Andy Gibbons [4 Apr 96] sent a long message to ITForum. In my view (not his) it had four separate components:
(1) He didn't like the phrase "learner centered" as a characterization of new-wave instructional approaches. I have my own objections, expressed in an earlier message, but they are different from his.
(2) He thinks experience is primary, and that verbal concepts are derivative. I think that this is deeply mistaken, as I shall try to justify below.
(3) He sees new wave instructional techniques as being a shift back away from a controlling teacher to a companion. I think this is a very interesting view. For instance it is compatible, as many theories and practices are not, with the extensive evidence that peer interaction promotes deep learning, yet does NOT depend on the peer knowing more in advance; in fact people can learn by interacting with less knowledgeable people. I don't have much to say about it, although I shall go on thinking about it. I think actually it is compatible with my quite different views on (2): in other words, I think it may be an enduring idea that does not depend on the particular justification for it that he gave.
(4) He thinks the shift is also from static, discipline-related content structures, to dynamic performance-related model centered structures and instruction. I also don't have much to say on this, though again it is clearly important.
Andy Gibbons made clear that he thinks experience primary, verbal concepts secondary. (e.g., "This learning-by-poking or learning-by-experiment is the natural pattern of learning. It is the method by which the overwhelming majority of our personal knowledge originates, even taking into account the mass of knowledge we acquire through formal studies.") I'll give four arguments for a different view: (a) Putnam's arguments that even everyday concepts like "gold" and "water" are not grounded in our personal experience but are socially distributed: we mean what others mean; (b) the nature of science as a socially distributed enterprise; (c) the example of buying a house; and (d) Laurillard's model of the teaching and learning process, and in particular its distinction between the level of public, formal descriptions and concepts, and the level of personal experience and action. The idea here is that both are important, as are learning the relationships between them: not that the latter is primary. In fact if I could only have one, I would go for the former.
(a) Putnam's argument that there is a sense in which we don't understand most of what we say: much of our knowledge is really socially distributed and not in our personal grasp. The philosopher Hilary Putnam argued in an essay that if you really ask yourself what you mean by words like "water" and "gold," then although we do know quite a lot about them, when faced with stuff that looks just like what we expect and yet might be simulated stuff (e.g., platinum or a cunning alloy for gold, or "heavy water," i.e., deuterium oxide, for water) we end up by saying that we mean whatever the ultimate expert means by those terms, e.g., the official government chemist. Putnam H. (1975) "The Meaning of Meaning" in Mind, Language and Reality (Cambridge, U.K.: Cambridge University Press).
(b) Science is a socially distributed process, in line with the above. As the over quoted remark of Newton about standing on the shoulder of giants indicates, the nature of science is that each individual adds a little bit to the shared accumulation. A scientist does not even check, much less reinvent, previous discoveries. There is debate about whether each new fact should be tested once or thrice: but certainly not by most scientists. Ideas of evidence and so on can be viewed as limiting science to what can be shared socially by having a checking procedure. But the whole point of science is to discover things once, then everyone can exploit it. This means that the body of science is transmitted, not by personal experience, but by other means: typically the printed word. The rest of our culture is rather similar. What is so distinctive about humans is that we so successfully share the results of experience without having to experience it ourselves. And as Putnam argues, we typically do not learn all that others know, but leave the knowledge socially distributed. When education is reduced to Gibbons' "lexical loop," it may have fallen below the ideal, but it is still empowering the learner with the keys (the index terms) to this socially distributed network of knowledge that is the essential heart of human power. This, for me, is an important point. If we teach students the official names for things, then they can participate in this socially distributed knowledge even if they don't understand much about it, e.g., they can look it up in the library, ask experts about it, etc. So much of the use of knowledge comes from knowing the public names for things, and not from complete personal understanding. (Though having both would be even better.)
(c) Consider the example of buying a house: one of the most important activities we undertake. If learning-by-experiment were the paradigm, I could only buy at random, and would not learn much in a lifetime. What I do, like most others, however, is to learn through words from others: this affects almost everything about this task. For instance, I have never had a fire in my home, nor witnessed one close to: instead I read about this (to me) theoretical risk and act accordingly. Similarly I hope to learn from other people's experience of burglary without having to try it for myself. Similarly for the beds, clocks, cookers, etc. that I will buy. In other words, Putnam's and not Gibbons's description seems to me to apply to this non-academic activity, and to the mode of learning that then determines my actions.
(d) In an earlier message to ITForum [Draper, 28 Apr 96], I described Laurillard's distinction, from her 1993 book Rethinking University Teaching, between the level of public, formal descriptions and concepts, and the level of personal experience and action. See that message for examples and a longer discussion. The former level corresponds to Putnam's point, while the latter corresponds to how Gibbons describes his viewpoint. The point is that all academic subjects have components at both levels; the ideal teaching and learning process will cover both, and in addition the links between them. Instruction most often emphasizes the level of public descriptions. Redressing the balance may be in order. But to swing to the opposite one-sided view is, in my view, deeply mistaken for the reasons I have given.
Finally: Gibbons gives this interesting view about replacing a controlling teacher by a "companion." The Laurillard model does not cover this, and it might perhaps with advantage be extended to incorporate Gibbons' idea. However it does emphasize a conversational model, with iterative interaction between teacher and learner at both levels (public and private, descriptive and material action). Gibbon's phrase of "learning by poking" is a close match to the Laurillard level of personal experience and action on tasks. But an analogous interaction at the conceptual level is also in that model, where the learner gets feedback on their attempts to express and re-express the concepts and descriptions being learned at that level. So I think progress will be served by applying Gibbons' approach to the full Laurillard model.