To Model or Not To Model: That Is the Question

In addition to the numerous treats I shared from this year’s NCTE convention, I also had the privilege of hearing Ellin Keene talk about talk—specifically about what kind of teacher talk enhances or impedes student understanding. Drawing on some of the work from her most recent book Talk About Understanding, she shared some trends and patterns she’d noticed during a year she spent viewing and analyzing teachers’ talk in classrooms. Among the things she noticed and named that all too often we do were the following:

• Cut students off before they have a chance to fully develop their thinking
• Accept students’ first thoughts without probing for deeper thinking
• Move on before we label students’ descriptions of thinking (i.e., naming for them what they’re doing) so that the thinking can be transferred
• Segue from modeling to student responsibility too quickly

The first three points I see all the time—and have been guilty of doing myself. And seeing them named so clearly reminds me of both the power of naming and the importance of giving students enough time to develop and test out their thinking. But the last point made me pause, because increasingly in my own practice, I’ve found myself moving away from explicit modeling in reading.

As Dorothy Barnhouse and I both noticed and discussed in What Readers Really Do, when we model how readers use strategies through a think aloud, what students too often take away is what we thought, not how. And they can be left (as I sometimes am in the wake of a great think aloud) feeling dazzled but daunted. Additionally, a mini-lesson based on a “Today I’m going to teach you” teaching point, followed by a “Now watch me do it” demonstration and a “Now you do what I do” link puts students in a passive role and re-enforces a vision of student as empty vessels in need of teacher filling.

In his great book on teacher talk Choice Words, Peter Johnston shows how this positioning can have even more consequences, which he describes as the “hidden costs in telling people things”:

“If a student can figure something out for him- or herself, explicitly providing the information preempts the student’s opportunity to build a sense of agency and independence . . . When you figure something out for yourself, there is a certain thrill in the figuring. After a few successful experiences, you might start to think that figuring things out is something that you can actually do. Maybe you are even a figuring out kind of person . . . When you are told what to do, particularly without asking, it feels different. Being told explicitly what to do and how to do it—over and over again—provides the foundation for a different set of feelings and a different story about what you can and can’t do, and who you are.”

For Johnston, the key to learning isn’t explicit teacher modeling but student engagement. And from 2008 to 2010 he was involved in a research study that yielded compelling proof of that. As he shared in a recent blog post titled “Reducing Instruction, Increasing Engagement,” he and his colleague Gay Ivey looked at four 8th grade classrooms where the teachers decided to put engagement front and center by cutting back on explicit instruction and modeling and offering instead edgy young adult fiction with adolescent-relevant themes that the students could read without strings attached—i.e., no book reports or quizzes. The results? In addition to reading an average of 42 books a year and becoming more open, responsible and empathic, the students increased their standardized test scores, in some cases by more than 10%. (A paper on the study can be found here.)

In our current age of anxiety, however, where accountability and data collection rule, it’s hard to image this study being replicated in any systemic way. But what if instead of modeling, we moved students more quickly into problem-solving mode by orchestrating experiences for them that positioned them to feel the thrill of figuring things out?

This was what I did the other day in a fourth grade bilingual room that was embarking on a thematic unit of study about overcoming adversity. These were students who could easily be seen as deficient—who ‘couldn’t’ infer, ‘couldn’t’ summarize, ‘couldn’t’ find the main idea. But as we began to read Yangsook Choi‘s The Name Jar, without a shred of modeling and no more support than a chance to turn and talk and a T-chart to record what they were learning and what they wondered about, their thinking was amazing.

From the cover, they wondered what a name jar was, why the book was called that, who put the names in the jar and why, and was the girl putting something in or taking something out? With these questions in mind and their curiosity sparked, I started reading, pausing periodically to let them turn and talk and share out what they were thinking out.

What they noticed was that on almost every page, something about names came up: the girl’s grandmother gives her a wooden name stamp when she leaves Korea; children on the bus make fun of her name; she lies about her name to her classmates; the Korean grocer says her name is beautiful; and she tries out various American names as she brushes her teeth. They also had two more burning questions: Will she decide to change her name? and Will she manage to make friends?

As they wrestled with these questions half-way through the book, they demonstrated a deep understanding of the girl’s predicament in a way that also showed their ability to refer to details when explaining what the text said explicitly and when drawing inferences from it (Reading Literature Standard 4.1) and to draw on specific details from a text to describe in depth a character or event (RL. 4.3). They were also well on their way to determine a theme of a story from details in the text (RL.4.2)—and none of that had been explicitly taught or modeled (though I did ask them to share what made them think what they did and ended by naming the work they’d done).

It’s possible, of course, that what allowed them to do this was the explicit modeling their teacher had done. But what if, as Johnston and Ivey conclude of the students in their study, “Being fully engaged and facing problems, they became strategic”? What if they automatically generated strategies because they were invested in what they were reading, not because someone told them that’s what good readers do? And what if in delaying the release of responsibility, we risk becoming helicopter teachers, hovering over our students heads to make sure they get it right in a way that deprives them of the opportunity to learn by their mistakes?

For the record, I do keep explicit teaching and modeling in my toolkit of teaching moves. But it’s not automatically the tool I first pull out, because sometimes less is more.

Superficial or Significant: The Challenge of Comparing

When a friend and colleague heard I was going to Las Vegas for NCTE so soon after being in Reggio Emilia, she thought it might be interesting for me to compare the two places. My initial thought was no, that’s too easy. The light, the noise level, the language—all different. The money, the history—all different as well, with Las Vegas, as we know it, a virtual newborn in the span of human time and some buildings in Reggio standing in place for more than one thousand years.

But then I thought of quote another friend and colleague recently sent me from Ron Ritchhart, Mark Church and Karin Morrison’s book Making Thinking Visible. Here the authors take a look at skills and thinking, like comparing, that appear in classification charts such as Bloom’s Taxonomy and Webb’s Depth of Knowledge, and they offer this advice:

“Rather than concerning ourselves with levels among different types of thinking, we would do better to focus our attention on the levels or quality within a single type of thinking. For instance, one can describe at a very high and detailed level or at a superficial level. Likewise . . . analysis can be deep and penetrating or deal with only a few readily apparent features.”

With Las Vegas and Reggio, I think I was simply ticking off “readily apparent features” without being terribly insightful, just as I described many students doing in last spring’s post on the limits of graphic organizers. Of course, sometimes a student will come up with something that does seem “deep and penetrating.” But I don’t think we always teach toward that, aiming instead at just teaching the skill without that attention on quality. Or put another way, we teach the concept of comparing without teaching the concept of significance.

The Common Core Standards, however, have dramatically upped the ante in ways that I think are important. In the case of comparing, for instance—a.k.a. Anchor Reading Standard 9—the focus should be on significant, not superficial, comparisons. But how can we instructionally help students move beyond what’s readily apparent to what’s more penetrating but often less visible—a step which often requires readers to look beyond the specifics of any one text to something that’s more abstract and general? Thinking about this, I’ve developed a theory that, when comparing, it’s often useful to focus exclusively on similarities between two things or texts that, on the surface, seem different, and explore differences when similarities are more apparent. Then once those have been mapped out, the next step is to dig into the differences within the similarities or the similarities within the differences.

I tested this theory out last spring with a group of middle school teachers who had gathered for two days to explore ways of helping students read complex nonfiction texts on a common topic or theme. To make this concrete, I asked them to read an excerpt of Philip Hoose‘s Claudette Colvin: Twice Toward Justice, which combines transcripts of interviews with Colvin with more expository text, using a text-based Know/Wonder chart to see how it could help students connect details within the text (e.g., figure out why the number ten was detested, which is mentioned on the first page below).

Then we read an excerpt of Ann Petry‘s biography Harriet Tubman: Conductor on the Underground Railroad, which appears in the Standards Appendix B as a middle school informational exemplar text. Here’s the beginning of the excerpt:

Rather than handing out Venn Diagrams, I asked the teachers to take out their notebooks and jot down as many similarities they could think of or patterns that recurred across the books, without judging any of their ideas—that is, nothing should be deemed too obvious or, conversely, too far-fetched. This helped them move beyond the most apparent similarities that both books were about African-American girls who as children experienced inequality based on race, to more insightful noticings such as these:

• Both girl’s parents were addressed by their first name by white people.
• Both girls learned lessons about the social structure they lived in very early in life.
• The social structure was enforced through threats of violence, insults and humiliation.
• Both girls felt fear, uncertainty and confusion.
• Both girls saw the adults around them afraid.
• Both girls were expected to take responsibility for something that was done to them, not by them.
• Neither girl’s parents could protect them.
• Both girls felt that there were unstated rules “in the air”.

As these were shared, I invited teachers to add ideas they hadn’t thought of before to their list. Then I asked them to look at their expanded list and think about which similarity seemed the most  important or significant to them and on another page of their notebook to briefly explain why. Using another think-to-write strategy, the Write-Around, from Harvey Daniels, Steven Zemelman and Nancy Steineke‘s Content Area Writing, I then had everyone pass their notebook to the person on their right. That person would then read what the other person wrote then write a quick response, extending, commenting, questioning, and probing what the writer before them had said, before passing the notebooks yet again to the right.

After several passes, the notebooks were returned to their owners who were eager to see how their original thinking had traveled and evolved. And at that point, they felt they would be prepared to have a more formal discussion or even to begin planning out a piece of writing. But perhaps, most importantly, they saw how this process could help lift their students’ thinking beyond the obvious or the superficial in ways that would help them, not just meet the Standards, but understand the undercurrents of a topic in that deep, more penetrating way.

Which brings me back to Vegas and Reggio. After giving myself some time to brainstorm, I did come up with something that was similar and more significant than the fact that both cities had two-word names that were often shortened to one. Both cities revolved around public spaces where people congregated and socialized. In Las Vegas, it was the casinos; in Reggio, the piazzas. And what seemed different within this similarity was the purpose of those spaces. In Reggio the piazzas helped the community connect and strengthen their social bonds, while the casinos were there to make money—with visitors like me forced to walk through the casinos just to get water or coffee.

These differences led to a final similarity: The purpose of these spaces reflected the cultural values of each of the cities, with those values again being different. Anyone want to place a bet on which one I liked best?

 

A Close Look at Close Reading

As teachers and schools continue to wrestle with implementing the Common Core Standards, I hear more and more talk—and more and more questions—about the term ‘close reading’. Interestingly enough, the term doesn’t appear in the actual Standards, though it crops up repeatedly in many Standards-related material, including the now famous—or infamous—videos of Standards author David Coleman dissecting Martin Luther King’s “Letter from Birmingham Jail.” And Text Complexity co-author Douglas Fisher has said that close reading is “the only way we know how students can . . . really learn to provide evidence and justification,” as the Common Core requires.

So what exactly do we mean by ‘close reading’? According to Timothy Shanahan, who’s become something of a spokesman for the Standards, close reading is “an intensive analysis of a text in order to come to terms with what it says, how it says it and what it means.” I agree completely that close reading allows a reader to understand what a text says and what it means, with what it means directly related to the author’s decisions about detail and language and structure—i.e., how it says what it says. But for me, analysis is an off-shoot of close reading, something I can produce, if I’m asked to do so, after I’ve read closely.

I think this because, by definition, analysis involves thinking about how the parts contribute to the whole, which presupposes an understanding or vision of the whole. Putting analysis in front of understanding seems a bit like putting the cart before the horse. And asking students through a text-dependent question to analyze a part before they’ve had a chance to consider the whole risks putting them in the position of the blind men in the old Indian tale who sought to understand what an elephant was by attending to its parts. One man touched the trunk and thought an elephant was like a snake; another felt the tail and concluded it was like a rope; while a third stroked the ear and thought it was a fan. None was able to make sense of the whole when asked only to consider a part.

My own vision of close reading is better captured in some of the guidelines colleges provide students. The Purdue Online Writing Lab, for instance, advises ‘tracking’ your understanding of a text through margin notes that often consist of questions, with an example that bares more than a passing resemblance to the kind of questions that come up when students are using a Know/Wonder chart, noticing patterns across a text, and wondering what the writer might be trying to tell them through the details he’s chosen.

Example of close reading annotation using Doris Lessing’s short story “A Woman on a Roof,” from the Purdue Online Writing Lab

Harvard also provides a “How to Do a Close Reading” guide to students, which breaks close reading down into a two-part process: First the reader observes facts and details in the text, then he interprets what he’s observed through inductive reasoning—that is, he builds an interpretation bottoms-up from the details, rather than by deductively starting with a claim and then finding evidence to support it. And they offer the following tips, which sound similar to the kind of thinking the fifth graders I described in a recent post engaged in (with the teacher transcribing their thoughts in lieu of annotating the text):

1. Read with a pencil in hand, and annotate the text, noting anything that strikes you as surprising or significant, or that raises questions.

2. Look for patterns in the things you’ve noticed about the text—repetitions, contradictions, similarities.

3. Ask questions about the patterns you’ve noticed—especially the how and why.

This two-pronged process has always seemed to me a lot like the scientific method. The reader attends to the details an author gives just as a scientist attends to the details of whatever phenomena he’s studying. And from those observations, each develops a hunch that attempts to explains what they’ve noticed, which in science we call a hypothesis. Then just like the scientist, the reader continues to probe and observe, testing her hunch out as she encounters new details and looks back on ones she’s read, revising, refining and developing her ideas until all the pieces fit—at which point she comes to a final understanding, which is like a scientist’s theory. Only then, I would argue, can the reader’s thinking be turned into a claim whose validity can be proved in a deductive fashion using many of the same details that helped her understand as evidence.

Unfortunately, however, some of the approaches that aim to support close reading rob students of the opportunity to notice and to develop ideas of their own—which, as Harvard says, ”is central to the whole academic enterprise.” Take Achieve the Core’s 8th grade Close Reading Exemplar for “Long Night of the Little Boats” by Basil Heatter, which recounts an incident from the Battle of Dunkirk when a ragtag flotilla crossed the English channel to rescue soldiers who were stranded on a beach during World War II.

My hunch is that the exemplar writers followed a process similar to Harvard’s to arrive at their own understanding of the piece (noticing, questioning, and interpreting, perhaps, automatically in their heads). They then rephrased their understanding as a question for the final writing task: “How did shared human values, both on the part of the little boat rescuers and the soldiers, play a part in the outcome of Dunkirk?” With that in place they then designed a series of questions and steps that would focus the students’ attention on details that were key to their own understanding’s development, such as:

The students neither own the noticings here, nor the development of the ideas. And the ‘help’ that teachers are asked to provide in order that students ‘see’ what they’re supposed to runs the risk of being as much an act of spoon-feeding as some of the pre-teaching practices that have come under fire are. Of course, it does increase the likelihood that students will meet the Standards. But they’ll do so by plugging in someone else’s language about details someone else has noticed to support an idea someone else has formulated. And that’s a far cry from the independent thinking that colleges want students to have.

To support that kind of independence, we have to design instruction that engages students in both components of the close reading process: to first be observers and questioners and then to use their observations and questions to, as Harvard puts it,  ”reason toward our own ideas.” That may, indeed, involve asking students questions, but those questions need to be open enough for students to engage in real close reading, not an overly-prompted knockoff.

And so to ensure that we don’t put the cart before the horse, let’s remember this when it comes to close reading:

Questions before Answers

Hunch before Claim

Understanding before Analysis

Some Questions about Text Dependent Questions

As the school year finally begins to wind down here in New York City, a new term is the air: text dependent questions. I first encountered the term in the Common Core Standards Publishers Criteria, which recommends that Standards-based instructional material includes a sequence of “rigorous text dependent questions that require students to demonstrate that they not only can follow the details of what is explicitly stated but also are able to make valid claims that square with all the evidence in the text.” And now Student Achievement Partners, the group founded by several of the Common Core authors, has issued a “Guide to Creating Text Dependent Questions” along with an ever-growing number of “Close Reading Exemplars” that show this method in action.

These text dependent questions stand in contrast to some of the common kinds of questions often heard in classrooms, such as questions about students’ own feelings or experiences and questions related to strategies or skills, like “What’s the main idea?” I agree that these kinds of questions are problematic and should be used sparingly. The first kind can shift students’ attention away from the text to their own thoughts, while the second can turn the act of reading into a scavenger hunt, as I explored a few weeks ago in my post on basal readers.

But text dependent questions seem problematic, as well. The Student Achievement Partners’ guide says that text dependent questions aim to “help students see something worthwhile that they would not have seen in a more cursory reading.” This is a goal I completely share. But the text dependent question approach relies on teachers directing and prompting students to what they want them to see, not on teaching in a way that empowers students to more independently notice what there is to be noticed through their own agency. And in this way text dependent questions run the risk of creating teacher dependent students instead of strong, flexible readers.

To see what I mean, let’s look at one of the Close Reading Exemplars from the Student Achievement Partners’ Achieve the Core site. Here eighth graders are asked to dip into a passage from Narrative of the Life of Frederick Douglass an American Slave, Written by Himself, which begins like this:

Like all the Exemplars, this one asks students to first read the passage silently to themselves, without any introduction or instruction. They then follow along for a second go through as the teacher reads the text aloud in order to offer “all students access to this complex text.” Then the questions start:

This read-listen-then-answer-questions sequence seems to almost guarantee that some, if not most, students will read and listen to the passage passively, waiting for the teacher to tell them what to do. It also seems to mirror standardized tests, where students don’t often begin to think until they hit the questions, rather than the moment they first begin to read.

The questions themselves also seem test-like; you can almost imagine them being followed by a choice of four possible answers. That’s because there seems to be one right answer, and the questions are seeing if you ‘got it’ or not. In this way, the questions are assessing comprehension, not helping students build it, which means that students who are able to comprehend will probably do fine, while those who can’t, will not. And one can only imagine how those answers might be pulled and yanked like a tooth from those struggling students through continued prompting.

But what if, instead, we taught students that every reader enters a text not knowing where it’s headed, and because of that they keep track of what they’re learning and what they’re confused or wondering about, knowing that they’ll figure out more as they both read forward and think backwards? This vision of what readers do acknowledges that reading is just as much a process of drafting and revising as writing is, with readers constantly questioning and developing their understanding of what an author is saying as they make their way through a text. And it supports the idea that readers are actively engaged and thinking about how the pieces of a text fit together, beginning with the very first line.

To make this process more visible to students, Dorothy Barnhouse and I developed our text-based Know/Wonder chart. Depending on students’ familiarity with the chart, we might briefly model how we use it in a way that encourages students to acknowledge their confusion by reading the first two sentences and noting the following:

Students who had noticed the title, might say that the narrator was a slave, which would help answer the first question and also raise a lot more, including how a slave got to be friends with white boys; where, exactly, was this taking place; how old is/was the narrator; and, as they read further on, how did he manage to get a book and was he allowed to take the bread or had he stolen it.  Reading forward on the lookout for answers to these student-generated questions, the students would pick up clues that engaged them in considering the third text dependent question about how Douglass’s life as a slave differed from those of the boys. And those students who hadn’t caught the title could hold on to the question, made visible by the chart, until later on in the passage where they’d encounter more clues. And at that point they’d need to think backwards to revise whatever they’d made of the text so far in light of this realization.

Thus, all this could happen the first time the students read the text with virtually no teacher prompting, because they’d be reading closely from the get-go, fitting details together like puzzle pieces to see the larger picture they revealed. And doing so without any prompting would contribute to an increase in both their engagement and their ability as readers. It would also be an experience they could transfer to the next complex text they read.

Additionally all this drafting and revising would eventually enable students to “make valid claims that square with all the evidence in the text,” in a much more independent way than the text dependent question method permits, because so much more of the thinking is theirs. So let’s not jump so quickly on the text dependent question bandwagon and consider, instead, making the process of meaning making more visible to our students, by offering instruction not directions and giving them time to practice–and perhaps remembering that asking a question doesn’t constitute teaching, nor does answering one always mean learning.