Is Teaching an Art or a Science?

by Robert Pondiscio
May 30th, 2012

That’s the question Dan Willingham poses in a new video.  As you likely know, Willingham is a University of Virginia cognitive scientist whose work focuses almost exclusively teaching and learning.  The video is worth watching, but – spoiler alert! – his conclusion is that teaching is neither art nor science, but “somewhere in between.”  He draws a parallel to being an architect, who understands enough about physics and materials science to design a building that won’t fall down.  But like an architect, a teacher “then also uses creativity and ingenuity to go beyond any strictures that science can offer, to create something wholly original, functional, and enduring.”

What is most interesting, and should spark the most informed and intense viewing of the video, is Willingham’s take on how science should inform teaching.

“I think what we know about how kids learn, and develop, interact and so on, can suggest boundary conditions. I mean by that things that, if you ignore them, will likely lead to trouble. For example, if you think that someone will acquire a skill without practice, that’s probably not going to work.  Providing practice for skills is a boundary condition. Another example, most kids benefit a lot from explicit instruction in letter-sound correspondences when they are learning to read, so providing that instruction is another boundary condition. Notice this doesn’t tell you how to implement practice, or how to implement teaching letter-sound correspondences—it just says that those things have to happen. I call these ‘must have’ principles.”

But in addition to the “must haves” there are a series of “could dos.”  In Willingham’s architecture analogy, there are ways to put a window in the middle of a brick wall that protect the integrity of the wall so it’s structurally sound.  But there is no rule saying you must have a window in a particular location, or have one at all.  This allows for a broad range of teaching approaches and activities, all of which could be good, useful or even elegant.  Just like architecture.

“The ‘must haves’ and the ‘could dos’ do not tell you what the house is going to look like.  The ‘must haves’ are boundary conditions, within which there is a HUGE amount of room for variation, and the ‘could dos’ are tools that you can use to help you get there, but you don’t have to use them if you don’t want to, as long as you respect the ‘must haves.’

Willingham is offering up his usual dose of fact-based common sense, but I’m tempted to suggest there might be little agreement on “boundary conditions” for teachers.   The preponderance of evidence may indeed come down on the side of phonics, as he suggests, but that hasn’t entirely settled the issue.  Whole language still has its adherents and repackagers.  Discredited ideas like learning styles remain hardy perennials.   The larger problem, to put it bluntly, is that education  pays insufficient evidence to science.

Boundary conditions in architecture or civil engineering are inherently self-policing.  Violate them and things fall down.  I’m having a difficult time thinking of a set of universally agreed upon “boundary conditions” for teaching.  But this creates an enormous opportunity for the cognitive scientists like Willingham to frame the discussion and offer evidence-based guidance on those “must haves.”  And an enormous obligation on the part of ed schools and programs that train teachers to pay attention.

It will take a whole lot of science to move the field past its self-image as an art, its neglect of science–and the tyranny of its philosophers.


  1. “Tyranny of its philosophers”? This is a naive statement. Read anything about how science actually works & you’ll find philosophy is of great importance as evidence only makes sense in light of a guiding framework (philosophy). Philosophers provide the questions that scientists seek to answer.

    Comment by Jerrid Kruse — May 30, 2012 @ 10:08 pm

  2. I accept the edit. Tyranny of its ideologues.

    Comment by Robert Pondiscio — May 30, 2012 @ 10:30 pm

  3. Art or science?

    In the US it appears to be more art, more touchy-feely, creative, and/or problem solving.

    In Japan, with their national data base of exemplary lessons, I’d have to believe they’re more of a scientific mindset and thereby have a better chance of success based simply on evidence of successful lessons.

    Comment by Paul Hoss — May 31, 2012 @ 6:10 am

  4. For the purposes of boundary-setting, I wish every future K-12 teacher had to read Stanislas Dehaene’s Reading in the Brain. It explains very clearly why systematic phonics instruction is crucial, and also why children and adults do use sight recognition at the very beginnng of learning to read and also when they have become fluent.

    Comment by EB — May 31, 2012 @ 8:39 am

  5. @Anthony Excellent comment, and I think — I fear — you’re exactly right. Another example is learning styles, which was covered on the very first day of my training in the NYC Teaching Fellows program. I suspect it will be a long time before we see the end of some of these “out of boundary” conditions.

    Comment by Robert Pondiscio — May 31, 2012 @ 9:48 am

  6. @Anthony @Robert I think reading strategies and learning styles represent two different types of problems. The problem with learning styles is that they are not supported by the bulk of the research community. Reading strategies are–the problem there is too much zeal.

    Comment by Dan Willingham — May 31, 2012 @ 10:35 am

  7. @Dan School me: Reading strategies might not meet your test for a “boundary condition” but they certainly have the imprimatur of science (do they not?) given their prominence in the 2000 “The Report of the National Reading Panel: Teaching Children to Read,” a putatively research-based report. Thus, wouldn’t the field reasonably conclude they are scientifically sound?

    This no doubt the point you were making in the video — research may show some limited efficacy for reading strategies. But how do we ensure teachers can be informed consumers and appreciate the nuances?

    Comment by Robert Pondiscio — May 31, 2012 @ 10:44 am

  8. @ anthony

    The irony of being asked to teach your children to think critically while failing to do so yourself is indeed rich.

    Perhaps the emphasis on reading strategies is like the blood letting to which Dan makes reference in his video? Unproven, often counter-productive, yet widely followed?

    Comment by matthew — May 31, 2012 @ 12:10 pm

  9. My experience in having tracked down the history of these ideas is that the reading strategies emphasis comes ultimately from people who view fluency with print as a fetish to be discouraged. Reading phonetically and fluently has a necessary side effect of liberating abstract thought. Political schemers know that and write about it. It’s time we knew it to.

    Likewise if you read Howard Gardner’s work from the early 80s, he is pretty graphic that he sees learning styles as a means of deemphasizing an academic orientation.

    And the art of teaching to me tracks back to John Dewey’s aesthetic vision for the classroom.

    Great video Dan.

    I think all of you will be saddened with the direction of today’s post on what is being pushed and mandated under the Education for Sustainable Development mantra.

    Next I will explain how the accreditors like AdvancED have become enforcers for this EFA and ESD vision coming in under Common Core.

    Comment by StudentofHistory — May 31, 2012 @ 12:25 pm

  10. I, too, object to the suggestion that “philosophers” are the problem here. I realize, Robert, that that wasn’t what you meant–but I want to defend them anyway. We need more of them–for in order to make sense of research, we must first establish what we are trying to do and why.

    For instance, reading strategies may be moderately effective in helping students tackle generic texts, provided the students know hkow to read in the first place and have adequate background knowledge. But tackling generic texts is a pretty drab goal. If one wants students to read substantial literature, then one must devote attention to the works themselves. Strategy instruction should be peripheral and subordinate to the literature.

    Likewise, phonics instruction is effective in teaching students to decode. IF that is the goal, then phonics instruction is an appropriat means toward that end. But there are many children who do not he troiuble reading. For them, the instruction should take a different form (and not that of Balanced Literacy, in my view).

    Education discussion often assumes the struggling student. It is important, of course, to find ways to help the struggling student. But this is not the whole of education and should not become the whole.

    Comment by Diana Senechal — May 31, 2012 @ 12:47 pm

  11. The architecture comparison is apt, there is a pretty vast difference between a sky scraper and a shack. I would also say that it epitiomizes the American education systen in terms of what we are investing in children.

    In terms of reading strategies, I think it is popular because to contemplate the alternative that a child needs to read a lot more, possibly relearn phoenics anddevelop significant background knowledge completly overwhelms failed systems. If you have a middle schooler reading 3-4 grades behind it is pretty hard to contemplate what it will take to get that child up to grade level if these are boundary conditions, you look for strategies because the system does not have the resources in time or people to change the trajectory. Thus the many leaning towers our education system is producing.

    Comment by DC Parent — May 31, 2012 @ 1:16 pm

  12. @robert @matthew @studentofhistory: the scientific evidence supporting the *short term* value of reading strategies is, indeed, very strong. It’s been replicated in lots of classrooms/schools. What has NOT been established is that continued practice of reading comprehension strategies does much good.

    @Diana et al I think what Robert meant by “philosophers” was not “people who offer a framework in which to think about data” but rather “people who guide practice by theories and ignore or misinterpret data that conflict with their theories.”

    Comment by Dan Willingham — May 31, 2012 @ 1:44 pm

  13. Dan-I always smile ironically when I read that teachers may give explicit instruction in explaining strategies for reading or math problems. But they are not to explain the phonetic basis of the English language or actually systematically work out a math problem.

    That should be inquiry. Right. Because acute frustration is a useful emotion.

    It is quite clear to me Common Core intends to dole out vocabulary whole word style over years. It really is everything Marie Clay ever envisioned. Which is not good. It’s also why there are lots of links to materials on New Zealand servers illustrating the literacy progression.

    Diana-the Coalition for Essential Schools was the template for Australia’s reforms and looks to me to be the template for common core’s actual implementation. Yesterday I was tracking the history behind public charters. Going back to some 1988 documents before the last go around of national ed reform in the 1990s, it graphically laid out the vision that ed reform should be targeting the 80% of students who could never do well in traditional instruction.

    That’s taking down all the able kids. Plus the reaction to all the state funding litigation over the years and the IDEA FAPE standard is “adequate” is all anyone is entitled to from a public school.

    I really am not just being contrary here with my insistences the implementation looks nothing like the PR sales campaigns for the standards.

    Comment by StudentofHistory — May 31, 2012 @ 2:09 pm

  14. Dan, yes, I realize Robert wasn’t disparaging those who ask what education is and what it is for.

    But the question needs to be raised again and again: what is it that we’re trying to do? Much of the confusion over education research is due to the burying of this question.

    For example, at a typical professional development session you’ll hear that “research has shown” that cooperative learning has both academic and social benefits. But even if this is so, one must ask whether the benefits constitute the whole of what students should gain from school. One should also ask whether cooperative learning might not have drawbacks and losses as well.

    An understanding of “boundary conditions” can help us distinguish between those goals and means that are non-negotiable and those that may vary from school to school or course to course. This can clear up a good deal of the muddle. So thank you for putting forward this interesting and helpful idea.

    Comment by Diana Senechal — May 31, 2012 @ 2:52 pm

  15. Anthony G. argues convincingly that most teachers need to pay more attention to the science of teaching and learning. My personal experiences as a school board member and as a parent are these: the older a particular teacher’s students are, the more likely that teacher is to realize the need for background information to comprehend text. A high school history teacher knows from her reading of history that if a student is to write effectively about, say, the Civil Rights movement, he’ll have to know a lot about the history of African-Americans: slavery, the Civil war and Reconstruction, separate but equal, Brown vs. Board, etc. But elementary teachers in CK schools have to be educated into the CK philosophy and cognitive science – very few of them even heard about it in college.

    Even more depressing is how the mantras of reading strategies and learning styles are almost universally believed by parents. This is true even among parents whose kids have attended CK schools for many years. The worst of all parents’ beliefs is that it’s not that important what a kid of any age reads, as long as she’s “reading something”, be it Harlequin Romances, comic books, or Dostoevsky. These scientifically unfounded beliefs are deeply embedded in American culture, and like almost all cultural beliefs, will be very slow to change.

    Robert’s best line is “Willingham is offering up his usual dose of fact-based common sense…” That’s true for us CK true believers. Alas, the fact-based sense is none too common.

    Comment by John Webster — May 31, 2012 @ 3:06 pm

  16. No problem here with philosophers and ideologues in education–as long as they fess up when indulging in ideology. All education is philosophical to a point. The real culprits are the ideologues who present their personal, philosophical preferences as scientific fact.

    There is a high priesthood in this business; they wear the robes, speak the holy words, and enforce the canonical law–but they expect the rest of us to obey their dogma because it is “research-based” and thus scientifically vetted. At least Jimmy Swaggart and Jim Baker never invoked research when spouting their silliness.

    Comment by James O'Keeffe — May 31, 2012 @ 4:49 pm

  17. James- this past week the National Academies put out a very troubling report to remake the nature of higher ed based on something called Discipline-Based Education Research.

    It makes quite clear what I have suspected for a while. Research based means theory in practice. Take good notes on the effects on actual students because they are the guinea pigs for seeing what happens.

    Even worse than that experimentation on students who are at least technically adults, there’s no research support for Common Core either.

    A November 2011 NSF funded report called “A Priority Research Agenda for Understanding the Influence of the Common Core State Standards for Mathematics” makes it clear there is little research to support what is being mandated nationwide. Plus they would like more funding to examine the effects in the districts and states most aggressive in their implementation.

    This is economic suicide. There is so much unappreciated cultural knowledge embodied in all these traditions being rejected.

    Comment by StudentofHistory — May 31, 2012 @ 7:04 pm

  18. @ James: “There is a high priesthood in this business; they wear the robes, speak the holy words, and enforce the canonical law–but they expect the rest of us to obey their dogma because it is “research-based” and thus scientifically vetted. At least Jimmy Swaggart and Jim Baker never invoked research when spouting their silliness.”

    I wish there was a “like” button for this.
    But I wouldn’t compare Lucy Calkins, Ken Goodman, and the rest of the lot to Jimmy Swaggart. They are more like Deepak Chopra, invoking science to persuade those who know as little about science as they do.

    Comment by alamo — May 31, 2012 @ 9:57 pm

  19. Alamo: I’ll meet you halfway. Throw in Tony Robbins and Shirley MacLaine, maybe Cher during her infomercial period, and I think we’re on the same wavelength.

    Comment by James O'Keeffe — June 1, 2012 @ 12:08 am

  20. @ James “No problem here with philosophers and ideologues in education–as long as they fess up when indulging in ideology. All education is philosophical to a point. The real culprits are the ideologues who present their personal, philosophical preferences as scientific fact. ”

    I couldn’t agree more. I think what is mostly missing from education debates is a clear understanding of (1) how values influence educational goals; (2) what science can and can’t do to further educational goals; (3) the challenges in implementing basic scientific findings in education.

    Comment by Dan Willingham — June 1, 2012 @ 11:37 am

  21. Dan,

    I have spent the last several years going through the footnotes and locating the cited supporting research. When certain influential people think they are off the record they get quite graphic on intent. Before things became controversial, they were quite graphic on intent.

    To me the intended goal of the designer or the intended goal when you think you are among fellow true believers is what is relevant. We parents and taxpayers then get some politically palatable or factually plausible stated goal that we are supposed to accept at face value. Why? Because they claim to be the professionals.

    In my mind and based on the evidence I have this simply encourages the accreditors and some of the colleges of ed and many central office admins to game the system and misrepresent everyone’s real intentions to the public paying the bills. And living with the consequences of the actual intended goals.

    Taxpayers and parents need to better understand the distinction between intended, stated, and enacted curriculum.

    Comment by StudentofHistory — June 1, 2012 @ 3:20 pm

  22. One of the most interesting aspects (to me) of Dan’s video is its emphasis on the difficulty of applying science to the classroom. It’s problematic to apply laboratory findings to the classroom, for instance, because classrooms lack the controlled conditions of the laboratory. It’s difficult, likewise, to apply findings from one classroom to other classrooms, because classrooms are highly variable.

    This doesn’t mean that research has no place; it just means that it should be approached with caution. A study may suggest a finding for a particular situation, but that does not necessarily apply to other situations.

    And even when you have a research conclusion that seems relatively robust, it may have little bearing on what you’re trying to do.

    Too many studies assume that a test score increase is the ultimate goal. So, for instance, you could have a study that suggested that the inclusion of Plato in the curriculum did not correlate an increase in ELA test scores. The conclusion would be that studying Plato wasn’t an “effective” means of raising achievement. But who says increased test scores are what you’re trying to achieve? What if you were trying to achieve an increased understanding of Plato?

    Likewise, there have been studies that suggest that the study of music does or does not correlate with high achievement in other subjects. But wouldn’t one want to see whether the study of music correlated with better performance/understanding of music itself?

    In other words, our educational priorities influence not only our interpretation of research but even research design.

    Comment by Diana Senechal — June 2, 2012 @ 9:14 am

  23. @Diana- you noted “In other words, our educational priorities influence not only our interpretation of research but even research design.”

    Do you think this difference is part of the debate between teachers, their unions, parents and policy makers? When people ask me about how my kids are doing with school, I often sigh and say well.. it is not so much their school but the American education system I feel frustrated by and how that then affect public schools. I really wish more often the debates would start with the old theory of bureaucratic politics, “where you sit is where you stand.”

    Comment by DC Parent — June 2, 2012 @ 12:58 pm

  24. DC Parent,

    Yes–I’d say that some of the confusion in education discussion is due to clashes of unspoken assumptions. People use the same word (e.g., achievement, curriculum) to mean widely different things.

    Comment by Diana Senechal — June 2, 2012 @ 4:44 pm

  25. I would greatly appreciate some advice from the expert teachers and cognitive scientists on this blog. My concern expressed here is right on-topic about teachers not using basic scientific findings in education.

    As I write this comment, I’m seething over the study guides my son’s teachers handed out for next week’s final exams (middle school). Each guide has a lot of material, most of it worth knowing long-term, in my layman’s opinion. Some of the facts and concepts have been recently reviewed in classes, but students are expected to review the vast majority of the material on their own.

    Sounds good, so far. One BIG problem: all the study guides were given to the kids last Thursday or Friday, and finals start Monday, three days later. In other words, the kids are expected to “cram” most of the material in a span of less than a week. This expectation for “cramming” has been common this year.

    I love what they study in this Core Knowledge school, and the teachers really do support the curriculum. But good grief – haven’t they heard of using “distributed practice” as the scientifically proven best way to memorize material over the long-term? I grasped this concept through personal experience in college 30+ years ago, and I’ve never had a course in education.

    How can these dedicated teachers not know about distributed practice, and how useless cramming is? For the teachers reading this comment, how long a timeframe should middle school students have to study for finals or other major exams, if distributed practice techniques are to be used to most advantage?

    My kids’ school will have a new middle and high school dean in a few weeks, and I plan to bring this matter up in a constructive way. Any advice from the CK crowd is most welcome.

    Comment by John Webster — June 3, 2012 @ 11:57 am

  26. John,

    My sense is that most teachers would share your concerns. I have often heard teachers say that they have to cram too much into the year and that the end of the year is rushed no matter what they do.

    No one, as far as I can tell, believes in cramming. So, why does it happen?

    I know that many CK schools must address state standards and local requirements, in addition to teaching the CK curriculum. That might account for some of the cramming. On its own, the CK sequence allows for plenty of wiggle room (from what I can see). It’s even possible to work in the additional requirements without frenzy–but it takes some sorting out and some practice.

    But no matter how well a teacher plans, the end of the year takes people a little by surprise. There are lots of events, and the weather gets warm, and before you know it, classes are over.

    What is the optimal timeframe for studying for finals? It depends a bit on the course. In some courses, such as mathematics, much of the review is built in. That is, you need to know the former material in order to understand the new material. You’re continually drawing on what you’ve learned before. In other courses, where the topics might not connect quite so tightly, more review time might be needed. Or else the teacher might find ways to work in review along the way.

    My guess is that the review packet was meant mainly as a resource, not as an assignment. A student could read it fairly quickly to determine which parts of it were solid in the memory and which might need some revisiting.

    Yes, it would generally be good to hand out review packets more than a few days in advance of the final. But the late distribution of the packets does not necessarily mean that the teachers lack appreciation of distributed practice. It may simply mean that things were hectic.

    In any case, there’s nothing wrong (in my view) with bringing up the matter courteously. There may well be a way to get the packets out earlier next time.

    Comment by Diana Senechal — June 3, 2012 @ 8:44 pm

  27. In theory, study guides are a sort of fail-safe: they are supposed to remind students of what they have already learned, and if the student reads a section of the study guide and finds that s/he doesn’t have a good grasp of it, s/he then knows to focus on that topic for review. But over time they have been turned into cram guides.
    This is not the fault of the study guide, but of the way they are used.

    And I say this as someone who went to HS and college before there was any such thing.

    Comment by EB — June 4, 2012 @ 8:33 am

  28. [...] Pondiscio at The Core Knowledge Blog highlighted a video from University of Virginia cognitive scientist Dan Willingham that poses the age-old question: Is [...]

    Pingback by In Another Place · Art or Science? — June 9, 2012 @ 10:48 am

  29. All chlidren can go All chlidren can go to the library. There are a wealth of chlidren’s authors who write about math, science, history, baseball anything and everything. Any parent can take their kids to the post office and teach them about mail. Any parent can take their kids to the grocery store. You don’t have to be well-traveled or well off to have life experiences. Life experiences happen to everyone, regardless of whether they are wealthy or not.

    Comment by Ahmed — June 27, 2012 @ 10:37 pm

  30. [...] Pondiscio described the video and analyzed Willingham’s answer to the question on the Core Knowledge [...]

    Pingback by EdNext Readers Poll: Is teaching more of an art or more of a science? | Courses Newsadmission news | study abroad | school courses | college courses — August 1, 2012 @ 1:28 pm

RSS feed for comments on this post. TrackBack URL

Leave a comment

While the Core Knowledge Foundation wants to hear from readers of this blog, it reserves the right to not post comments online and to edit them for content and appropriateness.