Learning with Examples

As you may have noticed, I've been working on the design of this blog: mostly color changes but also a fluid design for the content side. At first, I started trying to wrap the posts around at the bottom of the sidebar. I did my research, read the tutorials, but couldn't figure it out. I emailed Mark Bernstein, the designer of Tinderbox, this weblog's software application, suggesting that the feature be incorporated into later versions. He responded (Mark's generosity with his time is unbelievable!), offered to do it for me, and, one hour later, sent me my weblog file re-coded with the fluid design (a switch from the earlier wrapping style). And I continued with changing the colors, which is not a straightforward process for someone who is colorblind. (I use the Color Generator and patient friends.)

What's this got to do with learning with examples? Well, I've learned by observing what Mark did. Previously, I would duplicate an entire file to have a practice file; Mark simply added a new CSS note. Previously, I would export an entire document to see how it looked in html. While in Boston, I noticed that Mark just used the Preview button. And from the code he sent, I began to understand the difference between "float" and "absolute". In trying to re-design this blog,I spent two full days acquiring quite a bit of frustration but little understanding, as opposed to taking a few minutes to look at Mark's re-coding to learn where I had gone wrong.

Such incidental learning via examples underscores John Anderson's ACT-R learning theory. Anderson, a professor of psychology and computer science at Carnegie Mellon University, is the one who first posited two types of knowledge: declarative and procedural. I've posted on Anderson before (see "Lies teachers tell?" and "Forget IQ. Just Work Hard!".

Anderson and Schunn in their article "The implications of the ACT-R learning theory: no magic bullets" (pdf) write:

There are no magical properties conveyed upon a knowledge structure just because it was self-generated. If all things were equal it would be preferable to have children learn by generating the knowledge (due to the redundant encoding). However, because of difficulties of generation and dangers of misgeneration, things are not always equal and it can be preferable to tell the knowledge.


Thus, ACT-R's theory of procedural learning claims that procedural skills are acquired by making references to past problem solutions while actively trying to solve new problems. Thus, it is both a theory of learning by doing and a theory of learning by example.

Simply providing the learner with examples is not sufficient to guarantee learning in the ACT-R theory. The sufficiency of the production rules acquired depends on the understanding of the example.

Anderson and Schunn add, "For competences to be displayed over a lifetime, time on task is by far and away the most significant factor." That is, learners must practice a lot. The problem is one can practice the wrong skills, in which case "practice makes imperfect." In other words, learners need feedback and explicit guidance, often in the form of examples, to make their practice effective.

But how can examples be so effective? Perhaps because human beings learn mostly through imitating. Vilayanur Ramachandran, Director of the Center for Brain and Cognition, posits that imitation via mirror neurons is the driving force of human evolution:

With knowledge of these neurons, you have the basis for understanding a host of very enigmatic aspects of the human mind: "mind reading" empathy, imitation learning, and even the evolution of language. Anytime you watch someone else doing something (or even starting to do something), the corresponding mirror neuron might fire in your brain, thereby allowing you to "read" and understand another's intentions, and thus to develop a sophisticated "theory of other minds."

An earlier post "Be Happy, and Learn!", commented on Kathy Sierra's post "Angry/negative people can be bad for your brain" on the effect of mirror neurons on one's emotional state:

There is now strong evidence to suggest that humans have the same type of "mirror neurons" found in monkeys. It's what these neurons do that's amazing--they activate in the same way when you're watching someone else do something as they do when you're doing it yourself! This mirroring process/capability is thought to be behind our ability to empathize, but you can imagine the role these neurons have played in keeping us alive as a species. We learn from watching others. We learn from imitating (mirroring) others. The potential problem, though, is that these neurons go happily about their business of imitating others without our conscious intention.

Note, however, that this imitation is an unconscious process. I'm not quite sure of the relationship between consciously understanding and using examples and imitating those examples. Perhaps understanding comes through imitation + practice.

Although we wouldn't want to limit ourselves to learning by imitation, the fact that imitation is such a strong component of learning should give us pause when we read statements that denigrate imitation and position it in opposition to creativity.

Paul Butler argues for re-introducing imitation into composition in his article "Imitation as Freedom: (Re)Forming Student Writing":

For many years now, the use of imitation in the composition classroom has been waning. As Connors points out, articles on imitation, sentence combining, and generative rhetoric have steadily declined and have been almost nonexistent since 1995. Yet in composition classrooms all over the country, as we adopt various process techniques, we still hold our students accountable for the fundamental elements of good writing: organization, coherence, unity, and clarity, among others. Lisa Delpit has pointed out that our expectations are sometimes “hidden,” that they remain invisible to students as we encourage them to explore their ideas and work within the process model of teaching. Delpit’s argument, though intended to address the situation of minority students, also applies to students in composition classes around the country. Indeed, it seems the height of hypocrisy to use strictly process techniques when we expect high quality “products” from our students’ writing.

Along these lines of using examples and imitation, I commented previously on They say/I say : the moves that matter in academic writing, a book by Gerald Graff and Cathy Birkenstein that uses templates to help students see and be able to make the rhetorical moves of academia.

I think that most of us forget how often we use and appreciate examples when we enter new territory. For instance, when writing my first book review, I looked at dozens of other book reviews to understand this genre's requirements. I imagine if someday I write a grant proposal, I'll do the same, too. And, I imagine that most writers follow suit. If we learn this way, then why wouldn't our students do so, too? Why do we expect them to start from scratch when we don't? And with respect to EFL/ESL students who don't have a strong L2 cultural foundation for learning L2 writing by "osmosis," the case for making explicit the implicit is even more essential.

None of this is an argument for rote memorization of models. Rather, it's an acknowledgement that if we are wired for imitation, for learning with examples, then why not take advantage of our "wiring" when designing class activities?