This blog was stimulated by the recent work of Paul Kirschner, a colleague in the broad church of educational researchers. Recently I saw on facebook that he had produced some materials for teachers called, perhaps provocatively, ‘Fully Guided Instruction’. The title evokes his attack on constructivist teaching in the highly cited and influential article: ‘Why minimal guidance during instruction does not work: An analysis of the failure of constructivist, discovery, problem-based, experiential, and inquiry-based teaching’. I recommend both the article and the materials. But not uncritically or unequivocally. One positive value of Paul’s recent work for me is pushing us to think more about the awkward relationship between facts and values in educational research.
Constructivist theories of teaching which involve minimal teaching, the idea that teachers need to get out of the way in order to facilitate the learning of students, are an over-extrapolation, Paul claims of perfectly sensible constructivist models of learning. In other words, while it is true that children learn by constructing their own meaning it does not follow that they do not need to be explicitly taught and guided by teachers to help them do this. Apparently while it is true that experts in many domains learn together through collaborative inquiry the same is not true of novices who need to be explicitly taught. This argument made me think more about the general problem in educational theory of moving too fast from a theory of how students learn (what ‘is’) to a theory of how we should teach (what ‘ought to be’).
The shift here from what is to what ought to be looks like a particular instantiation of the fact/value problem in philosophy introduced by David Hume. [http://oll.libertyfund.org/titles/hume-a-treatise-of-human-nature p469]. Hume was aiming at moralists who pretend to ground their values on the way that things are. In 1739 he wrote:
‘In every system of morality, which I have hitherto met with, I have always remark’d, that the author proceeds for some time in the ordinary way of reasoning, and establishes the being of a God, or makes observations concerning human affairs; when of a sudden I am surpriz’d to find, that instead of the usual copulations of propositions, is, and is not, I meet with no proposition that is not connected with an ought, or an ought not’
A false dichotomy in educational research
Of course it is possible to show, through a control study or a scientific study illuminating ‘cognitive architecture’, that a given bit of the curriculum with a given assessment, can be taught more effectively in this way than in that way. That is useful research in a modest technical kind of way but it avoids the often much more pressing and obvious question that parents and governments expect educationalists to be able to help them with, which is the question about what should we teach and why. Research can show, for example, how best to memorise facts for an examination and why it is better to do it in stages rather than all at once (apparently – I have this only on the authority of Kirschner’s teaching materials), but research cannot show if memorising facts and reproducing them in examinations is what we should be focusing on in education as opposed to, for example, portfolios illustrating more embodied design skills. If you ignore the issue of values because you think it cannot be rationally grounded you become, by default, a purveyor of the ideologies of the past that somehow become sedimented into the curricula of today. As Biesta points out, education is values all the way through. Why, for example, do children have to spend so much time on trigonometry – knowledge no one ever seems to use in real life - and no time at all learning how to work with others in different countries in order to solve an international crisis in a mutually beneficial way? Is it really because trigonometry is more important? Who decides? A randomised controlled trial or experiment with cognitive architecture is not going to give us the answer to such questions.
On the other hand, some of the academics who are not frightened to tackle the big questions about what education is for, can be quite resistant to the idea that their ‘research’ is anything to do with science. I see a lot of research out there claiming to be educational research that others might think of as little more than political journalism. The authors start by asserting their values in favour of good things like ‘justice’, ‘anti-racism’, ‘post-colonialism’, ‘feminism’ etc and against bad things like ‘neo-liberalism’. Because they have been open in adopting an ethical ‘stance’ it is therefore somehow seen to be acceptable that the resultant research is obviously biased in such a way that no one who does not share those initial assumptions is likely to read it or be persuaded by its claims. The problem with this approach is that there cannot be any progress in our shared knowledge, just different groups of people confirming themselves in their own assumptions and prejudices. This taking a stance approach means that the values advocated do not get properly challenged and developed through argument and evidence.
These two types of educational researcher – the fact-focussed type that often gets labelled as ‘positivist’ by those who do not like this sort of thing as well as the more values focussed type that often gets called names like ‘post-modernist’ by those who do not like that sort of thing – are united in one key assumption: accepting Hume’s fact/value distinction. The one side asserts the importance of facts and thinks you cannot argue rationally using evidence about values so excludes them from science, the other side asserts the importance of values and agrees that these cannot be put on a proper research footing so exclude themselves from science. But what if the claim introduced so casually by Hume nearly 300 years ago is simply wrong? What if we can derive values from an investigation of the facts?
Of course Hume has a point if he is merely saying that we need to distinguish between accounts of what is the case backed by evidence and argument from serious researchers as opposed to stories about what ought to be the case invented by moralists who are biased by an obvious ulterior motive. But, although still common, that kind of reasoning is intellectually speaking, an extreme case or straw man argument. On close examination even the best empirical research is found to have a values base and depends on interpretation and argument within a larger dialogue. Similarly the best values-based theory can be persuasive through appeal to reason and evidence in much the same way as empirical science. The dialogue of empirical science and the dialogue of values are part of the same larger dialogue, which I want to call, following Oakeshott, the borderless dialogue of humanity. While the findings of this borderless dialogue as shared enquiry are often ambiguous and subject to possible regression there is still real progress over time in slowly expanding and deepening our shared understanding of ourselves and our situation as natural being within nature.
Values enter into research when we select what to look at, when we decide how to look at it and when we interpret the meaning of what we think we see. (Standish, 2001). So values are always implicit behind ‘experimental designs’. The large number of different research programmes being followed mean that frameworks of assumptions do get tested indirectly over time. On the whole, in the long run, those research programmes that better reflect the truth of things will survive (Lakatos 1978) and those that just reflect the passing intellectual fashion will fade away. When I say ‘the truth of things’ I include values as well as ‘facts’. There are many reasons for thinking that some values are objectively better than others and there is evidence that these gradually emerge over time. Yes of course all knowledge has to be socially constructed and mediated by language and culture and so on but that does not mean that there is no underlying truth of things that we can tend towards and get to know better albeit always ‘as if through a glass darkly’ (Bible, St Paul, Corinthians 13:12).
What I am suggesting here is that values as well as facts can be subject to systematic design-based research. This is not a very new idea. In the Kalama Sutta, the historical Buddha (around 2600 years ago) is reported as claiming that there is no need for faith in values, they do not come from revelation nor are they just subjective choices, we have to try them out for ourselves and then we will see which ones work to promote human flourishing, happiness and liberation and which ones tend to suffering, anxiety and death. Do not believe me or anyone else, the Buddha asserts, just try it and see. And that, of course, is what we do over time in educational research. We count not only the exam results but also the side effects, the employment patterns, the number of suicides and the relative amount of happiness and unhappiness, and we debate these findings and try to adjust education systems accordingly. Educational research is part of the borderless dialogue of humanity about how we should live and about what kind of future we want. In that borderless dialogue facts mix with values, and what ‘is’ blends seamlessly into what ‘ought’ to be.
In fact what we are looking at when we think we are looking at ‘data’ is always already informed (meaning shaped from within) by values. Face-to-face dialogue observed in classrooms is an excellent example of how this works. Dialogue is a naturally occurring event in the world producing ‘data’ for research but it is also always a social achievement. Dialogue is not possible without a certain effort at ‘mutual attunement’ between participants. This implies shared expectations that have the force of moral ideals. To engage in dialogue I need to be able to see myself to some extent as if from the other’s point of view. Seeing things as if from the other’s point of view is an ethical ideal but it is also a necessary ideal for successful communication and mutual understanding. It is true that we are often aware of unsuccessful communication but this awareness in itself implies that communication must be successful to some extent at least some of the time. In other words some ideals are also necessary social facts essential to the survival of communities (Habermas, 1994 see also Axelrod 2006).
It is widely held that while physical facts are universal, values are relative, changing with each culture. But empirical research suggest that there is one value that seems to run counter to this. This is called the ‘golden rule’ often phrased as ‘do as you would be done by’.According to those who have studied such things essentially the same ideal of reciprocity can be found in every cultural tradition. It seems likely that the ubiquity of the golden rule is a consequence of the ubiquity of face-to-face dialogue. In other words a fact found in every culture and essential to the continued existence of every culture, dialogue, gives rise to a universal human value, reciprocity. It seems that Hume was exaggerating a little in claiming that you cannot get ‘ought’ from ‘is’.
Constructivism has been the orthodoxy in my area of empirical educational research, educational technology design, for a long time now. Paul Kirschner is right to question this. He is right to claim that there is little support for constructivist approaches to teaching in any empirical evidence about how we learn. But such claims were not the only basis for constructivist teaching. On the whole constructivists have values, they tend to want children to be self-directed, to be able to learn by inquiring together with others and basically to be free thinking autonomous individuals. Many reasons and arguments are put forward for these values in education including the argument that we need lifelong learners in a time of rapid change and the argument that the jobs of the future require creative, flexible interdisciplinary thinkers due to the automation of normal jobs.
I have always been suspicious of constructivism. I tend to think that there are cultural traditions that we need to be inducted into before we can be usefully constructive. So I am not unhappy with Paul Kirschner’s account of the need for guided teaching before full participation in knowledge construction. This is bit like the distinction Neil Phillipson and I made in our recent book between learning the dialogue so far (guided teaching) and learning how to take the dialogue forward in the future (teaching dialogue, see ‘Dialogic Education: Mastery Learning through Thinking Together’).
Like the constructivists that Paul criticises I too sometimes feel that I move too fast from an empirically based theory of how we learn, learning through being drawn out to take the perspectives of others in a dialogue, to a theory of how we ought to teach – education for ‘dialogue as an end in itself’. I try to be reflective about my own bias. However my arguments for that shift from ‘is’ to ‘ought’ are quite varied and not just based on evidence from studies of learning. I also make claims about the cognitive affordances of new digital communications technology and the needs of education in the future. In other words the value of dialogic education is supported by arguments put forward with evidence in the borderless dialogue of humanity as well as arguments and evidence in the more bounded field of empirical educational research.
So how are these two contexts related? I mean how can we connect bounded inquiry into learning of the kind that Kirschner refers to and the much bigger unbounded shared inquiry into educational aims and values that I also think is important.
I think what we have here is a double-loop process rather akin to what Argyris called ‘double-loop’ learning, a concept widely used in professional education and organisational learning.
My suggestion from this example is that what appears to many researchers as an unbridgeable divide between facts and values within educational research is perhaps better understood as the real difference in quality and temporality of these two intertwined research loops. On the one hand focused empirical research within a theoretical framework that embeds values and on the other hand the rather larger and rather longer dialogue of theory that can question and develop the assumptions of research frames. Both loops can be seen as united in a larger conception of science as collective ‘wissenshaft’ or shared knowledge. Both call upon evidence from experience and both make arguments that seek to persuade. While research findings from the smaller loops of empirical science based on narrow frameworks of assumptions can seem to progress faster and be more certain for a while than the findings of the larger borderless transdisciplinary shared enquiry of humanity this is an illusion because in fact the cogency of the assumptions behind narrow research programmes depend upon justifications that can only be made within the larger dialogue.
When the larger dialogue moves on the apparent progress of narrow scientific programmes can become irrelevant. Consider why we no longer conduct much learning theory research on rats in mazes and pigeons in cages when once this was the norm. I suspect that this is not because such experiments stopped delivering good scientific results in the narrow loop sense but it is more likely because their relevance to human learning in social contexts became no longer obvious in the second loop or deep loop. Once a programme seems irrelevant it no longer attracts funding and new recruits. To attract funding and recruits it is necessary to be able to participate in the larger dialogue of humanity and not just talk to an in-group.
The larger dialogue is rational but in an expanded, open-ended, embodied, and creative way that includes emotions and values. Against those researchers who think that scientific method is a fixed way of delivering true facts I am arguing that scientific progress is only possible because scientific activity is always embedded within and part of a larger open-ended dialogue which includes values and the question of what we ought to teach.
Education depends upon values. Those values are not subjective or arbitrary. They arise from an ongoing long-term shared inquiry into our nature and our situation. That borderless dialogue depends upon arguments and evidence often drawn from more focussed empirical enquiries such as the exploration of cognitive architectures relevant to learning referred to by Paul Kirschner. However, these focussed empirical investigations ultimately depend on the larger dialogue around them for their cogency. These two loops of dialogue, the focussed dialogue of shared enquiry within a research programme and the long-term global dialogue about values, work together. Together they make up something we could call science in the broadest sense since the word science in origin simply means knowledge. But perhaps what we are really describing here - the slow and often fragile collective growth in insight and understanding of the human race collaborating together - could better be called education. This movement of science in the larger sense combines not only the education of individuals being inducted by teachers into ongoing long term dialogues of culture, but also the self-directed learning of humanity - an always ongoing borderless and transdisciplinary shared inquiry into the truth of things.
Argyris, C. (2000). Double‐Loop Learning. Wiley Encyclopedia of Management.
Biesta, G. J. (2015). Beyond learning: Democratic education for a human future. Routledge.
Habermas, J. (1995). Justification and application: Remarks on discourse ethics.
Hume, D. (1739/2003). A treatise of human nature. Courier Corporation.
Kirschner, P. A., Sweller, J., & Clark, R. E. (2006). Why minimal guidance during instruction does not work: An analysis of the failure of constructivist, discovery, problem-based, experiential, and inquiry-based teaching. Educational psychologist, 41(2), 75-86.
Lakatos, I. (1976). Falsification and the methodology of scientific research programmes. In Can Theories be Refuted? (pp. 205-259). Springer Netherland
Oakeshott, M. (1959). The Voice Of Poetry In The Conversation...
Phillipson, N., & Wegerif, R. (2016). Dialogic Education: Mastering core concepts through thinking together. Routledge.
Standish, P. (2001). Data return: The sense of the given in educational research. Journal of philosophy of education, 35(3), 497-518.
Wegerif, R. (2008). Dialogic or dialectic? The significance of ontological assumptions in research on educational dialogue. British Educational Research Journal, 34(3), 347-361.
Wegerif, R. (2013). Dialogic: Education for the Internet age. Routledge
Wegerif, R., Postlethwaite, K., Skinner, N., Mansour, N., Morgan, A., & Hetherington, L. (2013). Dialogic science education for diversity. In Science Education for Diversity (pp. 3-22). Springer Netherlands.