How to teach critical thinking: an immodest idea

This entry is going to proceed through some maddeningly vague statements, so I will start with the bottom line: currently my best idea for teaching critical thinking in social sciences or related areas is to have students wrestle with an important substantive question or puzzle in a reasonably-well-bounded area of knowledge. The structure may be the writing of a history essay such as those the Concord Review publishes, an International Baccalaureate extended essay, a seminar class with a central question, or any other experience that has the “substantive wrestling” part. That forces students to develop skills in the context of learning or using background knowledge in an area, and some part of that limited but important set of skills will then transfer to some nearby areas of knowledge. And that is enough to justify the exercise.

The term critical thinking means nothing in itself while able to hold a broad range of meanings to those who use the term. We should teach critical thinking, or so many say, but what does that mean? Usually that call does not mean a mandate for a course in formal logic, though maybe we should require a course in formal logic for all who call for teaching critical thinking. Often it means encouraging people to be moderately skeptical about the claims they read or hear, something that shouldn’t really need much encouragement for anyone over the age of 9. Sometimes it means teaching one of the vague algorithm-like processes tossed around in connection with other vague terms such as problem-solving.1 Sometimes it is the modern equivalent of faculty psychology and the 19th century argument that learning a particular subject would strengthen some form of mental muscles.

In the 20th century, the most interesting argument in favor of teaching some form of critical thinking was the idea that we should teach disciplinary perspectives when we teach subjects. I associate Jerome Bruner with that argument, and his optimism about the ability to teach how to think like a physicist, an historian, and so forth. In 1959, Bruner was a key player in the Woods Hole Conference discussing American education after Sputnik. In The Process of Education, Bruner made a broad argument that young children had the capacity to understand disciplinary perspectives and the underlying structure of a field. As one of the key figures in postwar cognitive psychology, Bruner’s influence was widespread.

Bruner turned out to be both correct and incorrect, as often happens when people make bold arguments. He was correct that you can and should teach young children more than rote facts. But when educators took his argument to mean that we should teach more abstract ideas such as set theory in math, it turned out that “disciplinary structure” was not the only consideration in selecting what and how we should teach children.2 While I am bemused by and skeptical of E.D. Hirsch’s prescriptions, he is almost certainly correct that you need to learn stuff and not just abstract how to learn something we’re not teaching you at the moment.

But just learning stuff is not enough; somehow those who focus on the importance of prior knowledge often talk as if a miracle happens in cognition if you just keep learning more stuff in a subject.3 One difficulty with intermediate and middle grades is that while we know how to teach decoding and reading comprehension, addition and subtraction and multiplication and division, and we can buy textbooks that say they are “Common Core-aligned” or approved by such-and-such textbook commission, we often do not really know how to help students make significant jumps towards the type of disciplinary understandings that Bruner would like children to make, and I think most of us would agree should be significant goals within any subject. In the teaching of history, for example, there is a great deal of literature on different purposes of teaching history in K-12, and also of the type of cognitive work that historians do, but there is little great research I am aware of discussing how to teach putting primary sources in context, sorting through different perspectives on a period, and in short performing the cognitive work common in history and that belongs in an historian’s critical thinking repertoire, for want of a better term. We know many students can make that leap–many somehow figure out how to write an essay in response to a DBQ (document-based question) in Advanced Placement classes in high school. But I suspect many of them make the leap independent of planful instruction. Somehow exposure works for many students, but not all and not most.4 Many undergraduate students enter my classes with fairly sophisticated understandings of history as a discipline, others do not, and I have not discovered the magical elixir or quantum of stuff that leads most students in the second group into the first during a semester.

So a belief in young children’s ability to think like an historian or mathematician without knowledge in history or math is magical thinking, and likewise any belief that the right tactic is to throw enough stuff inside children’s heads with a thoughtful adult nearby. Is the solution to teach some generic version of “critical thinking” or meta-cognition (the ability to recognize what type of tactic might be appropriate for a specific assignment)? There is quite a bit of research on meta-cognition, but the last time I needed to be aware of a large chunk of it (back in the mid-90s), my strong impression was that the potential benefits of meta-cognitive strategies were subject-specific (and strategy-specific).

All of this suggests that we don’t currently have the pedagogical content knowledge chops to where we could say, yes, we know how to teach what counts as critical thinking in history, sociology, anthropology, economics, etc.5 For some other subjects, there is quite a bit of money to learn what helps children make significant leaps in conceptual understanding. Not a huge amount of money–educational R&D is a pittance compared with other fields–but there is at least a reasonable stream of money in math education methods. Not so history, or other social-science fields. Nor should we expect significant money for that in the near-term future.6

That depressing gap does not mean we are entirely bereft of tools. Some of those mentioned above can be useful for engaging students in particular spots — using primary sources in assignments, having class debates, or role-playing (whether a one-class activity or the long-style assignments in Reacting to the Past games). The main problem is not the creativity of K-12 and college teachers but the scale and scope of the tools we commonly use. What in addition to sufficient knowledge in a subject helps students to start thinking like someone formally trained in the discipline? Please don’t tell me, “two or three courses”–that’s just the enough-stuff claim all over again.

There are some options for thinking more strategically about courses. Yesterday Swarthmore history professor Timothy Burke explained his redesign of an honors class in African history, which prompted me to revisit a 2008 blog entry I had written about discursive learning in liberal-arts classes and also think about the best undergraduate class I had in helping me to think like an historian. Burke explains the goal of his redesigned class as follows:

It’s still going to be about the historiography of modern Africa but I am going to build the class much more extensively around learning to dissect, interpret and operate large bodies of citation, information and reference when the aim is primarily to understand the state of scholarly and intellectual conversation about a subject rather than to produce a work of research on a specialized topic. I want to show my students how to become active agents in parsing and reassembling historiography.

Historiography is simply historical literature, so this is essentially a course in understanding what makes a research field tick in history. To some, this may look like the type of navel-gazing that only a history major would take, and only then if required for majors. But it’s not. I don’t mean that Burke’s course is guaranteed to draw all sorts of people who are motivated to debate why historians take different approaches in African history, but that this type of class has two important features. One is the invitation to explore both the substance and the politics of a field; I took a few courses of that nature at my liberal-arts college, one of which led to a much deeper understanding of W.E.B. DuBois than ever would have been the case without that course. That particular course taught by Paul Jefferson had a 30-page paper on an African-American historian as main requirement, so the investment of time in DuBois in particular as a choice of mine. That functioned essentially as a higher level version of the enough-stuff thesis: “throw enough stuff at a history major, or better yet invite the history major to throw enough stuff and herself or himself.” And that worked. From my experience in a course with overlapping goals, I think Burke will draw a number of interested students.

But beyond the substantive invitation, there is another feature of Burke’s course that is important, and that feature is setting up the historical debate over African history as a focal problem for the course. I had one course like that as an undergraduate, and unlike the class where I chose the topic for a long paper, in the other case the professor identified the focus for the entire class that lasted as the central question week-in and week-out. This was a course in early modern European history taught by Susan Mosher Stuard, and she framed the course with a single question: “Why the industrial rise of Europe, and at that time?” Almost twenty-eight years later, I still remember the first class where she gave an example of alternative outcomes (why didn’t extensive silk production in the late medieval period serve as a trigger for industrialization?) and previewed some of the claims we’d be reading over the semester. Then every week we’d read a sample of the historical literature and Stuard would prod us to discuss the general argument of the reading for that week and how the latest reading helped us understand the broader question. Over the course we built up quite a bit of detailed knowledge of early modern Europe, a good catalog of the historical arguments, and experience in debating those arguments.

There were lots of wonderful qualities to Susan Stuard’s teaching, and her skill in organizing a course around a compelling central question was one of them. She did not need to work that hard to sell the question to a group of liberal-arts undergraduates, mostly history majors, but she worked quite hard behind the scenes to keep us on target and productive. Among the moments I remember is one evening when she told us that it was much easier to teach our generation about the concept of societies in abstract than previous generations, all because of Star Trek. She continued, explaining that every time the Enterprise visited a planet, the episode explained something about the society, including the politics and often a sketch of the planet’s economics, religion, and culture. There is a great deal lost in the words I have just written; Stuard said it with such warmth and enthusiasm that it was much more affirming than the description above… and made it clear and expected that we as students should be thinking even more about the broad structures of societies.7 That course was enormously successful in teaching a good part of what counts as generalizable “critical thinking” in history: assimilation of both historical details and arguments, skill in ranging between detail and broad generalizations, understanding the way historical debates unfold and the assumptions about human nature and societies embedded in major historical arguments.

There is a description of this type of course: Robert Rotenberg defines a seminar as a class with a central problem or question. That is more narrow than the common use of seminar, and it is broader than the usual discussion of case-based or problem-based learning. In Rotenberg’s problem-based seminar, the entire term revolves around the question, a structure that should encourage students to engage with the subject at multiple levels. Instead of assuming that a student picking a paper topic will explore it at length, the course itself has that focus as an explicit plan. The last time I taught such as class, in fall 2012, I framed the graduate history of childhood course around the question, Is childhood for children? Several students told me months later that they kept thinking about that question as they drove home from class. So at least from the standpoint of student engagement, it worked from my perspective as a teacher. I KNOW that it served that role when I experienced it as a student.

That specific type of class probably is not required for a student to think like an historian, sociologist, economist, or political scientist. I think any long engagement with a substantive question is the closest we can get to the experience required to develop what counts as critical thinking for a discipline or subject, at least now. And as I explained above, that engagement can come in many forms. Susan Stuard constructed a class to encourage it. Paul Jefferson pushed individual students to do so independently in his course, Afro-American Historiography. And there are examples of those structures in high schools, moreso in the independent-project realm: the journal of high school history essays, The Concord Review, and the junior-year long paper called the extended essay in International Baccalaureate programs. And there are many others. In lieu of better-funded R&D in teaching social sciences in K-12 and undergraduate classes, that’s probably a reasonable rule of thumb: not just an assemblage of stuff but a focus on certain well-bounded stuff that is a student’s regular wrestling partner.

If you enjoyed this post, please consider subscribing to the RSS feed to have future articles delivered to your feed reader.


  1. There is nothing wrong with problem-solving, but because it is not recognized as a technical term, abstract problem-solving processes lose power in comparison with more specific examples in contexts. []
  2. I learned set theory in first grade, and so was one of the millions of children taught “new math.” It didn’t hurt me to be exposed to it, but I don’t think I could have grasped the concept of compact metric spaces at age 6. []
  3. Extra-credit for wonks on the history of educational psychology: compare and contrast with Herbart’s concept of the apperceptive mass. []
  4. You can replace DBQ with other ways of spending time in a history class: role-playing, debates, etc. When well-constructed, they’re all great for motivation if nothing else. But I am not persuaded that any one is necessarily more effective ways to learn to think like an historian than others. []
  5. For the purposes of this discussion, I am thinking of history as a social science. To put it in the humanities would not change this argument much. []
  6. The first IES grant focused on history was issued just this year for “a tablet-based interactive role-playing game that immerses 5th through 9th grade students in the history of the Great Depression.” Mmm… sounds like a great project, and it’s clearly not about pedagogical content knowledge. []
  7. In an essay I came across some time ago, Stuard explains that the French Annales school had a significant impact on feminist history by ignoring mostly-male political events and making clear that ordinary human beings were important. []

7 responses to “How to teach critical thinking: an immodest idea”

  1. Gary McGath

    It’s a very good point that critical thinking can’t be taught purely in the abstract, that it has to be connected to specific questions. But I think you go too far in the direction away from recognizing a core set of skills in what you call meta-cognition (which are really simply basic cognitive skills).

    These skills include the ability to recognize logical fallacies, emotional manipulation, incomplete information, and one’s own tendencies to favor one view over another. Thinking habits that incorporate these skills are applicable to any body of knowledge, including newly acquired ones.

    Things get tricky, though, when actually teaching critical thinking. The examples teachers use will usually involve fallacies in support of positions the teachers disagree with. This is natural but tends to undercut the point of the teaching. Teaching it right requires exposing sloppy thinking by one’s allies, and most people don’t like to do that.

  2. EB

    Wonderful post. I agree that the Core Knowledge sequence is lacking in attention to how students move to critical thinking. They do emphasize connections, but maybe not enough questions. (or: CK is necessary but not sufficient). At the K-12 level, too many teachers “teach the controversy” without giving students any grounding in the facts OR the skills that would enable students to use real disciplinary techniques.

    One thing that this post makes clear is that college-level classes (and especially seminars) can engage with disciplinary techniques that K-12 (and especially K-8) classes should not introduce prematurely. Yes, teachers can facilitate discussions where students draw connections and compare approaches in social science areas (and the other disciplines too). But it is dispiriting to see elementary-aged students being led through lessons designed to teach them to think like physicists or historians or mathematicians almost to the exclusion of allowing them to engage with the actual knowledge that would fascinate them.

  3. EC

    “we know how to teach decoding and reading comprehension”

    Hm. I’m not so sure about that… I think your suggested approach–long-term wrestling with a substantive question–would do much more for students’ reading comprehension than what a lot of teachers think they “know.”

    1. CCPhysicist

      Looking only at the reading comprehension of students in a community college science class, they can decode the important new words but often lack the background knowledge to comprehend the explanation of a definition based on the knowledge of more basic terms that is assumed by even the best textbook authors.

  4. CCPhysicist

    Just a shout-out in support of your indirect (footnote 2) reference to Heine-Borel and Cantor in the context of critical thinking. A recent ad on TV with the punchline “How about infinity times infinity?” suggests that students in the middle elementary grades could confront the issues raised in a few of the chapters of “One Two Three Infinity”. However, it seems that the still-neglected skill of basic arithmetic should be introduced first.