Monday, January 01, 2007

_isms as filter, not blinker

Learning theory, like politics, is full of _isms: constructivism, behaviourism, cognitivism and now a new one, connectivism. What should we do about these _isms? Are they a useful guide to what to think and do?

In October I wrote this on the home page of the learning evolves wiki:
Learning theories might help deliver radical curriculum reform. I think that’s why we have all those -isms (constructivism, behaviourism, connectivism, etc.) and that although -isms can be dangerous we still have them and they might be necessary. Because how else could we have a big change without a theory to justify it and help us think about it? Should we stick to -isms or should we be more pragmatic and just cherry pick different useful ideas out of the various theories? I’m not entirely sure but I am more inclined to think that we need big change. That might mean the -isms are necessary. You might develop a new unit of work under the influence of constructionism, for instance. The learning theory is indispensible to the curriculum reform effort.
What I have noticed is that these _isms do not stand still. They evolve, they listen to criticism and move on. I've also noticed that learning theorists, who have a different favourite _ism to mine, might still come up with significant findings in their empirical studies that I find hard to reject or ignore. So, although it is possible to make perfectly valid criticisms of Skinner's behaviourism or the theoretical foundation of cognitivism that is not the end of the story.

That premable helps explain my responses to some recent blogs by Stephen Downes which has become a dialogue between him and Karl Kapp about learning theories. In reponse to this post by Karl Kapp, Stephen began with a blanket rejection of behaviourism:
... it remains puzzling that so much of the instructional design community remains rooted in behaviorism - this more than 30 years after the theory was abandoned everywhere else
- Definitions: ABCD Objectives
I wrote this in disagreement on Stephen's blog:
Philosopher Daniel Dennett has extended the core correct concept of behaviourism (generate and test) into the inner environment. It's not correct to say that the "theory was abandoned everywhere else". Actions which are followed by rewards are often repeated. Doesn't that make us all behaviourists, despite many excellent critiques of Skinner?

Daniel Dennett. Why the Law of Effect Will Not Go Away. Chapter 5 of his book, Brainstorms
From my perspective see more detail about Dennett's argument: Dennett's Creatures

Karl Kapp replied and Stephen continued with his critique:
Kapp writes, "For mission critical items, we cannot write an objective like: The nuclear technician, upon encountering a meltdown of the primary reactor will use a discovery method to explore possible options for stopping the meltdown... We really need something like: The nuclear technician, upon encountering a meltdown of the primary reactor will follow a defined set of steps to stop the meltdown." But this is not true, and the proof is this: if it were true, then the human performance could be replaced by a machine. If you are working simply on stimulus-response, then you are working on programmable behaviour. But we use humans in nuclear reactors (and elsewhere) just because we understand that 'knowing' involves a set of cognitive processes - like recognition, inference, association - between stimulus and response. The difficulty is, of course, convincing politicians, customers, and (apparently) instructional designers of this
- Design: Behaviourism has its place
I responded again on Stephen's blog:
stimulus: behaviourism
stephen downes response: dehumanising

Now that the best chess player in the world is a machine should we stop playing chess or reconceptualise it as an undesirable human activity?

I haven't researched it but think it very likely that in nuclear power plants activities previously carried out by humans are now carried out by machines. It has happened everywhere else. Machines now do a variety of maths type behaviours better than humans (algebra, chess, etc.). If machines evolve further and start displaying visceral emotions then how should we deal with that? An alternative to dehumanising humans would be to humanise machines. At that point behaviourism might make a come back.
Changing topic, Karl Kapp then posted about cognitivism and Stephen's response was:
I have always depicted cognitivism as a response to behaviourism and also as a philosophy of learning and of mind to which I stand essentially opposed (and no, that does not make me a behaviourist). "The idea is that the learner is a complex information-processing system and to understand how learning occurs, one must understand how information processing occurs within the human brain... in the cognivitist's view learning occurs internally and through the social interactions with others." Now how could I disagree with that? It is with the central concept: 'information processing'. The mind is not like a computer, at least, not like most any computer we've build, and depicting the mind as analagous to (and governed by the rules governing) symbol system processors is to misrepresent it in a fundamental way. In my view
- Definition: Cognitivism
Again I have left a comment on Stephen's blog, as follows:
I agree with you that the architecture of the mind is very different from that of a computer (including connectionist machines.)

Nevertheless, at the level of empirical studies some who fit under that broad umbrella do useful work IMO, eg. Ericsson on the role of deliberate practice in the acquisition of expert performance

ie. Ericsson is emphasising deliberate practice I suppose because that does fit an information processing model whereas someone with a constructivist perspective, for example, is looking intently at some other aspect of learning (eg. rich, exploratory learning environment but with some implied or overt guidelines)

It seems to me that each _ism is offering something useful without any of them being complete or stand alone in their own right
My Conclusion: _isms are important but use them as a filter, not as a blinker


Miguel said...

So, are you saying that while the -isms provide valuable insights into human learning, they are imperfect, undergo constant revision, and, as such, they only suffice as magnifying lens on a microscope? A way of understanding human learning?

What makes any -ism superior to another? Why should Siemen's connectivism be any more than an exciting metaphor for human learning than constructivism? Is it only that one has been around longer than the other?

As we humanize machines, what it is to be human is explored more. I can be MORE human because a machine takes care of the drudgery of work.

Finally, how hard is it to debunk a learning theory? Learning theories seem to serve only insomuch as they help us get passionate about human learning and the possibility that we may be able to predict how people learn...but isn't that process of learning individual? Shouldn't any learning theory help us better enable people to learn individually? How can technology help us do that?

Bill Kerr said...

hi miguel,

your first para: yes

second para: theory-practice spiral is the best (only reliable) way to test theories. more here . That is my bias openly declared.

third para: I agree

fourth para: There are a lot of different learning theories! I'm discovering that it does require hard work to get your head around them. I think some ideas are better than other ideas, it's not helpful to just generalise, "we are all individuals". I don't think permanent lecture mode will work very well for anyone. I like the ideas of seymour papert and alan kay wrt technology being used to advance learning further.

Karl Kapp said...


I particularly like your comment about the concept of taking a little bit from each school of thought, I really believe that is the essense of good educational design. Creating learning utilizing an entire tool kit of philosphies, techniques and ideas.

Also, a good job summarizing the discuss that the three of us have been having on the topic. Kudos.


Bill Kerr said...

Karl has expanded on his comments at his blog, here is an extract:
"So the next question you ask is “What is the best, how do we know what makes sense or what doesn’t?” I suggest that lower level learning (lower cognitive load) requires a behaviorist approach (memorize, recognizing, labeling) as does the expectation of outcomes that must be measured. I then suggest that procedural and rule-based learning requires an emphasis on Cognitivism and finally, problem-solving, collaboration and creativity require a view of Constructivism"

Also there is a new entry by Karl on Constructivism and Stephen Downes and myself have added comments there

Paul Kremer said...

Hi Bill, I am a frequent visitor to Karl Kapps' blogspot, I stumbled on his page when searching for learning design research and material for my first attempt at a book.
I have posted a couple of excerpts on a new blog spot at
I would love your feedback. And any additional information you may have that would assist in my quest, to help society, would be great also.
PS Love your blogspot, I will certainly learn a great deal from you and your opinions.

Thanks Rgds Paul

Pennyfree said...

I think that the statement “it seems to me that _ism is offering something useful without any of them being complete or stand alone in their own right” really puts a cap on it from an educational standpoint. As in all educational curriculums, no one objective should stand alone and no one subject should stand alone. We need the integration of across-the-board curricula just like we need the integration of all of the _isms. PENNY

Dreana Marshall-Stuart said...

In the blogs, discussion on Behaviourism aimed at quantifying behaviour, to make it more scientifically observable. I do agree that Behaviourism is observable, and can effectively be measured through research, experiments, interviews, surveys, case studies, and questionnaires. Behaviorism focuses on a new behavioral pattern being repeated until it becomes automatic. I agreed with this premise especially when we look at the experiments conducted by Skinner in terms of conditioning.

The reality is that although cognition is not observable, we know it exists. Are we to conclude that cognition explains the thought process behind the behavior of a person. Yes, changes in behavior can be observed, but how do we effectively use these changes as indicators as to what is happening inside the learner's mind? The process of learning cognitively which was compared to that of the Central Processing Unit (CPU) or brain of a computer where we as humans process information through receiving, storing and retrieving it was disagreed upon in Stephen's response. I have somewhat to agree with Stephen based on my personal belief in the concept “GIGO” – Garbage In, Garbage Out. So although scientists believe humans process information like computers, we must also remember that the information computers receive, store and retrieve is what the user puts in. So if we indeed take in ‘garbage’ that is what we will store and retrieve – can we then consider that learning has indeed taken place.
I agreed with Bill Kerr’s comments that each –ism being offered on its own does have something to offer us however incomplete they maybe. I also agree that if we take bits and pieces from each theory we can use them collectively to our benefit. The suggestion to use the bits taken from each theory as a scaffolding process was also an excellent idea. It is true that no one theory can stand alone and be strong in itself, but if we apply different aspects of each theory to fill a particular learning need then we may be able to design a course that can be consider exceptional in terms of the learning processes. I think Bill said he beast when the stated that the essence of good educational design may be had in taking a little bit from each school of thought.
As instructional designers we are going to find ourselves applying both of these theories. For instance, as instructors we are going to find ourselves heavily relying on Bloom's Taxonomy of Learning, which was constructed from a cognitive perspective. We have now find ourselves shifting to a Cognitive form of teaching over that of behaviorism as we become more concern with the internal mental processes of the mind and how they could be used in encouraging effective learning. For example, in using the behaviorist approach to designing a lesson we will have probably broken down a task into small steps in an effort to shape the learner’s behavior. Now, if we were to use the cognitive approach we would have analyze the task differently. Yes, we would have broken the task down steps albeit smaller steps or chunks of information. We would have then used that information to develop the learning process to move from simple to complex learning, building on prior schema as with Bloom's Taxonomy of Learning.

Solomon Nariwo said...

The notion that behaviourism was abandoned is somehow setting strict boundaries on the different theories and I also do not agree. Most of the corporate trainings assumes the behaviourism theory (especially for softskill trainings). The objectives are targeted towards a change in response (or behaviour),you hear words like competence, goals and ultimately, there are reinforcements
factored into such trainings especially from the HR departments. I share Bill's view that each of the -isms has something to contribute and should be considered in contextual relevance. They should be applied on a case by case bases.
Just like comparison between classical mechanics and relativistic theory, one does not replace the other but in my opinion, they fill up the gaps where necessary and are thus both necessary.

Software Trainer/Technology Coordinator said...

Greetings Bill Kerr:

It has been 3 years and the discussion continues.

Each-ism represents the track record of an idea that has been created, tested and found true in a majority of cases. Not 100%, but more than 50% which makes them all viable. Since no two people are alike, we will never get to a point where one -ism will serve all. Therefore, the prudent course of action at this point may be to extend this dialog beyond the brain trust and move it amongst the students.

Explain each -ism and provide examples then ask the question... "How do you learn?" This course of action may prove beneficial as teachers evaluate the learning style of their class and design instruction to meet the needs of learners.

IDT Student said...

As I read the postings and comments, (although some are 3 years old), I can’t help but think about the most recent horrific incident that took place in the Gulf of Mexico. The BP oil spill claimed 11 lives and endangered the Gulf for many years to come.

This incident comes to mind simply because we are talking about how the mind processes information and how our behavior reacts to the information we process. From a training perspective, I’m sure there was safety training that each employee on that rig had to attend. Yet, when the incident was about to erupt, nobody was prepared. BP’s CEO Tony Hayward stated that millions was spent on training. So, why did this happen if these employees were trained to recognize and react to emergencies? Was it a lack in memory retrieval? Was it poor training to start? Perhaps they did know what to do, but didn’t want to disrupt or make waves in case of being was wrong? Or was the necessary behavior to properly react to this emergency lacking “ism”?

I would assume, although not validated, that the safety training BP provided was the same or very similar for all employees who work on a rig. I use this one case as an analogy, but numerous documented disasters identified the safety training successful, but executed application of the information was not.

Behaviorists theorize stimulus/ response, but the human mind has to fill in the activity taking place between the stimulus and the response. A theory and “ism”, to me, is only as good as the outcome of the situation. Computers have certainly changed our world and from a functional standpoint, probably for the better. However, the growth of technology continues to become replacements for people. Most of these theories were transcribed before computers became a household accessory. We’ve experienced many real world situations outside of the laboratory that indicates theories are only starting points for further investigation.

Kaatri said...

I believe most of the _isms have a place. From our text, Learning Theories and Instruction, I have developed the impression that each _ ism has validity and usefulness in certain circumstances, but not in all. In most cases each has a place, behaviorism is effective as a means to learn facts, problem solving theories describe how facts and experiences become solutions, constructivism is appropriate particularly when we are working on projects of interest or in our careers where creativity and innovation is called for. Each _ism is required for functional learning, but we may acquire needed building blocks for a project 10 years before they become the base to a constructivistic project.

My other observation is, although there are many ways to discuss memory, cognitive informational processes, and learning processes, they all revolve around/ depend upon particular brain processes: encoding, organization, elaboration, meaningfulness, links with schema, analysis, placing learning in broad associated networks which can be activated and require motivation with conditional procedures regulating the appropriate brain process. The –isms are highly related.

One study, “Speech Perception and Language Acquisition in the First Year of Life”
Annual Review of Psychology (1/2010) is an example of an attempt in the field of psychology to meld these _isms together to aid in applying our knowledge of learning effectively. The purpose of this article was to weld differing theories of infant language acquisition as complimentary pieces of the whole. Their conclusion: “We have reviewed evidence suggesting that nativist and empiricist proposals are incomplete if they fail to include innate dispositions and learning in a broader, integrative, biologically anchored language acquisition theory. In addition, we have shown that a third type of mechanism, perceptual and memory constraints, needs to be evoked to provide a full account of early acquisition.”

In conclusion, it seems wise to theories but not to pontificate, to use accumulated wisdom to create multiple learning pathways and environments to meet each learners unique abilities and providing creative challenges that motivate learners.

Roger's Blog said...

I to agree that isms should be used as filter, not a blinker and that all of them can be used at anytime in the learning process. as the behaviorist believe that human behavior resulted from specific stimuli that elicited certain responses. John B Watson (1878-1958) Narrated by Diane Gornell. I know just using the behaviorist thoery lone does not prove that learning has taken place. It just the start . We again need to cherry pick them as needed. without understanding how a task or how the behavior works learning will not take place, but if you add in the cognitive or as stated in the book learning theories and Instruction by Margaret Gredler Metacognition in which first you must understand what skills, strategies, and resources a task requires, learning truly couldn't take place( Gredler, 2009) all of these isms must combine in order to move learning forward. So, with that said, I do agree that it has to be a filter.