Saturday, January 22, 2011

Blue's Clues and High-Definition Writing

My two-year-old son is a huge fan of the children's show Blue's Clues, though he doesn't fully understand it yet. Watching the show recently with him, I came up with a way to illustrate a point I often try to make with students about what good writing does, and I thought I'd share it here.

Let me start with a claim that will seem strange to anyone who knows Blue's Clues.

CLAIM: Blue's Clues always has more than three clues, in every episode.

If you've ever seen the show, you'll probably recall that Blue leaves her paw prints on the clues, but you may be wondering if I can count, since she always leaves three paw prints. One of the catchphrases of the show is, in fact, "We've found all three clues!"

Now, I could just leave you with my thesis statement: The show always gives more than three clues. But by itself, that's pretty cryptic. You might reject it outright, since you don't know why I said it.

So now I have to defend it. Let's consider an example. In one episode, Joe and friends are putting together an "alphabet train" -- a series of boxes, each labeled with a letter of the alphabet and containing an object that starts with the appropriate letter. Joe asks Blue what should go in the "Z" box, and Blue leaves three paw-print clues to tell him what she thinks should go in the box (a stuffed-animal zebra). By the time you or your child is sitting in the "thinking chair" and trying figure out the puzzle, you have the following clues:

1. You know it goes in a box.
2. You know it starts with the letter Z.
3. You know it involves the color white.
4. You know it involves the color black.
5. You know that every time the clues are shown, Blue's Clues plays a snippet of African drum music that's unique to this episode.

That's more than three clues. Only three came with paw prints, but the others are still clues. The show regularly (and cleverly) tucks additional clues into its program. Sometimes there are musical clues, like the one above, or there are pictures hanging on the wall in the background related to the puzzle's answer. But there's always more than three clues for the viewer to think about.

What's this have to do with writing?

Well, from time to time, students will complain to me about writing or reading assignments, saying that we college folk seem to like stuff that's really long-winded. Last term, a student asked me, point-blank, "Why not just come right out and say what you mean? I don't think it's necessary to go on and on about it."

But if I'd done that with my observation about Blue's Clues, you might have rejected my point entirely -- it's a strange point, and without an explanation looks like it must be wrong. To make my point clear -- or even remotely acceptable -- I had to explain what I meant. I had to give an example.

Now, if I'd said, "Blue's Clues has three clues," I wouldn't have needed to explain much. It's a fairly obvious point and not immediately controversial. What's to explain? It can be stated in a single sentence and left at that.

The same dynamics apply to thesis statements in student papers or assigned readings.

If your thesis is familiar and noncontroversial ("Murder is wrong!"), writing three pages on it probably feels strange. It should.

On the other hand, if your thesis is controversial or surprising ("Computer hackers are worse than murderers, and should be executed"), you need to do more than just "get to the point." The point isn't enough. You need to give reasons. (The computer hacker example is real, though Landsburg is playing Devil's advocate a bit, and having fun. Click the previous link to read his argument.)

Which brings us to two of the most significant problems in student argumentation:
  1. Some students write three pages about why murder is bad, and thus say way too much on something that didn't really need it.
  2. Others say we should execute computer hackers (or something similarly surprising), and don't explain sufficiently why they hold such a strange position.
The best writers, conversely, stake out a thesis that requires defense and then engage in what I like to call high-definition writing.

To understand what I mean by "high-definition writing," imagine watching Avatar in high-definition 3D, and compare that experience to watching the same movie on a small black-and-white television with poor reception.

They both have the same plot. They're the same movie.

Would you get the same value out of each of them?

No, of course not. As a skilled film viewer, you'd probably prefer the high-definition experience. It enables you to immerse yourself in James Cameron's vision much better. The tiny, fuzzy, black-and-white image has too much distortion -- sure, you can follow the plot okay, but you're probably going to be distracted by all the technical issues.

That scenario is very similar to two essays about executing computer hackers: a low-definition one that makes its point, but leaves the reasons murky or muddles them with distracting errors; and a high-definition essay that states reasons clearly, provides examples, anticipates objections, and is edited closely enough that we're not distracted by errors. Both papers say we should execute hackers, but one of them is far more likely to persuade us that the writer isn't ... well, insane.

High-definition writing enables us to immerse ourselves in the author's strange, personal world without being distracted by technical issues. We emerge at the end of the experience having seen the world differently, having seen the world through another person's eyes. We can go to our friend and represent your point of view accurately, even if you're not there to speak up for yourself.

In short, while Blue's Clues highlights three clues for its viewers, I prefer to emphasize just two clues for student writers:
  1. Practice coming up with points of view that require defense.
  2. Practice defending them, perhaps by getting into friendly "Devil's Advocate" arguments with friends.
Of course, just like Blue's Clues, I might have embedded a few other hidden clues along the way.

Saturday, January 1, 2011

The Value of Duh

A recent article on is titled "Duh! The Most Obvious Scientific Findings of 2010."

Although writer Jeanna Bryner makes some unexpected moves in the article that I find refreshing, acknowledging that some of the "duh" studies she describes might actually have some value, the tone of the headline disturbs me, in part because it echoes a message I've been hearing a lot lately in radio commentaries, editorials, and student papers.

The gist of the message is, "Those silly scientists! They could have just asked me, and I would have saved them a lot of time and money!"

The people sending these messages are often missing the message themselves.

Let's set aside, for the time being, the very real possibility that the journalists are only focusing on the headlines and are missing critical, important, helpful details in the deeper parts of the studies they're discussing. That's often the case, but even if it weren't, the commentators are missing the point.

The point is this: There is value to "duh" studies.

Look at it this way: Every time someone releases a study that surprises us by coming up with an unexpected finding, we pay respect because they've taught us something new.

But the only way they could arrive at an unexpected conclusion is to test the "obvious" stuff we think is right. Scientists learn not to trust instincts. What we think is right is often wrong. So they test everything.

When researchers come up with something surprising, we are illuminated.
  • We discover that all objects fall at the same speed, no matter what their weight is, even though many people (even today) would never predict that.
  • We discover that time passes differently for satellites than it does for people on the ground -- you'd think, if you go with "obvious" instincts, that a minute is a minute is a minute. But no, the satellite's minute isn't the same as yours -- we actually have to make regular adjustments to satellite clocks or else your GPS and cell phones will stop working.
But any search for the unexpected is by its very nature unpredictable: You don't know where the unexpected stuff is going to be. So you're going to hit some dead-ends, and end up verifying some couch potato's half-drunk observations. It's inevitable. And then he'll laugh at you -- as he reads about your work on the Internet, even though the Internet would never work at all if we only relied on our instincts about what is obvious.

Fine, you say. But why report the findings, if they turn out to be obvious?

That's a good question, and, as it happens, there are two good answers:

1. It's useful to researchers to know when the obvious stuff is right. As I noted before, scientists (and other sorts of researchers) learn after a while that gut instincts and "obvious" conclusions can be wrong. So it's reassuring every once in a while to learn, "Oh that assumption I've been making all these years is correct. It means the 30 articles I've written over my career are still possibly valid!"

2. Huge misunderstandings can develop when people don't report results.

Let's say 1,000 scientific studies have determined that, surprise, broccoli is good for you. They don't bother to publish the findings because, well, everyone knew that. Then scientist 1,001 comes along, and simply because incorrect results pop up randomly from time to time, he ends up with numbers that say eating broccoli is worse for your health than eating rocket fuel. Wow. He's wrong, but he doesn't know that. He reports it. Now the only study published on the health effects of broccoli says it's worse than eating rocket fuel. If the other 1,000 scientists had published their work, it'd be easy to take that weird finding with a grain of salt. But as far as the world knows, the only study done on the effects of broccoli shows it'll kill you stone dead.

This principle cuts both ways: It's important to report results, no matter how expected or unexpected they are.

Economist Steven Landsburg (the first freakonomist, before Freakonomics was ever written) wrote a column in Slate a while back in which he talked about minimum wage studies. Years ago, economists had decided that minimum wage increases must kill lots of jobs. It seemed obvious: There would be less money to go around -- you can hire 100 people for $1 each or 10 people for $10 each. And, in apparent support of that obvious conclusion, studies were sometimes published showing that increasing minimum wage reduced the number of jobs on the market.

But it turns out the impacts of minimum wage increases aren't so severe or obvious. Later, statistical analysis showed something suspicious about those previous studies: The findings didn't get stronger when sample sizes increased. If the pattern were real, they should. Economists eventually concluded that lots of their colleagues had been doing studies on minimum wage impacts, but throwing out the results when they didn't match expectations. Self-censorship led to confusion for a whole field.

Similarly, although there are very good reasons to think that alarms over climate change are legitimate and deserve attention, skeptics frequently argue that dissenting opinions frequently get squelched in official channels. If so -- if only articles that report on expected findings are getting published -- that's as bad as publishing only the unexpected would be. (The famous article by Naomi Oreskes, in which her analysis showed no disagreement with the established consensus on global warming in more than 900 scientific abstracts, may be a bad sign, viewed in this light.)

I'm not saying that the majority is wrong in the latter two cases, or that there's no such thing as a waste of grant money. Without doubt, there are studies that didn't deserve a penny. But they should be judged by their methodologies, not their conclusions.

In our age of rapidly disseminated information and promiscuous skimming of headlines, we need safe-text practices to keep us clean of memetic diseases. A good start is to be wary of any commentator who snorts with derision at a study simply because its conclusions are expected or unexpected.