Teachers as Architects


I think architecture serves as a better comparison field than medicine does. Architects, like teachers, usually have multiple goals they try to satisfy simultaneously. Safety is nonnegotiable, but architects may also be thinking to a greater or lesser extent about energy efficiency, aesthetics, functionality, and so on. In the same way, some goals for teachers are nonnegotiable — teaching kids to read, for example — but after that, the goals are likely to vary with the context. In addition, architects make use of scientific knowledge, notably principles of physics, and materials science. But this knowledge is certainly not prescriptive. It doesn’t tell the architect what a building must look like. Rather, it sets boundary conditions for construction to ensure that the building will not fall down, that the floors can support sufficient weight, and so on.

In the same way, basic scientific knowledge about how kids learn, about how they interact, about how they respond to discipline — this knowledge ought to be seen as a boundary condition for teachers and parents, meaning that this knowledge sets boundaries that, if crossed, increase the probability of bad outcomes. Within these broad boundaries, parents and teachers pursue their goals.

-Daniel Willingham, When Can You Trust the Experts? (p. 221)

I really like this comparison, especially as someone who has compared teaching with medicine in the past. It’s making me think about what my “boundary conditions” are for learning. What is the scientific knowledge that I use to inform my teaching? Here are six principles that I think about with some frequency, and have found useful in my classroom:

  • Practice is essential to building flexible, durable knowledge. This practice should be spaced over time and interleave topics whenever possible.
  • The teacher’s job is not to make learning easy. While learning should not be so difficult that students are likely to give up, activities that are hard can lead to more powerful learning because the brain is more active in retrieving information and making connections.
  • A growth mindset is a major factor in learning. Shifting students’ mindsets  is difficult, but influencing these beliefs has a great deal of influence on future learning.
  • Problem solving and critical thinking cannot be taught independently of content. While these are critical goals, a strong base of knowledge across a range of content areas is an essential prerequisite for problem solving and critical thinking.
  • Treating students like experts is an inefficient way to lead them to expertise. Experts and novices think in qualitatively different ways, and expertise requires a great deal of carefully structured deliberate practice that is often distinct from the activities that experts excel in.
  • Incentives and feedback play a complicated role in students learning. Grading and other normative comparisons can reduce intrinsic motivation, inhibit future learning, and reduce the effectiveness of feedback.

None of these are prescriptive, but I find all of them useful. While they don’t offer specific guidance on how to teach, they can be particularly useful when I find out that students haven’t learned something. If I teach a topic, assess students, and learn that they don’t know what I thought they would know, I can use these principles to try to diagnose where something went wrong.

I’m curious if I’m off base with anything here, or if other principles belong on this list. For instance ,there’s a great deal of literature on feedback, explicit instruction vs constructivism, formative assessment, and more, but much of that research is conflicting or difficult to distill into concrete boundary conditions. I’m also curious about the complementary knowledge necessary for teaching that is not based on scientific principles, and how that all of that knowledge can fit together.

7 thoughts on “Teachers as Architects

  1. Carolina Vila

    That’s a great quote.

    I’m also curious about the “treating students like experts” bullet. Is this something I’ve somehow missed? What is the rationale for treating students like experts before they’re even close to fitting that title?

    1. dkane47 Post author

      Good question. There’s a great chapter in Willingham’s “Why Don’t Students Like School?” that addresses the expert-novice issue. I think the best example is in science class — “kids need to think like scientists” –> kids need to walk outside, formulate a hypothesis about something, design an experiment, test it, and report back the results. Then kids flounder because they don’t have the tools to be successful. It’s an extreme example, but there’s a certain logic to it, that students should have chances to do “authentic” work rather than dull practice. Faulty, in my opinion, but I do see the logic behind that. And I think there’s a danger in confirmation bias, where some kids who are high achieving are likely to be successful in those situations. If the teacher focuses on those successes they may continue doing it, believing that they are teaching kids to “think like scientists”.

      I used to do something similar in math. I would teach “problem solving” by giving kids hard brain-teaser like problems and asking them to try hard and persevere. I was acting as if putting them in the situation of being an expert would teach them how to be experts, when what they actually needed was to have the knowledge they needed broken down and structured in a useful, engaging way. The same confirmation bias challenge exists here — some kids will be successful, but the needle isn’t actually moving anywhere for what’s usually a large majority.

  2. Ryan

    Nice list.

    The expert/novice thing reminds me a saying we have in the software world: “you are not the user”. We need an explicit process of understanding the user rather than designing software that seems to fit our own perspective and preferences (https://en.wikipedia.org/wiki/User-centered_design) Similar in teaching is the divide between content knowledge and pedagogical content knowledge (https://en.wikipedia.org/wiki/Lee_Shulman#Pedagogical_content_knowledge_.28PCK.29). Also you should definitely read the expert/novice chapter of How People Learn if you haven’t: https://www.nap.edu/read/9853/chapter/5

    I’ve recently refined my understanding a bit more beyond “feedback is complicated” thanks to the feedback chapter in Embedding Formative Assessment. It’s about feedback as what change it invokes in the student: how does it change my relationship with the student? how does it cause the student to think about their mindset and approach to learning? how does it cause them to change their approach to some concept in future work (or in revising this work)? By the way, the reason feedback is complicated (at least according to my preferred cognitive theories) is that feedback is an essence of everything you do as a teacher. E.g. when merely speaking a sentence, there’s some amount of feedback against the student’s expectations of what you might say.

    1. dkane47 Post author

      That’s a great reading list! I definitely hear you on feedback — that’s a topic I personally want to explore more. But what Embedding Formative Assessment explores is a bit less research-based than the topics I chose above, and because of that I decided to leave it out. I would love to have some more concrete boundary conditions related to feedback, though — especially when not giving feedback (spending that time doing something else to plan for lessons) is the best choice!

  3. Pingback: Deliberate Practice: Mental Models | Five Twelve Thirteen

  4. Pingback: Limitations of Research | Five Twelve Thirteen

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s