, and informal digital networks of learners have evolved, new theories of learning are emerging.

With a knowledge of alternative theoretical approaches, teachers and instructors are in a better position to make choices about how to approach their teaching in ways that will best fit the perceived needs of their students, within the very many different learning contexts that teachers and instructors face. This is particularly important when addressing many of the requirements of learners in a digital age. Furthermore, the choice of or preference for one particular theoretical approach will have major implications for the way that technology is used to support teaching.

In fact, there is a huge amount of literature on theories of learning, and I am aware that the treatment here is cursory, to say the least. Those who would prefer a more detailed introduction to theories of learning could, for an obscene price, purchase Schunk (2011), or for a more reasonable price Harasim (2012). The aim of my book though is not to be comprehensive in terms of in-depth coverage of all learning theories, but to provide a basis on which to suggest and evaluate different ways of teaching to meet the diverse needs of learners in a digital age.


Although initially developed in the 1920s, behaviourism still dominates approaches to teaching and learning in many places, particularly in the USA.

Behaviourist psychology is an attempt to model the study of human behaviour on the methods of the physical sciences, and therefore concentrates attention on those aspects of behaviour that are capable of direct observation and measurement. At the heart of behaviourism is the idea that certain behavioural responses become associated in a mechanistic and invariant way with specific stimuli. Thus a certain stimulus will evoke a particular response. At its simplest, it may be a purely physiological reflex action, like the contraction of an iris in the eye when stimulated by bright light.

However, most human behaviour is more complex. Nevertheless behaviourists have demonstrated in labs that it is possible to reinforce through reward or punishment the association between any particular stimulus or event and a particular behavioural response. The bond formed between a stimulus and response will depend on the existence of an appropriate means of reinforcement at the time of association between stimulus and response.  This depends on random behaviour (trial and error) being appropriately reinforced as it occurs.

This is essentially the concept of operant conditioning, a principle most clearly developed by Skinner (1968). He showed that pigeons could be trained in quite complex behaviour by rewarding particular, desired responses that might initially occur at random, with appropriate stimuli, such as the provision of food pellets. He also found that a chain of responses could be developed, without the need for intervening stimuli to be present, thus linking an initially remote stimulus with a more complex behaviour. Furthermore, inappropriate or previously learned behaviour could be extinguished by withdrawing reinforcement. Reinforcement in humans can be quite simple, such as immediate feedback for an activity or getting a correct answer to a multiple-choice test.

Skinner and his machine 2

Figure 3.1 YouTube video/film of B.F. Skinner demonstrating his teaching machine, 1954

You can see a fascinating five minute film of B.F. Skinner describing his teaching machine in a 1954 YouTube video, either by clicking on the picture above or at:

Underlying a behaviourist approach to teaching is the belief that learning is governed by invariant principles, and these principles are independent of conscious control on the part of the learner. Behaviourists attempt to maintain a high degree of objectivity in the way they view human activity, and they generally reject reference to unmeasurable states, such as feelings, attitudes, and consciousness. Human behaviour is above all seen as predictable and controllable. Behaviourism thus stems from a strongly objectivist epistemological position.

Skinner’s theory of learning provides the underlying theoretical basis for the development of teaching machines, measurable learning objectives, computer-assisted instruction, and multiple choice tests. Behaviourism’s influence is still strong in corporate and military training, and in some areas of science, engineering, and medical training. It can be of particular value for rote learning of facts or standard procedures such as multiplication tables, for dealing with children or adults with limited cognitive ability due to brain disorders, or for compliance with industrial or business standards or processes that are invariant and do not require individual judgement.

Finally, it should be noted that behaviourism, with its emphasis on rewards and punishment as drivers of learning, and on pre-defined and measurable outcomes, is the basis of populist conceptions of learning among many parents, politicians, and, it should be noted, computer scientists interested in automating learning. It is not surprising then that there has also been a tendency until recently to see technology, and in particular computer-aided instruction, as being closely associated with behaviourist approaches to learning, although we shall see that this does not necessarily follow.


An obvious criticism of behaviourism is that it treats humans as a black box, where inputs into the black box, and outputs from the black box, are known and measurable, but what goes on inside the black box is ignored or not considered of interest. However, humans have the ability for conscious thought, decision-making, emotions, and the ability to express ideas through social discourse, all of which may be highly significant for learning. Thus we will likely get a better understanding of learning if we try to find out what goes on inside the black box. Cognitivists therefore have focused on identifying mental processes – internal and conscious representations of the world – that they consider are essential for human learning. Fontana (1981) summarises the cognitive approach to learning as follows:

The cognitive approach … holds that if we are to understand learning we cannot confine ourselves to observable behaviour, but must also concern ourselves with the learner’s ability mentally to re-organize his psychological field (i.e. his inner world of concepts, memories, etc.) in response to experience. This latter approach therefore lays stress not only on the environment, but upon the way in which the individual interprets and tries to make sense of the environment. It sees the individual not as the somewhat mechanical product of his environment, but as an active agent in the learning process, deliberately trying to process and categorize the stream of information fed into him by the external world.’ (p. 148)

Thus the search for rules, principles or relationships in processing new information, and the search for meaning and consistency in reconciling new information with previous knowledge, are key concepts in cognitive psychology. Cognitive psychology is concerned with identifying and describing mental processes that affect learning, thinking and behaviour, and the conditions that influence those mental processes.

© Agile Development Blog, 2013

© Agile Development Blog, 2013

Figure 3.2: Some of the areas covered by cognitivism, based on Bloom’s taxonomy (1956). Note that this becomes a reductionist exercise, as psychologists delve deeper into each of these cognitive activities to understand the underlying mental processes.

Cognitive approaches to learning cover a very wide range. At one end, the objectivist end, cognitivists consider basic mental processes to be genetic or hard-wired, but can be programmed or modified by external factors, such as new experiences. Early cognitivists in particular were interested in the concept of mind as computer, and more recently brain research has led to a search for linking learning to the development and reinforcement of neural networks in the brain. In terms of practice this concept of mind as computer has led to several technology-based developments in teaching, including:

  • intelligent tutoring systems, a more refined version of teaching machines, based on analysing student responses to questions and redirecting them to the appropriate next steps in learning. Adaptive learning is the latest extension of such developments;
  • artificial intelligence, which seeks to represent in computer software the mental processes used in human learning (which of course if successful would result in computers replacing many human activities – such as teaching, if learning is considered in an objectivist framework.)
  • pre-determined learning outcomes, based on an analysis and development of different kinds of cognitive activities, such as comprehension, analysis, synthesis, and evaluation
  • certain instructional design approaches that attempt to manage the design of teaching to ensure successful achievement of pre-determined learning outcomes or objectives.

On the other hand, many other cognitivists, coming from a more constructivist epistemological perspective, would argue that mental states or even processes are not fixed but constantly evolving as new information is integrated with prior knowledge, and new strategies for seeking meaning are developed by the individual. Thus teachers who place a strong emphasis on learners developing personal meaning through reflection, analysis and construction of knowledge through conscious mental processing would represent much more of a constructivist epistemological position. It is here that the boundaries between cognitivist and constructivist learning begin to break down.

Cognitive approaches to learning, with a focus on comprehension, abstraction, analysis, synthesis, generalization, evaluation, decision-making and creative thinking, seem to fit much better with higher education than behaviourism,  but even in k-12 education, a cognitivist approach would mean for instance focusing on teaching learners how to learn, on developing stronger or new mental processes for future learning, and on developing deeper and constantly changing understanding of concepts and ideas.

Put simply, brains have more plasticity, adaptability and complexity than current computer software programs, and other factors, such as emotion, motivation, self-determination, values, and a wider range of senses, make human learning very different from the way computers operate, at least at the moment. Education would be much better served if computer scientists tried to make software to support learning more reflective of the way human learning operates, rather than trying to fit human learning into the current restrictions of behaviourist computer programming.

Nevertheless, cognitivists have increased our understanding of how humans process and make sense of new information, how we access, interpret, integrate, process, organize and manage knowledge, and have given us a better understanding of the conditions that affect learners’ mental states.


Both behaviourist and some elements of cognitive theories of learning are deterministic, in the sense that behaviour and learning are believed to be rule-based and operate under predictable and constant conditions over which the individual learner has no or little control. However, constructivists emphasise the importance of consciousness, free will and social influences on learning. Carl Rogers (1969) stated that: ‘every individual exists in a continually changing world of experience in which he is the center.’ The external world is interpreted within the context of that private world. The belief that humans are essentially active, free and strive for meaning in personal terms has been around for a long time.

Constructivists argue that individuals consciously strive for meaning to make sense of their environment in terms of past experience and their present state. It is an attempt to create order in their minds out of disorder, to resolve incongruities, and to reconcile external realities with prior experience. The means by which this is done are complex and multi-faceted, from personal reflection, seeking new information, to testing ideas through social contact with others. Problems are resolved, and incongruities sorted out, through strategies such as seeking relationships between what was known and what is new, identifying similarities and differences, and testing hypotheses or assumptions. Reality is always tentative and dynamic.

For many educators, the social context of learning is critical. Ideas are tested not just on the teacher, but with fellow students, friends and colleagues. Furthermore, knowledge is mainly acquired through social processes or institutions that are socially constructed: schools, universities, and increasingly these days, online communities. Thus what is taken to be ‘valued’ knowledge is also socially constructed. Thus knowledge is not just about content, but also values. One set of values are those around the concept of a liberal education. According to this ideology, one of the principal aims of education is that it should develop a critical awareness of the values and ideologies that shape the form of received knowledge. This then suggests a constant probing and criticism of received knowledge.

One consequence of constructivist theory is that each individual is unique, because the interaction of their different experiences, and their search for personal meaning, results in each person being different from anyone else. Thus behaviour is not predictable or deterministic, at least not at the individual level. The key point here is that learning is seen as essentially a social process, requiring communication between learner, teacher and others. This social process cannot effectively be replaced by technology, although technology may facilitate it.

It can be seen that although constructivist approaches can be and have been applied to all fields of knowledge, it is more commonly found in approaches to teaching in the humanities, social sciences, education, and other less quantitative subject areas.

Online collaborative learning

The concurrence of both constructivist approaches to learning and the development of the Internet has led to the development of a particular form of constructivist teaching, originally called computer-mediated communication (CMC), but which has developed into what Harasim (2012) now calls online collaborative learning theory (OCL). She describes OCL as follows (p. 90):

OCL theory provides a model of learning in which students are encouraged and supported to work together to create knowledge: to invent, to explore ways to innovate, and, by so doing, to seek the conceptual knowledge needed to solve problems rather than recite what they think is the right answer. While OCL theory does encourage the learner to be active and engaged, this is not considered to be sufficient for learning or knowledge construction……In the OCL theory, the teacher plays a key role not as a fellow-learner, but as the link to the knowledge community, or state of the art in that discipline. Learning is defined as conceptual change and is key to building knowledge. Learning activity needs to be informed and guided by the norms of the discipline and a discourse process that emphasises conceptual learning and builds knowledge.

This approach to the use of technology for teaching is very different from the more objectivist approaches found in computer-assisted learning, teaching machines, and artificial intelligence applications to education, which primarily aim to use computing to replace at least some of the activities traditionally done by human teachers. With online collaborative learning, the aim is not to replace the teacher, but to use the technology primarily to increase and improve communication between teacher and learners, with a particular approach to the development of learning based on knowledge construction assisted and developed through social discourse. This social discourse furthermore is not random in OCL, but managed in such a way as to ‘scaffold’ learning, by assisting with the construction of knowledge in ways that are guided by the instructor, that reflect the norms or values of the discipline, and that also respect or take into consideration the prior knowledge within the discipline.


Connectivism is a relatively new theory of learning or epistemology (there’s not even agreement about which it is), it is still being refined and developed, and it is currently highly controversial, with many critics. Siemens, Downes and Cormier constructed the first massive open online course (MOOC), Connectivism and Connective Knowledge 2011, partly to explain and partly to model a connectivist approach to learning. More recently, Downes (2014) has spelled out, in a presentation called The MOOC of One, some of the relationships between individual learning, the contribution of individuals to knowledge and its flow, and networks of learners, within a broad interpretation of connectivist theory. In this presentation Downes sets out some design principles for  connectivist ‘courses’ or cMOOCs, such as:

  • learner autonomy, in terms of choice of content and how they choose to learn
  • openness, in terms of  access to the course, content, activities and methods of assessment
  • diversity: varied content, individual perspectives and multiple tools, especially for networking learners and creating opportunities for dialogue and discussio
  • interactivity: ‘massive’ communication between learners and co-operative learning, resulting in emergent knowledge
Figure 2.1: A map of connectivism, © Stephen Downes, 2011 (accessed via

Figure 2.1: A map of connectivism, © Stephen Downes, 2011 (accessed via

Connectivists such as Siemens and Downes tend to be somewhat vague about the role of teachers or instructors, as the focus of connectivism is more on individual participants, networks and the flow of information and the new forms of knowledge that result.. The main purpose of a teacher appears to be to provide the initial learning environment and context that brings learners together, and  to help learners construct their own personal learning environments to enable them to connect to ‘successful’ networks, with the assumption that learning will automatically occur as a result, through exposure to the flow of information and the individual’s autonomous reflection on its meaning. There is no need for formal institutions to support this kind of learning, especially since such learning often depends heavily on social media readily available to all participants.

There are numerous criticisms of the connectivist approach to teaching and learning, which include:

  • there is no control on the quality of content, or on contributions from participants;
  • assessment strategies, such as peer assessment, are primitive and unreliable, thus making reliable or valid recognition of achievement more difficult;
  • the kinds of learning that take place in connectivist MOOCs or courses are not necessarily academic, in the sense of meeting the requirements for academic knowledge, as defined in Chapter 2;
  • many participants struggle with the lack of structure and are overwhelmed by the volume of content generated by other learners;
  • most students need a high level of explicit support in learning from an ‘expert’ teacher and this is lacking in connectivist courses
  • this kind of learning requires learners already to have at least some level of more formal or traditional education before they participate if they are to fully benefit from this kind of learning experience (and there is substantial evidence that MOOC participants tend to have an already high level of post-secondary education).
  • thus this kind of learning is more appropriate for non-formal learning or communities of practice than for formal education.

Some of these criticisms may be overcome as practice improves, as new tools for assessment, and for organizing co-operative and collaborative work with massive numbers, are developed, and as more experience is gained. More importantly, connectivism is really the first theoretical attempt to radically re-examine the implications for learning of the Internet and the explosion of new communications technologies.


Different theories of learning reflect different positions on the nature of knowledge. With the possible exception of connectivism, there is some form of empirical evidence to support each of the theories of learning outlined here.

However, while the theories suggest different ways in which all people learn, they do not automatically tell teachers or instructors how to teach. Indeed, theories of behaviourism, cognitivism and constructivism were all developed outside of education, in experimental labs, psychology , neuroscience, and psychotherapy respectively. Educators have had to work out how to move from the theoretical position to the practical one of applying these theories within an educational experience. In other words, they have had to develop teaching methods that build on such learning theories. The next section of the book examines a range of teaching methods that have been developed, their epistemological roots, and their implications for teaching in a digital age.

Over to you

Your feedback on this will be invaluable. In particular:

  • are theories of learning still relevant in a digital age? Is it important to discuss these?
  • is the description of the various theories accurate and useful; if not, what should be changed?
  • are there important theories or theoretical positions that have been missed?


Bloom, B., Englehart, M., Furst, E., Hill, W. and Krathwohl, D. (1956) Taxonomy of Educational Objectives, Handbook I: The Cognitive Domain, Longmans Green, New York, 1956

Downes, S. (2014) The MOOC of One, Stephen’s Web, March 10

Fontana, D. (1981) Psychology for Teachers London: Macmillan/British Psychological Society

Harasim, L. (2012) Learning Theory and Online Technologies New York/London: Routledge

Knapper, C. (2010) ‘Changing Teaching Practice: Barriers and Strategies’ in Christensen Hughes, J. and Mighty, J. eds. Taking Stock: Research on Teaching and Learning in Higher Education Toronto ON: McGill-Queen’s University Press

Lewin, K. (1951) Field theory in social science; selected theoretical papers. D. Cartwright (ed.). New York: Harper & Row.

Rogers, C. (1969) Freedom to Learn Columbus, OH: Charles E. Merrill Publishing Co.

Schunk, D. (2011) Learning Theories: An Educational Perspective (6th edition) New York: Pearson

Discussion of MOOCs: more links and questions

Rodriguez, O. (2012) Vast Lurker and No-lurker Participation in Open Online Courses: MOCCs and the AI Stanford-like courses respectively Osvaldo Rodriguez, March 3

Siemens, G. (2012) MOOCs for the Win! ELearnspace, March 5

Lewin, T. (2012) Instruction for Masses Knocks Down Campus Walls, New York Times, March 4

Following my two earlier posts on MOOCs (Some critical reflections on MOOCs and More reflections on MOOCs and MITx) and a response from Stephen Downes and Sui Fai John Mack, I am adding three more posts that deliberately or accidently continue the discussion.

George Siemens

First let’s start with George Siemens, one of the original designers of a particular kind of MOOC based on a connectivist approach. His post sets out some of the history behind the development of MOOCs and responds to a post from Clark Quinn and also to my two posts.

In conclusion to his post, George Siemens write:

It is important to realize that MOOCs are not (yet) an answer to any particular problem. They are an open and ongoing experiment. They are an attempt to play with models of teaching and learning that are in synch with the spirit of the internet. As with any research project, it is unlikely that they will be adopted wholesale in traditional universities. Most likely, bits and pieces will be adopted into different teaching models. Some systems will offer open online courses as a means of drawing attention to their university. Others will offer MOOCs because it’s an effective way of getting out an important message or to raise awareness about certain topics.

Any or all of those adoptions of MOOCs are not really a concern for me. I’m more interested in experimentation and exploring new modes of interaction online. I’m not concerned about whether or not existing university systems adopt MOOCs for undergraduate education or whether they serve to improve continuing education. That kind of discourse appropriates MOOC concepts to support the narrative of the existing education system. Which is fine.

But that is only one way to look at MOOCs.

Osvaldo Rodriguez

Osvlado is an Argentinian blogger and he has a very interesting post that compares participation rates between different kinds of MOOCs. He makes the distinction between MOOCs such as the Stanford AI MOOC based on a cognitivist-behaviourist methodology and ‘connectivist’ MOOCs, such as #Change 11 and EduMOOC.

He ran a statistical analysis comparing EduMOOC with the Stanford AI MOOC that showed that after a few weeks, active participation in EduMOOC had dropped from 2,700 to just over 100. The Stanford AI MOOC started at around 160,000 active participants then dropped rapidly to a pretty steady 25,000 active participants a week. He argues that although less than 10% of the original ‘starters’ in the EduMOOC actively participated, there were a lot of ‘lurkers’ reading but not otherwise participating, whereas with the Stanford AI MOOC, students could not lurk; if they did not take the obligatory exams they were ‘non-completers’.

His summary:

From previous studies it has become evident (George Siemmens 2012) that we are in the presence of different formats:
    • the AI-Stanford participants have totally different learners goals and preparation than those in MOOCs.
    • there exists a very different nature of the subjects studied: engineering  and  educational theory.
    • the AI-Stanford course falls into the cognitive-behaviorist pedagogy category and the MOOCs  into the connectivist.
The retention and lurker behavior described above adds another differentiation to the previous list.
In my view, though, the study raises more questions than it answers.
  • First what counts as success? It seems to me that 25,000 students successfully completing a course on AI is pretty good. However, losing over 95% of participants in the EduMOOC and ending up with barely 100 active participants does not seem very successful.
  • Does this mean that cognitive-behaviourist design is more successful than a connectivist design? I don’t think so. Other factors have to be taken into consideration.
  • Is connectivism on its own sufficient to achieve success on an online course, or could other strategies, such as better software for organizing content and better learning design, increase participation rates, maintain active presence during the MOOC, and above all lead to deeper learning?
The New York Times
This article is a report on a number of MOOCs, but focusing particularly on the Stanford experience.
First I found it interesting that  ‘an additional 200 registered for the course on campus, but a few weeks into the semester, attendance at Stanford dwindled to about 30, as those who had the option of seeing their professors in person decided they preferred the online videos‘. This says a lot about the quality of face-to-face teaching, as well as the online course. If you design a course in a very cognitive-behaviourist way it lends itself to automation. Shouldn’t the face-to-face class have been doing something different?
Second, as usual, mainstream journalism over-hypes the development. Headlines such as ‘Instruction for Masses Knocks Down Campus Walls’ and ‘Welcome to the brave new world of Massive Open Online Courses — known as MOOCs — a tool for democratizing higher education’ sets up (again in my view) unrealistic expectations for what is a very interesting but still developing phenomenon. It also by implication undervalues the already excellent online teaching and open learning programs that are going on in a less publicized way in formal education. For instance, reading the NYT, one would think that the UK Open University had never existed, nor that many leading universities and colleges are offering thousands of online courses and have been doing so for many years.
In particular, the designation of MOOCs as ‘democratizing education’ really needs to be carefully examined. Presumably, 135,000 learners who wanted to learn about AI were disappointed or unable to follow the Stanford AI course. If this was really about democratizing education, these students should have been accommodated in some way.
With regard to connectivist MOOCs, I worry whether they are just preaching to the converted by reinforcing participants’ existing knowledge or values, or whether they lead to significant change in learners. They may do or they may not. We need more research on this. Octavo’s simple research study, although valuable, just reinforces the need for more thorough research, and we also need more experimentation, with different designs and approaches.
So I am definitely agreeing with George Siemens here that MOOCs are a fascinating, valuable ongoing development, but let’s approach them in the spirit of critical analysis and research, with the aim of constant improvement and refinement.

IRRODL on Connectivism

IRRODL (the International Review of Research in Open and Distance Learning) has just published a very special issue focused on connectivism.

Connectivism claims to be a powerful new learning theory that exploits the power of networks and networking to support learning. The term was first coined in 2004 by George Siemens (Athabasca University), who along with Grainne Conole (Open University, UK) is the guest editor of this issue. To the editors’ knowledge this is the first full peer-reviewed journal issue focused on connectivist ideas, ideals, practices, and criticism.

The nine articles in this issue were winnowed from a much larger set of submissions to provide both supportive and critical commentary and research results.

From the editorial:

this special issue of IRRODL presents a somewhat confusing landscape. Some themes are emerging around the relationship of connectivism to existing theories of learning and social interaction (communities of practice, actor-network theory, and activity theory being most prominent). Critiques of connectivism also reveal themes: the need for ongoing research, the suitability of existing theories in answering the questions that connectivism attempts to address, and the status of connectivism as a theory of learning….It seems futile to debate the merits of connectivism versus behaviourism, cognivitism, or constructivism. Instead, several questions arise. Which theory best maps to the reality of a particular subject content? Which theory most effectively embraces the ‘adjacent possible’ of our technologically based society? Which theory best meets current and future learning needs of learners?

Personal comments

I have to say that I have struggled for some time both to understand exactly what is the theory of connectivism, other than the importance of online social networks for learning, which I agree with but don’t find particularly helpful in pragmatic terms, and I’ve also struggled to understand the extent to which learning actually takes place within a connectivist framework. In other words, I have been looking for a more coherent theoretical framework, and some empirical evidence to support the theory. In an unstructured way that is a feature of connectivism, this edition of the journal goes a long way to providing what I have been looking for. However, it is clear that I need a community of practice to help me pull all this together!

I found the article by Rita Kop to be particularly helpful. She provides a concise definition of connectivism:

Connectivists advocate a learning organization whereby there is not a body of knowledge to be transferred from educator to learner and where learning does not take place in a single environment; instead, it is distributed across the Web, and people’s engagement with it constitutes learning.

and draws attention to three potential limitations in its application:

  1. the need for critical literacies and the power relations on the network;
  2. the level of learner autonomy;
  3. the level of presence.

She argues that These [limitations] can all be overcome by what has in traditional formal educational practice been seen as crucial to teaching and learning: social interaction.

Her article and several of the others then demonstrate how social interaction, in a variety of ways, can provide the ‘essentials’ for learning through informal networks.

The articles collectively also demonstrated the permeability of learning through both formal and informal learning structures. For the individual learner, these barriers or distinctions don’t make a lot of sense – they will often need both forms of learning.

There is too much in this edition for me to summarise – this is a case where you need to read every article to build up a more coherent picture of connectivism, and as I said it’s like a rich meal – I’m still struggling to digest a lot of it.

However, my final impression is that arguments over which is the most effective form of online learning, formal learning through structured courses, or informal learning through non-heirarchical social networks, are pointless. Informal learning has always been important, and the Internet immensely strengthens the opportunities and range of informal learning. At the same time, as Rita Kop makes clear, the conditions for effective learning through informal networks are quite demanding and will often not be present. Formal learning can (if properly designed) offer opportunities for learning that may not possible in many contexts through informal learning. In a sense, there is much to be learned from both approaches to learning. It is the context that will determine which is the most appropriate for a particular individual at a particular time, and learning will not be neatly separated by these boundaries.

This is one of the most important editions of IRRODL, and although the reading isn’t always easy, I strongly recommend the edition for anyone who wishes to understand the future development of online learning.

Table of Contents

George Siemens, Grainne Conole

Research Articles

Interconnecting networks of practice for professional learning
Julie Mackey, Terry Evans

The challenges to connectivist learning on open online networks: Learning experiences during a massive open online course
Rita Kop

Emergent learning and learning ecologies in Web 2.0
Roy Williams, Regina Karousou, Jenny Mackness

EduCamp Colombia: Social networked learning for teacher training
Diego Ernesto Leal Fonseca

Three generations of distance education pedagogy
Terry Anderson, Jon Dron

Connectivism: Its place in theory-informed research and innovation in technology-enabled learning
Frances Bell

Frameworks for understanding the nature of interactions, networking, and community in a social networking site for academic practice
Grainne Conole, Rebecca Galley, Juliette Culver

Dialogue and connectivism: A new approach to understanding and promoting dialogue-rich networked learning
Andrew Ravenscroft

Proposing an integrated research framework for connectivism: Utilising theoretical synergies
Bopelo Boitshwarelo