What Seems "Natural", and to Whom, and Why?

Natural Reading?  OK.  Natural Teaching? Maybe not.
A couple of times on this blog I have suggested that "natural reading" should be an important component of literacy education.  I still think so, but I recently read an interesting scholarly paper that made me realize that "natural" is a pretty subjective term.  I should have known this, since "all natural" is a classic slippery phrase in food marketing, and maybe in some sense I did know it, since my own use of the term "natural reading" was always intended to be partly humorous.  After all, who would be in favor of unnatural reading?  But while I still think "natural reading" is a good idea, I'm less sure, after perusing this 2009 paper by Mckeown, Beck and Blake (Rethinking Reading Comprehension), about natural teaching.

Scholarly articles are a bit more objective (i.e. unnatural?!)
Articles about education aimed at a policy audience, or a popular audience, are often terrible, distorting, inflating or ignoring the often ambiguous data.  Nevertheless, the data, though ambiguous, are often worth looking at.  Science is supposed to be objective--and it is fairly objective, compared to the gross distortions that often come in work aimed at a popular or a policy audience. Reading the scientific literature, looking at the actual data, is a useful and interesting check on the received ideas and self-serving propaganda that you find in a lot of magazines or in publications like the National Reading Panel.

In the articles presenting the results of their experiments the same scholars often take a much more objective and moderate view than they do when they're writing for an audience of teachers.  Teachers are looking to be told what to do; scholars are looking for an argument (And then there are teachers like me). Writing for policy advocates or managers, you want to use the data to promote your favored policy; writing for teachers, you want to offer specific advice for things the teachers can actually do; but when you're writing for your fellow scholars, who are always looking to nitpick, because that's their job, you have to be somewhat more guarded in your assertions. (Being guarded and objective may not be natural, but it has its advantages!)

The paper
This post is about a 2009 paper by Isabel Beck and Margaret Mckeown, two big names in reading research and coincidentally the same researchers who back in the early eighties did the questionable research on vocabulary that is still being used to promote the idea that explicit vocabulary instruction increases reading comprehension.  The paper is about a study comparing two different ways of teaching reading, one a "content "approach" and one a "strategies" approach.  The study seems to show, interestingly, that the "content" approach is superior, but the paper is as interesting for its asides as for its data, which data are not, as usual, particularly conclusive.

Aside #1: We don't really know what works (or do we?)
The first interesting passage in this paper came in its introduction. Mckeown and  Beck are prominent, veteran researchers, and yet they paint a fairly grim picture of the current state of knowledge in reading research, essentially saying that very  little is known about the best way to teach reading.  This humble admission of ignorance, while not unusual in the sober scholarly literature on reading, is in striking contrast to the countless  books and magazine articles that offer specific advice to teachers on the explicit pretense that the advice is grounded in "the research", and Mckeown and Beck's humility also contrasts with such pretended authorities as the National Reading Panel and the "What Works" publications.  The professors who taught my workshop last week were wonderfully open about how unclear the research literature is, but many of the articles they provided were of the popular kind that I have come to see as fundamentally dishonest.

"The research on strategies and content approaches," Mckeown and Beck write, "provides little guidance on what in the instruction was responsible for the outcomes.  It could be the case that simply more time and attention to text is the key that leads to improvement."

Maybe spending more time and attention on text is the key.  Ya think? As a teacher who has spent way too much time doing my own talking and having my students either listen to me or do something that is not focused directly on the text, and as someone who has walked around my school for years seeing way more teachers talking than students talking, this last conjecture seems eminently reasonable: most of the time in most reading lessons is probably not spent either on actual reading or on students looking closely at the text and talking about it, so if you make teachers spend time actually looking closely at an actual text,  that might be expected to lead to more learning.

Amazingly enough, after presenting this eminently reasonable hypothesis, Mckeown and Beck immediately say, "We doubt that is the case." They don't explain the basis for their doubts, saying only, "it is more likely that some activities are more effective than are others" (222). Well, yes, some activities are probably more effective, but that is hardly any reason to doubt the hypothesis that spending time and attention on the text itself is important; perhaps those activities that focus student attention more sharply on the text are more likely to be effective.  But perhaps such a relatively simple hypothesis is too simple for Mckeown and Beck. Like the policeman in Poe's great tale, scholars are heavily invested in their complicated, time-intensive methods, and may respond to a suggestion that the answer is simple with anxiously incredulous laughter: ""Ha! ha! ha! --ha! ha! ha! --ho! ho! ho [...] oh, Dupin, you will be the death of me yet!"
  
The study itself
In any case, the study that the paper is primarily about compared two approaches to teaching reading, the "strategies approach" and the " content approach."  In a strategies approach, discussion of a text is a sort of meta-discussion, focusing primarily on which "reading strategies" might be useful in answering a given question about the text and only secondarily on the text itself.  In a content approach, student attention was focused on the content of the text through "meaning-based questions."  The study's results seems to show a "content" approach doing better than a "strategies" approach, basically because the strategies approach leads kids to focus on the strategies and not on the text itself, whereas a content approach encourages kids to think more deeply, and pay closer attention to the text. It seems to me that the content approach might be considered the more "natural."

The study worked with six classrooms in the same district, replacing one of the week's 5 regular reading periods (working in a basal reader) with a scripted lesson.  Two of the classrooms used a scripted lesson that focused on comprehension strategies; two classrooms used a scripted lesson based on discussing the content of the reading; and two of the classrooms used a scripted lesson based on the questions included in the basal reader.  The students who had discussions focused on content performed somewhat better than students in the the other classrooms on such tasks as recall and offered LONGER answers to discussion questions.  This is an interesting result, since it suggests that asking students to talk about what happened in the text is just as helpful, and possibly more helpful, than offering explicit instruction in HOW to talk about the text.  As with other research I've looked at, this study provides little support for explicit skills instruction in English class.  But the most interesting part of the paper was a result that the authors presented almost as an aside.

Aside #2: Natural teaching may not be natural learning
Although the content approach looks to me to be the more natural one, the teachers in the study didn't experience it that way.  When asked, "How natural did the approach feel?", the teachers who used the strategies approach were happier with their approach, saying that the strategies approach felt "very natural." Amazingly enough, the content approach teachers said that the content approach felt less natural.  One teacher said that it wasn't "natural at first [...] I always wanted to put my two cents in." The other teacher said, "It's not natural to not go deeper. It's hard to just let them think on their own and not pull the information from them."

What's interesting about this is that teachers do not "naturally" use a natural approach.  Instead, what seems natural to a teacher is to put her own "two cents in," instead of asking questions to elicit the students' own thinking. To a teacher, to go "deeper" apparently means to "pull the information from them." The strategies approach, on the other hand, felt natural to the teachers who used it, perhaps because that approach's explicit skills instruction allowed the teacher to feel she was putting her own "two cents in."

One reasonable explanation for the teachers' feelings is that teachers like to, well, teach.  They are teachers, after all.  But more teaching does not always mean more learning.  What's natural to the lion may not seem so natural to the zebra.

Conclusion: we may have to work unnaturally hard to foster natural reading and learning.
Natural reading, and natural learning, do not necessarily happen naturally.  School is an unnatural environment, and to create natural events in an unnatural environment probably means hard work.  That's only natural.