Who breaks a butterfly on a wheel: Kirschner, Sweller and Clarke

I think it is important to say what this is and is not. It is not an advocation of progressive techniques neither is it a refutation of traditional techniques. Rather it is  an attempt to look at the polemics of educational discourse or rather educational blogging discourse. Progressive education is receiving a, some might say, well deserved kicking. I think it is only partially deserved but I also think it is based on some dubious research and an underlying political ideology.

In this blog I am going to look at some of the underlying pedagogic arguments in the form of a refutation of Kirschner, Sweller and Clarke. Most of the arguments I have seen are based upon one aspect or other of this paper.

It always seems disingenuous to acknowledge the virtues of papers you are about to critique, however regardless of my opinions of it’s merits, I still think this is a valuable and important contribution to educational debate. It is possible to disagree with (and indeed be quite scathing about) something or other, and at the same time accept the contribution of a particular point of view.

Although I don’t refer to it explicitly, I do differentiate between epistemic constructivism and pedagogic constructivism. The former describes the philosophical theory of constructivism; Vygotsky, Berger and Luckman etc and the latter the instructional, Bruner etc.

The Widest of remits

The paper has the widest of remits. Not just Constructivism or  Minimal Guidance During Instruction but also Discovery, Problem-Based, Experiential, and Inquiry-Based Teaching.In many ways this is justifiable from the perspective that sometimes a particular approach doesn’t come with neatly packaged titles for research

Problematic Assertion no 1

However within the first paragraphs the first problematic assertion is constructed. Here we are led to believe that teachers deliberately withhold “essential information”. I think it unlikely that any teacher has actually done that or that any particular approach advocates teachers learners without essential information.

On one side of this argument are those advocating the hypothesis that people learn best in an unguided or minimally guided environment, generally defined as one in which learners, rather than being presented with essential information, must discover or construct essential information for themselves (e.g., Bruner, 1961; Papert, 1980; Steffe & Gale, 1995).

Problematic Assertion no 2

But it doesn’t end there in the next paragraph we are presented with problematic assertion number 2

On the other side are those suggesting that novice learners should be provided with direct instructional guidance on the concepts and procedures required by a particular discipline and should not be left to discover those procedures by themselves

So we are actually talking about novice learners but there is also another clever sleight of hand. The implication is actually not minimally guided instruction, which could actually be quite structured but one whereby learners are “left to discover those procedure by themselves”.

So the premise of the paper is whether novice learners will learn, if teachers withhold essential information, and then leave them to discover procedure by themselves.

What are we led to believe?

We are led to believe that Jerome Bruner, one of the most respected educational psychologists (or was anyway) and Seymour Papert, an MIT Compute scientist, said that novice learners should not be given essential information but construct it themselves without any help whatsoever, or at best with almost no help whatsoever. Now call me an old cynic but I doubt either ever said that.

In fact Papert says this:

This vision advances the definition of Constructionism and serves as an ideal case against which results that have been actually achieved can be judged. In particular, it illustrates the sense of the opposition I like to formulate as Constructionism vs. Instructionism when discussing directions for innovation and enhancement in education.I do not mean to imply that construction kits see instruction as bad. That would be silly. The question at issue is on a different level: I am asking what kinds of innovation are liable to produce radical change in how children learn. Take mathematics as an extreme example. It seems obvious that as a society we are mathematical under performers. It is also obvious that instruction in mathematics is on the average very poor. But it does not follow that the route to better performance is necessarily the invention by researchers of more powerful and effective means of instruction (with or without computers)

Seymour Papert and Idit Harel’s book Constructionism (Ablex Publishing Corporation, 1991).

Papert actually describes the notion that instruction is bad as “silly”.

Good Constructivism, bad Constructivism

Half way through the paper Kirschner et al say this, “the constructivist description of learning is accurate, but the instructional consequences suggested by Constructivists do not necessarily follow.”

Constructivism seems to be right and wrong at the same time. Here the assertion is that it is the interpretation of constructivism that is the problem not constructivism itself, which is fine but unfortunately constructivism has suffered as a consequence.

It is ironic therefore when Kirschner et al say this:

Because students learn so little from a constructivist approach, most teachers who attempt to implement classroom-based constructivist instruction end up providing students with considerable guidance.

This is a reasonable interpretation, for example, of qualitative case studies conducted by Aulls (2002), who observed a number of teachers as they implemented constructivist activities in their classrooms. He described the “scaffolding” that the most effective teachers introduced when students failed to make learning progress in a discovery setting.

Kirschner,  Sweller and Clarke

Ironic because, scaffolding is one of Bruner’s own concepts and a constructivist approach to teaching and learning.

Cognitive Load Theory (CLT)

At this point we get to the central themes of  Kirschner,  Sweller and Clarke:

… long-term memory is now viewed As the central,dominant structure of human cognition. Every thing we see, hear, and think about is critically dependent on and influenced by our long-term memory. If nothing has changed in long-term memory, nothing has been learned. Any instructional recommendation that does not or cannot specify what has been changed in long-term memory, or that does not increase the efficiency with which relevant information is stored in or retrieved from long-term memory, is likely to be ineffective.

I’m not sure who considers long term memory as the central dominant structure of cognition. I suppose it depends in which area you are working. The amygdala is often cited as a very important influence as are the executive functions.  Regardless the point of this passage seems to be that what you forget is not much use to you. Well I can’t dispute that but as a central premise of knowledge based education you might think it hardly inspirational.

The main theme of Kirschner,  Sweller and Clarke’s work is Cognitive Load Theory (CLT). The implicit assumption is that learning is really about storing long term memories. Presumably the presumption here is that learning how to search for information is not stored anywhere. This highlights a criticism of CLT that in fact it differentiates between what is good LTM, and what is bad LTM or, doesn’t happen at all.

Working memory is the cognitive structure in which conscious processing occurs. We are only conscious of the information currently being processed in working memory and are more or less oblivious to he far larger amount of information stored in long-term memory.

All problem-based searching makes heavy demands on working memory. Furthermore, that working memory load does not contribute to the accumulation of knowledge in long-term memory because while working memory is being used to search for problem solutions, it is not available and cannot be used to learn.

This is not science. This is making a value judgement on searching for stuff. Possibly a good point but not a good basis for scientific theory. Learning how to learn is valuable knowledge.

Indeed, it is possible to search for extended periods of time with quite minimal alterations to long-term memory (e.g., see Sweller, Mawer, & Howe, 1982). The goal of instruction is rarely simply to search for or discover information. The goal is to give learners specific guidance about how to cognitively manipulate information in ways that are consistent with a learning goal, and store the result in long-term memory.

In this passage of text (above) we see that the goal of learning is to store “long term memory”. Again this may be a good point but it is a value judgement. Part of the process of developing new ideas is learning how to test old ones, searching for new ideas etc. Simply stuffing LTM with knowledge maybe someone’s idea of a good education, and it may well be a good education, but it remains a value judgement and not a basis for a scientific theory.

The consequences of requiring novice learners to search for problem solutions using a limited working memory or the mechanisms by which unguided or minimally guided instruction might facilitate change in long-term memory appear to be routinely ignored.

Or possibly some people think that stuffing Long Term Memory (LTM) with knowledge is not the purpose of education.

The proffered solution is this:

In contrast, studying a worked example both reduces working memory load because search is reduced or eliminated and directs attention (i.e., directs working memory resources) to learning the essential relations between problem-solving moves. After a half-century of advocacy associated with instruction using minimal guidance, it appears that there is no body of research supporting the technique. In so far as there is any evidence from controlled studies.

Again I’m not sure why studying worked examples couldn’t be described as “minimal instruction”. Aspects of discovery could be built into a very structured “worked example” approach. The problem seems to be that Kirschner,  Sweller and Clarke seem to regard minimally guided instruction as being one where learners are dumped in a room and left to their own devices. In our SAT’s obsessed education system that is hardly likely to happen or ever have happened.

I can’t imagine any teacher leaving a seven year old in a classroom with a chess set, with the instructions “go figure”, which seems to be the definition of discovery based learning adopted by Kirschner et al, judging by some of the research the paper cites.


Overall it’s an interesting but ultimately dissatisfactory paper, based on polemical definitions of an approach that probably doesn’t exist in the form outlined. In many way’s it is hard to take it seriously, and reads like one of those papers that is talking to a small research audience who have defined a discourse far removed from the classroom.

Who breaks a butterfly on a wheel indeed, progressive education may well turn out to be a blip on the horizon but the problem is not, and never will be, about new ideas, approaches and methods per se but rather it lies in the appropriation of ideas by those who know little about education and who then distort them  for their own purposes.

6 thoughts on “Who breaks a butterfly on a wheel: Kirschner, Sweller and Clarke

  1. ” but the problem is not, and never will be, about new ideas, approaches and methods per se, but rather it lies in the appropriation of ideas by those who know little about education and who then distort them for their own purposes”

    Well said.

    I sometimes wonder why those who seem to know the least often seem to feel that they know the most.

    Great post.

    1. I sometimes wonder why those who seem to know the least often seem to feel that they know the most.

      Indeed and thanks

  2. One possible reason why the Kirshner article may have such life in online education debates (aside from the fact that it is polemical opinion, which some people love), is that 7 years ago a grad student sympathic to its arguments spread it, and his own bias, all over Wikipedia. And, as we know, a lot of people get their information from Wikipedia now: http://edtechdev.wordpress.com/2007/12/26/an-argument-for-knols-over-wikipedia-and-citizendium/

    But citing the Kirshner article has now become a good proxy signal for someone with a limited understanding of teaching, learning, educational research, and theory – someone who just read stuff on Wikipedia.

    1. Thanks Doug couldn’t agree more. I also think its longevity is related to its accessibility. Quite simply it sounds plausible

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s