The ideology of child-centredness and discovery learning has won many converts among science teachers, or at least among curriculum planners, witness the enthusiastic reception of Nuffield Science in Great Britain or the activities of the Federation for Unified Science Education in the USA. It appears now to be on the ascendant in the Caribbean, as several proposed CXC courses reveal, or as was obvious from many contributions to the 1979 conference of the Association of Science Teachers of Jamaica in Kingston. But as often happens when one fashion displaces another, its own excesses and blind spots begin to appear and become the focus for reaction or despair. In particular, it is becoming more widely thought that discovery based science teaching does not, as was hoped, initiate pupils into the procedures of real science but rather encourages a fraudulent game of guessing the 'right' answer. On the bad old traditional system, everybody was clearly aware that the teacher had the right answers; on the new, pupils are supposed to be discovering answers for themselves but are in fact quite aware that they have got to come up with the one that the teaching is keeping mum about (cf. Driver [1]). More recently, Mathias [2] has used this, along with other considerations, to castigate all school science, traditional and modern, for failing to be part of what it purports to be teaching.
It is perhaps worth noting that the same criticism might seem fatal to another venerable suggestion for overcoming the sterile dogmatism of traditionalist science teachingthe idea that science should be approached via its history. As this history is more deeply studied it is becoming ever clearer that a serious grasp of the history of science requires much more expertise (in sociology, history, logic, to begin the list) and knowledge (of long forgotten theories and observations, to note but one kind of item) than most science teachers could fairly claim. So science through its history would almost certainly not be science, or history, but rather a fictitious sketch of what never happened, and one that might find it hard not to give the impression that our predecessors were pretty stupid. It might be fun, but I suspect novels are more rewarding. (The history, or perhaps even an unhistorical rational reconstruction of science is, however, of considerable use in tackling questions in philosophy of science and methodology, and so, given my stress on the central role of methodology, may play a significant part in curricula designed to implement my proposal. But not as history, rather as a simplified case with historical analogies. Since school science hardly ever catches up with the theoretical advance of real science there is a sense in which it is mainly such a dabbling in simplified history.)
Before proceeding to the specific case of science, it is worth looking at Mathias' objection in general. He contrasts school science with real science and condemns the former because it is not part of the latter. School science is, of course, not intended to be a separate branch of knowledge; it is intended to be the teaching of science, real science; it is intended, to borrow Professor Peters's expression, to be the initiation into science. Now is it wrong, in general, that the procedures of initiation into X are not themselves part of X? In general, it would seem that it is not. Surely no one would object to simulated flying on the grounds that it is not real flying and so should not be used in teaching people to fly. Nor does throwing people in the deep end seem to be the most efficient way of teaching them to swim. For all I know, it might indeed have been one of Peters's motives for lighting upon the term 'initiation' to characterize education that initiation rites are often very unlike the activities they give on accesscopulation is thankfully not like circumcisionwho then should suppose that school history or school science or school French must have any discernible link with real history, science, or French?
But, of course, many of us do, and with reason, for there are many activities which can only, or at least can most efficiently, be learnt by participating in them. I have at least six books that can tell me how to cook a soufflé, but I know I shall never learn until I start trying actually to cook one. Perhaps one day we will have squandered enough resources to be able to do computer simulations of a soufflé, but as things stand, in this case and many others, the best way to learn to do X is to do X.
We may conclude, then, unhelpfully, that we must take each case on its merits when it is objected that the school subject, the initiation is too far from the real thing.
Part of the problem presented by Driver, and by proponents of discovery methods in science teaching, is to reach agreement on what the real thing is, in the case of science. Driver, for instance, contrasts science as discovery with science as a body of accumulated knowledge, or belief if you wish to be cautious. If we stress the former, as Driver and others do, it is then clear that there is something fraudulent bout getting pupils to discover for themselves correct answers known in advanceit is neither genuine discovery (since the teacher at least already knows what is to be 'discovered') nor is it simulation of the original discovery (since the intellectual situation of the discoverer was almost certainly different from that of anyone in the classroom). And even if there was not, it would seem perverse to expect pupils to recapitulate in a few years the discoveries of many generations; if we have made some progress, why hold pupils back from enjoying the fruits of it? (I do not find it possible to take seriously those who seriously deny that we have made progress, who aver that there has only been a shift of fashionable paradigm.)
Keeping our eyes fixed on creative discoveries, we are next assailed by the discouraging truth that it is impossible to teach people to be discoverers of this sort. It requires but little reflection on matters of simple logic to see the unavoidability of the Popperian emphasis on guess work, on conjecture, in the formulation of explanations or theories. The absence of any effective rules for constructing worthwhile theory is even clearer if we turn away from commonsense to simple chemistry or physicswe explain the behaviour of a gas by talking about the behaviour of millions of molecules; we account for the distribution of some features of animals by reference to the combination of genes in fertilization; and so on and on. In none of these cases is there anything in the data to be explained that will lead us, by any statable rule, to the satisfying theoretical explanation. No doubt there are useful tips of the trade to be had here, and there are ways of applying existing theory to new situations that can and must be taught (as importantly stressed by Kuhn [3]); but in short, and uncharitably, 'hack' work in science can be taught, just as it can in painting, rhyming, history, or cookery, but we cannot thereby teach people to be another Cézanne, Guérard or Galileo. If we cannot teach this, it would seem more honest not to pretend to.
With this last aim in mind, the aim that is may salvage from the wreck of discovery oriented science teaching, a re-examination of what goes on in our classrooms is an urgent necessity. To take but two examples from casual observation of science lessons, I have seen competent and motivated teachers set up a sequence of experiments intended to support a conclusion without any controls; a teacher asking a class what they may deduce about the behaviour of proteins from the observed behaviour of one protein. And this is to omit the type of experiment Mathias mentions that reflects such hopeless confusion that it could never be saved. Of course, excuses can be offeredin ordinary life we often get by without controls, because we have a simple before/after situation in which the persisting before set-up can be regarded as the control; in ordinary language we use 'deduce' more loosely than would a logician, so students are not really misled by being asked to make impossible deductions. But should we accept these excuses, if our concern is to get pupils to appreciate rigorous critical testing or to appreciate the tentative nature of generalizations. Science is precisely not like our everyday muddling along; and it works incisively only when its practitioners are clear about what can be deduced from what, and what can only be offered as a provisional explanation. And, finally, as Driver notes, there is some evidence to suggest that pupils are indeed misled by the incoherence of their teachers.
My suggestion is, then, that the realistic part of the new emphasis on discovery avoids the strictures we have looked at; we should abandon pretensions to teach creative discovery but keep the emphasis on tentativeness by focussing on testing, by teaching rigorous methodology. This is something worth grasping in its own right, and is indeed an essential part of science in its full bloom. It is moreover something general, resting as it does on logical foundations. It should, then, present fewer intrinsic difficulties to generalization beyond the sciences to the examination of the claims of politicians, journalists, preachers, or the pupils themselves.
This last point suggests, what is moreover true, that there is nothing especially scientific about my proposal. Canons of critical testing of hypotheses might be more easily taught by playing at detecting a crime; problems in sampling might show up more clearly in a mini-social survey than in a biology lesson. Since the bases of methodology are logical, this is only to be expected; and it seems to me an advantageous feature, particularly in a context where resources for straight science are scarce or non-existent.
But while we do not need to rely on scientific examples all the time, is there any reason to suppose that my proposal would hinder the teaching of the currently accepted facts in any branch of science? Not as far as I can see. Rather it should soon become obvious that all observation and testing rely on background assumptions of just such accepted facts. Thus, to take but one example (of the sort Cohen has elaborately examined [4]), if you want to test the hypothesis that all birds of a certain species are brown, and if you have some background knowledge that there is pressure towards white plumage in arctic regions, you will be better employed finding specimens of your birds from the Arctic than in collecting more of the temperate examples that might have suggested the hypothesis in the first place. There is then no reason against, and every reason for, a teacher conveying enough of our background knowledge to make an investigation or activity fruitfulit would have saved some wasted time if one teacher had warned his class on floral diagrams not to collect compositae: a trivial example of a tremendous failing. It is perhaps unnecessary to add that in recommending attention to scientific rationality I am not proposing that nothing else be done in the time devoted to science; all I am saying is that part of the time and effort be directed to this end, that it should be taken seriously and not left in the slip-shod, messy state we can too easily see around us.
It is, however, necessary to avert one other misconceptionI am not proposing that pupils be taught the logic of scientific investigation as an explicit, second-order subject of the type Mrs Warnock [5] fears will enter the curriculum on the heels of Hirst. Flew [6] reminds us of John Locke's down-to-earth assertion: 'God has not been so sparing to men to make them barely two-legged creatures, and left it to Aristotle to make them rational'. And scientists have not had to wait upon the still unfinished work of confirmation theorists. But all the same, one way of getting people to see the why and wherefore of testing, sampling, and so on is to engage in the abstract study of argument; and I would suggest that, for the teachers at least, such an explicit study would be of value, so that they can explain and justify the techniques their students will be learning to use, so that they can see why you can't deduce universal generalizations from observations, why controls are necessary, why statistical procedures work as they do, and so on. I do not want a proliferation of methodology or even philosophy of science courses; all I want is that pupils, and their teachers, understand and can apply (supposing there is any difference between these) techniques of scientific investigation and argument of the degree of rigour found in real science. This will not guarantee Einsteins or Galileos; in the uncharitable terms I use earlier, it is 'hack' work, but vitally necessary 'hack' work, and also potentially liberatingthe practice of freedom without Freire's continental trappings?
URL http://www.uwichill.edu.bb/bnccde/epb/logicinlab.html