Critical thinking and domain specificity

As a consequence of a pleasant Twitter thread over the weekend, I am currently revisiting two papers that have heavily influenced how I think about the idea of ‘critical thinking’. The two papers, by the same authors, are:

• Bailin, S., Case, R., Coombs, J.R. & Daniels, L.B., ‘Common misconceptions of critical thinking’, Journal of Curriculum Studies, 31.3, (1999a), pp.269-284
• Bailin, S., Case, R., Coombs, J.R. & Daniels, L.B., ‘Conceptualising critical thinking’, Journal of Curriculum Studies, 31.3, (1999b), pp.285-302

The sharp-eyed amongst you will have noticed that these two papers appeared together in the same issue of the Journal of Curriculum Studies a good twenty years ago. They were both set as reading on my PGCE, and have shaped my views on the matter ever since. I am reliably informed that both can be accessed with Charted College of Teaching Membership.

What I particularly like about the two papers (particularly the first) is that they set out to do some serious conceptual analysis by cutting through the myriad of terms associated with ‘critical thinking’. I am a big fan of such approaches (and it is thus no surprise that I rate R.S. Peters very highly) for education suffers from a proliferation of terms. We seem to pride ourselves on inventing new terms, or modifying existing terms with adjectives to mean something different, or using an existing term but giving it a different meaning. This proliferation, which might superficially be a sign of a vibrant field, is I think the sign of a struggling one, for it makes it difficult to create common shared ideas that work as glue to hold a community of thought together. Parsimony is our friend here, and there is not enough work going on to reduce rather than increase the terminology used in our field. The two papers I am promoting here do however do just that.

I shall make an attempt here to summarise the arguments, although I would strongly recommend reading the originals if you can as I shall no doubt fail to convey fully the complexity of the argument offered by the papers. The first paper sets out to deconstruct commonly-held assumptions about ‘critical thinking’. In particular, the authors target the idea that ‘critical thinking’ is some kind of generic skill, challenging the notion that “generic operations that can be learned in themselves, apart from any particular knowledge domains, and then transferred to or applied in different contexts (1999a, p.271)”. Instead, they argue that “The kinds of acts, such as predicting and interpreting, which are put forth as generic skills will, in fact, vary greatly depending on the context, and this difference is connected with the different kinds of knowledge and understanding necessary for successful completion of the particular task (1999a, pp.271-2).” This is now a well-rehearsed argument in education for it has been popularised by an recent emphasis on cognitive psychology which has tended to conclude that expertise is not generic but context-specific.

The idea that ‘critical thinking’ is a mental process also comes under the spotlight in the first paper. Here, the concern is that we cannot observe mental processes as we can observe physical processes, and that we have no grounds for assuming that the process involved in one form of (e.g.) ‘evaluation’ or ‘abstraction’ is necessarily the same as that involved in another. The authors note that this is an enticing idea as, if we can have thinking processes that transfer easily from one domain to another, then we can all save ourselves a lot of time as teachers by simply focusing on teaching the generic process. They point out, however, that as attractive as this is, we do not actually have any grounds on which to conclude that this is in fact the case.

The third conception of “critical thinking” taken to task in the first paper concerns the idea that it is a set of procedures to be followed (such as hypothesis testing, or weighing up pros and cons). The issue raised here is that what matters is not so much the procedure, but what it means to do that procedure well. As the authors note, using the ‘pros and cons’ model as an example,

The pro and con reasons that the individual comes up with may address only the most trivial aspects of the issue; so, too, the brainstorming of alternatives may miss the most sensible alternatives. Learning to engage in such activities has little educational merit unless these things are done in such a way as to fulfil relevant standards of adequacy. Students have, after all, performed these sorts of tasks for much of their lives. (1999a, pp.278-9)

The concern, then, is that the procedure is less important than knowing when a procedure has been done well. The challenge here is that the norms concerning what it means to ‘do something well’ are very frequently related to the specific nature of the problem being addressed. As the authors put it, “what drives increased competence in thinking is greater mastery of the standards for judging an appropriate tack to take in a particular context, not learning pre-programmed, supposedly generalizable, procedures (1999a, p.279).”

This emphasis, on what it means to ‘think well’, becomes central to the argument advanced in the second of the two papers. Whereas the first paper takes to task existing conceptions of critical thinking (skill, process and procedure), the second paper advances an alternative model for what it might mean to think critically.

So, where do they go? The thrust of the argument in the second paper is to reconceptualise ‘critical thinking’ not as some skill, process or procedure, but rather as a measure of quality. “It is the quality of the thinking, not the processes of thinking, which distinguishes critical from uncritical thinking (1999b, p.288).” To think critically is to think well.

The authors then set out a list of five conditions of ‘thinking well’. These are:

• Background knowledge – in short, a person’s ability to think well about a problem is very heavily determined by how much he or she knows or can find out about that problem.
• Operational knowledge of the standards of good thinking – the emphasis here is on understanding the practices of which one is part and what its standards are.
• Knowledge of key critical concepts – it is here I think that the authors come closest to what is often taken to be ‘critical thinking in a curricular sense. They offer examples such as “concepts for distinguishing metaphorical and literal language; necessary and sufficient conditions; assumptions, presuppositions and implications of an argument; and aesthetic, moral and prudential judgements (1999b, p.293).”
• Heuristics – here the authors come close to an internal contradiction, having challenged the notion of heuristics in the first paper, but they do specify that “The most powerful heuristics tend to be those designed to guide persons in carrying out rather specific kinds of tasks. Procedures designed to apply in all cases of critical thinking, such as the ‘problem-solving procedure’, are likely to give little help in solving any particular problem (1999b, p.293).”
• Habits of mind – by this the authors are referring to certain predispositions or values that a person has, and which might include “respect for reason and truth” and “open-mindedness”.

The second paper finishes by considering the curricular implications of the conditions of ‘thinking well’ that the authors provide. Although the authors do recognise some general ideas, the overall conclusion is that

“teaching critical thinking is best conceptualized not as a matter of teaching isolated abilities and dispositions, but rather as furthering the initiation of students into complex critical practices that embody value-commitments and require the sensitive use of a variety of intellectual resources in the exercise of good judgement (1999b, p.299).”

 Little account in the article is given of what kinds of ‘critical practices’ these might be, although “history, art, music, science, mathematics, technological studies and vocational studies” are mentioned (1999b, p.299). The authors do however hedge their bets right at the end and note that, because all possible forms of practice cannot be captured in the school curriculum (e.g. buying a car), it might be necessary to have ‘special’ courses in critical thinking. The authors sign off before dealing with how that might contradict their own conclusion that “we cannot safely assume that learning good judgement in applying a principle or standard in one context will generalize into good judgement in applying the principle or standard in other contexts (1999b, p.299).”

Those who have read my blog regularly will immediately see the influence of these two papers on my own educational thought. The emphasis on thinking well within practices is crucial and, although the authors do not dwell much on the psychological literature, it would in general seem to support their conclusion. This is why my hackles rise every time someone talks of “applying critical thinking to history” or “using critical thinking techniques in science”. If we follow Bailin et al’s argument, then to do history well or to do science well is to think critically: both practices already contain within them normative standards for what good thinking within those disciplines looks like, and it is to these ends that we ought to be directing our pupils. On such a conclusion, if our concern is to help pupils become ‘critical thinkers’, then probably the best thing we can do is to teach them history, science, mathematics, and so on. I agree with the author’s point that this also means teaching technological and vocational studies well, but there are wider issues here that I have addressed elsewhere concerning whether schools are the right places to be achieving this.

What, then, of the need for lessons specifically in ‘critical thinking’. If nothing else, the argument advanced in Bailin et al should be for us to exercise a good degree of caution. Although the authors hedge their bets on this right at the end of their second paper, much of their argument would seem to undermine the case for general lessons in critical thinking. I think this tension can be resolved by identifying those very high-level domains which might be worth learning. Good examples of such domains include logic, statistics, linguistics and (most obviously) philosophy. Each is a domain with its own clear standards of what it means to think well, and yet each, by its nature, finds itself drawn on in a great many contexts.

The role of these high-level domains is something I shall return to in a future post.