Thursday, January 06, 2011

Why I care about the difference, and dependence, between facts and skill

Lately I have found myself drawn into commenting on Ta-Nehisi Coates’ blog over at the Atlantic. For a regular columnist, I find him a sophisticated commentator on issues of race, class and American history, as well as a writer whose prose is a joy to read. But what really drew me in was the other commenters on the blog, who can range from an interesting conservative construction worker from Baltimore, to a former teacher/ current boat builder to hyperliterate out-of-work librarians and former teachers. In other words, kindred spirits.  During a few recent posts (one on the Huck Finn controversy), I found myself ranting about something that most there seemed to feel was tangential to the issues at play; the fact that skills are based on a rich knowledge of facts. But it got me thinking about why I care so much about cognitive psychological distinctions that few people seem to care about. My answer is that the fallacies I rail against are the foundation of much of modern day education reform, and have an impact on even poor old Mark Twain.

What is the relationship between facts and skills? Many people believe that we need basic cognitive skills to learn things. In other words, the skills come before the facts. For example, you need basic reading skills to appreciate literature, learn more history. You need basic math skills like adding, subtracting, etc to move up to algebra and geometry. You need critical thinking skills to evaluate scientific findings. While this view makes sense to many of us, who know how to read, and know basic math, as we imagine that the rest of the things that we know are dependent on our basic reading and math skills.
This view is wrong.
What we think of as academic skills are based on a rich foundation of background knowledge. We think of all of these facts as furniture, decorating a structure held up by strong skills, but it is exactly the opposite. Rich content knowledge makes skills possible, not the other way around. People certainly had rich amounts of knowledge before the widespread use of the printing press, much of it was not gained through reading. Kids manage to learn incredible amounts of information before they can read. But we take all of this knowledge for granted, because we have forgotten that we even needed to learn it. At one point in your life, the meaning of the word “banana” was a fact that you had just learned. At another point, the word “forget” or “moment” or “count” or “knowledge.”  We have forgotten those moments, so we don’t realize these things are memory; we don’t count them as knowledge. Obviously the previous sentence would be gibberish without the vocabulary that I just mentioned. In addition, without “forget, moment, count or knowledge” you wouldn’t start to learn the new words of “realize” and “memory.”  Facts let you learn more facts.
 I was reminded of the importance of background knowledge as my boys were learning to read. Caleb was having trouble decoding (going from the groups of letters to the sounds they make) and testing behind in reading. But at the same time, we were reading to him, he was surrounded by picture books, and he enjoyed narratives and learning new things, whether they be the difference between types of camels (“the dwomadawy has one hump, but the Bactwian has two, and is from Mongolia, dad”) or the difference between a town, a city, a state, a county, a country, and a continent. While he took a little bit longer than others to be able to decode, Caleb’s reading “skills” have now miraculously vaulted forward. The point (other than to brag about my kid) is that he was never really behind. He had a rich background of facts, of building materials, and was held up just briefly waiting (well, actually working hard with the help of dedicated teachers) for decoding to happen. I don’t mean to minimize the fantastic job they did at his school with helping kids who have trouble reading, or the hard work that Caleb put in, but decoding happens. It takes longer for some kids than others, and it is no small thing to teach, but it does happen. And when it does, because most of us who know how to read as adults can decode just fine, the foundation for learning more is not our decoding skills, but our background knowledge. My colleague Dan Willingham has a great post about this, interpreting some test scores showing how our performance relative to other countries is lower in high school. The point is that it is lower in high school precisely because we are hammering on decoding and forgetting about the rich content knowledge. While the emphasis on skills is prevalent in elementary school, its effects are felt in high school, and in college too. One of my colleagues told a heartbreaking story recently, in which a student came up to her during a final and asked what the word “cumulative” meant. This student did not need more basic reading skills. He needed more facts. And he needed them in middle school, when he was busy being drilled for the SOL's.
This is an example for reading, but it is equally true for “critical thinking,” even in college. We would love to have the general ability to dissect a problem, analyze its parts, and evaluate its solutions.  But such a skill does not exist. I can do that fairly well with a psychology experiment (ok, maybe a cognitive psychology experiment, ok, maybe a visual perception experiment), but if I am evaluating the politics in Iran, or a medical diagnosis, or a book on the history of physics, I am kidding myself if I don’t realize that I am an amateur (and to the extent that I am an advanced amateur, it is because of extensive amateur reading I have done on medical diagnoses, or the history of physics).  Skills are generally a lot more domain restricted than we would like to think.
I have written here about how this leads people to propose silly redesigns of the college curriculum, with classes on “7 Essential Skills You Didn’t Learn in College,” one of which was applied cognition (“the neuroscience you need”). All of those skills are dependent on background knowledge. Other people propose that we just teach kids “how to search” and “how to evaluate online sources” and then they don’t need to know any facts because they know how to find whatever they need. Unfortunately, this is not a skill. Knowing how to parse google search listings for the right links, or peruse Wikipedia and actually learn something while avoiding the controversial or unsupported errors is a skill that is dependent on content-specific knowledge.  
What is the best way to acquire facts? Reading lots of books. That’s it. Being interested in stuff, and reading about it. Somehow our educational system has forgotten this. If a student gets interested in Star Wars, or furniture design, or spiderman, or environmental law, or drug policy, give them a few books, and let them go to town. Of course, we should have an idea of what we think they will find interesting, and give them some of that too (race in America? Social psychology? Autism?).   But their patience with our books will wane if we keep telling them to put theirs down.
Which brings me to the basis of education reform, and the furor over Huck Finn.  Many education reform arguments go like this:
1)American education is failing our students, and we are falling behind our international competitors. 
This statement is often made based on international tests, such as the PISA, or the lack of improvement on our own standardized tests such as the NAEP. These tests are most often of basic reading and math. It often goes unchallenged that these are reflective of the rest of our education system.
2) To catch up, we need to catch up on these tests (or, if our reforms cause us to catch up, we would notice it on these tests).
This has two meanings. First, we need to narrow the “achievement gap.” What the achievement gap most often means is that performance of poor, urban (mostly black and Hispanic) students is below performance of white and wealthy students on these tests of basic skills. This often leads to the well intentioned effort to work these poor children harder (more test prep drilling, school on Saturdays, no recess) so that they can catch up on these tests. This approach has not been a resounding success.
The second is that we need to narrow the gap between the US and other countries on these basic skills. For this, the logic is somehow that to remain economically competitive, we must be educationally competitive. Yet if we look at the innovations that have kept the American economy strong over the past 50 years, they are not the amazing basic skills of the workforce, but the amazing creativity of a relatively small group of Americans. This creativity has occurred with the aid of different sorts of programs. Malcolm Gladwell outlines how Bill Gates was able to access a computer very early on in his education.
3) The best way to catch up is to increase accountability and teacher quality
Accountability measures have taken the accountability out of the local hands (principals) and urged greater standardization. To have standardized accountability, you must rely on a measure everyone agrees matters. Enter tests of basic skills. Principals and teachers, railroaded by the particular kind of standardized accountability instituted by reformers, drop everything and do more test prep.
Which leads them to drop books from the curriculum, because who has time to read a book when you need to boost your test scores ten points or lose your job, or get your school closed? And so, yes, Huck Finn has always been censored, and always been controversial. But teachers have been teaching it. But teachers are just exhausted from dealing with the pressure to raise test scores, who has the time and energy to deal with trying to explain to teenagers the complex motives of Huck Finn, and the satirical wit of Mark Twain? 

In the end, we have the paradox that our students spend a lot more time practicing these basic skills and less time learning facts, when the time comes to actually show that they have these skills, they don't perform well because they have no background knowledge, not to mention not much interest, because in our mad dash for skills, and accountability and performance, we stopped asking them what they were interested in.


The Triscuit's Dad said...

I'm buying your argument here for the most part, and certainly agree that the reformers have it wrong. But is there really no such thing as critical-thinking skills, independent of content? I thought of LSAT/GRE logic problems, or puzzles in general. While it's problematic to base school admissions on these kinds of puzzles, don't they measure SOME kind of problem-solving skill? And isn't that skill valuable? (Or maybe that's too trivial an example.) I guess I wonder if you're overstating the "no skills" point. Are you arguing that there is no such thing as a critical thinking skill, or just that any such skill has to be learned through exposure to content (of SOME kind), and that's what we've forgotten?

Cedar said...

This is a good point, and I may have overstated my case just a little bit.

There are elements of our thinking that apply to more than one domain. There are cognitive abilities that apply to more than one domain.
You are taking what some people think of as abilities (logical reasoning ability on the LSAT or GRE) and categorizing it as a skill. You are right that this "ability" can be improved.
But I think the generality of these abilities is debatable, and the real work they are doing in your thought is minimal.

It also might be worthwhile to distinguish between skills and strategies. Strategies can be learned and adopted fairly quickly, and don't really need practice to be mastered. Skills need to be practiced to be improved.
Skills are generally far less general than we would like to imagine. And the ironic part is that quite often, the more that you practice that skill, the more expert you become on it, the more narrow your skill becomes. One of the best squash players in the world was an above average ping player (when I played him in college).
Which is not to say that strategies should not be taught (I try to give my students study strategies for tests in my courses, or note taking).
Finally, I am not saying that there is no such thing as critical thinking, just that I think it is far more narrow than most people realize.

Cedar said...

To clarify on the abilities point:
Someone can test very low in logical reasoning ability, but learn a whole hell of a lot about cars, and be amazing at deducing, diagnosing and fixing cars. They are not fundamentally limited by their reasoning ability, but by how much knowledge they have, and how their knowledge is structured.

UnemployedDrifter said...

This post is a much more thorough and nuanced expression of something i have argued (mostly to myself) over the years. The denigration of facts as somehow a meaningless aggregation of dates that can always be looked up gives short shrift to how crucial the assembly of knowledge is in fleshing out the world.

Standardized tests are generally barren. Which begs the question in my mind of their value. They can't be avoided perhaps, but narrowing the focus of a child's education to try and ensure s/he can at least pass one of these tests is probably a net negative, even for the most underperforming of students.

I'm curious about whether you think our lack of a national curriculum is a strength or a weakness? slightly off topic, i know.

Cedar said...

@Unemployed Drifter:
Thanks for reading!
I think that tests are an amazing tool for learning. Teachers give tests for a reason, studying is better for learning than reading. I think that many standardized tests themselves can be useful as a rough measure, and as a diagnostic tool. But I think that you pay for what you get. If you are willing to pay for subject matter experts to work with testing experts to spend some time with every essay, and design good short answer and multiple choice questions, then you can have a pretty good assessment of learning in that domain.
Even the math and reading tests themselves can be useful to teachers to benchmark their kids relative to the class and others their age.
What makes these tests a problem is trying to make them solve problems they weren't designed to solve.
My problem is not with the tests of basic reading and math skills, which have their place, but with the increased accountability pressure that makes us conclude that learning we can easily measure is the only kind of learning we care about.
I am not sure about a national curriculum. In general, I think there should be more local control, and less national accountability. I would be in favor of some national curricular materials, let's say a few great US History textbooks, written by a committee of eminent historians, history teachers, and a few parents being available to teachers, but not being forced upon them.
I would much rather leave the teachers with both more control than we have now, and more resources. To the extent a national curriculum is able to provide resources without restricting control, I would be in favor. But I am dubious. National resources in education lately seem to come so full of strings that they are actually counterproductive (see: Race to The Top).

UnemployedDrifter said...

These are good points about testing. There is an incentive value in studying FOR something- as I have found out in my inchoate, auto-didactic adult life- and not all tests are created equal. So, too, as a diagnostic tool for basic skills, they can have their use to identify underperforming- or overperforming- students.

When i used the words standardized tests, I had in my mind bubble fill-in multiple choice question ones inflicted upon my youth. Surely short-answer, essay, and other formats have already been incorporated meaningfully into these high-stakes tests; this is an improvement.

I do think the reductive nature of any testing mechanism makes their value (excepting perhaps basic math and reading assessments) inversely proportional to how wide a target demographic you are testing. Or at the very least, the bigger the group you are testing, the harder it is to make that test a good one.

well said about learning we can measure being conflated with learning that is most crucial- a fundamental pitfall of testing. And while not a reason to ditch testing as a whole, it should give pause to anyone trying to measure performance, whether a classroom teacher for forty kids, or a committee for forty million.

As for national curriculum, I say science and math at the very least. History will never happen (unless we all adopt Texas history as the standard); and science probably no time soon either, as we must ensure no infringement on the constitutional right (article 59.345) of Kansan schoolchildren to have equal time to weigh intelligent design/creation science against other vague, incomplete theories purporting geneological fellowship amongst, humans, monkeys and slime...

Good post. Thanks for writing!

Sam said...

So what do you think of ED Hirsch?

Cedar said...

I am generally in favor of the approach of his Core Knowledge Foundation. I don't think it is a magic bullet, and I would like to see his vision paired with an approach that respects some independence for teachers and students, and fostering students' intrinsic motivation and curiosity, which I don't think their approach addresses specifically. But I am generally very sympathetic to the overall approach.

from their website:
In all of its publishing, support and advocacy work, the Core Knowledge Foundation is guided by the following principles:

Our work is not driven by ideology, but logically by science, history, and research.

For the sake of academic excellence, greater equity, and higher literacy, elementary and middle schools need to teach a coherent, cumulative, and content-specific core curriculum.

The persistent gap in reading achievement in U.S. schools can never be reduced until the knowledge gap is reduced. And the knowledge gap will not be reduced unless broad, rich content knowledge is integrated into the many hours devoted to language arts instruction.

We recognize that every school and community is different, and each student and teacher has individual interests and strengths. Schools teaching the Core Knowledge curriculum should still have ample time over the course of the school year to address any additional state or local requirements not reflected in the Core Knowledge Sequence.

An effective curriculum must be coupled with effective teaching. We believe teaching excellence requires a mastery of subject matter, as well as the ability to engage students, build language competency, use assessment to drive instruction, scaffold instruction to meet individual needs, and provide targeted feedback to students to further shape their learning.

We need to see the reading comprehension problem for what it primarily is–a knowledge problem. There is no way around the need for children to gain broad general knowledge in order to gain broad general proficiency in reading.

—E. D. Hirsch, Jr.