Looking at Complex Texts More Complexly (or What’s Wrong with this Picture?)

Clifford Loves Me -SunAlsoRises

By now many of us have experienced or heard about the effects of using Lexile levels as the sole arbiter of text complexity. In her wonderful post “Guess My Lexile,” for instance, Donalyn Miller looks at the absurdity of putting book with widely different reader appeal and age appropriateness in the same book bin because they share a Lexile level (as my own favorite Lexile odd couple, Clifford and Hemingway, do, with both clocking in at 610L). And for those of us who strongly believe in the power of choice and interest-based reading, young adult writer Mike Mullin shares a chilling story in a blog post about a mother frantically searching for a book that her dystopian-loving 6th grade daughter, whose Lexile level was 1000, would be allowed to read for school. The Giver—out. Fahrenheit 451—out. Margaret Atwood’s The Handmaid’s Tale—out, all because of Lexile levels which, in its arbitrariness and control, seems like something out of those dystopian books.

text complexity triangleWhile I can’t vouch for the intentions of the Common Core authors (as I can’t for any writer without direct communication), this is not what’s stated in the Standards themselves. In Appendix A’s “Approach to Text Complexity,” the Common Core authors offer a three-part model for measuring text complexity, which they capture with a now familiar graphic. This model, they clearly state, “consists of three equally important parts”—the qualitative dimensions, the quantitative dimensions, and the reader and the task—all of which must be considered when determining a text’s complexity in order to address “the intertwined issues of what and how students read.” Yet how often does that actually happen?

The Arrival coverThe sad fact is that too many schools, reading programs and test makers rely on quantitative measures such as Lexiles to make text selections for students because it’s simple and easy. Lexiles can be found with a click of a mouse, while assessing the qualitative measures is harder and much more time consuming, even when we use rubrics. That’s because the rubrics are often filled with abstract words that are open to interpretation, and they use what seems like circular logic—e.g., saying that “a text is complex if its structure is complex—which doesn’t seem terribly helpful. And how do you deal with a wordless book like Shaun Tan‘s The Arrivalwhich I recently explored with teachers from two schools that were looking at text complexity? Ban it from classrooms because, without words, there’s nothing to quantitatively measure?

Like other short cuts and quick fixes I’ve shared, dismissing a book like The Arrival, based on a non-existent Lexile level, risks short-changing students. The book requires an enormous amount of thinking, as the teachers I worked with discovered. And interestingly enough, their thinking mirrored that of the students of fourth grade teacher Steve Peterson, who wrote about his class’s journey through the book on his blog Inside the Dog. Both the fourth graders and the teachers had to make sense of what the author presented them by attending carefully to what they noticed and what they made of that. And while some of the initial ideas they came up with were different (the teachers thought the portraits on the page below were of immigrants, not terrorists, as some of Steve’s kids first did), the process was the same.

TheArrivalFrontispiece

Both students and teachers had to constantly revise their understanding as they encountered new details and images that challenged or extended their thinking. And both debated the meaning of certain details in very similar ways. The teachers, for instance, argued whether the dragon-like shadow that first appeared in the picture below was real or a metaphor for something like oppression, while in a second post, Steve recounts how his kids debated whether the bird-like fish that appear later in the book were real or a metaphor for wishes.

TheArrival6

The teachers only read the first part of the book, after which I passed out the rubric below, which many states seem to be using, and asked them how they’d qualitatively assess this text. Being wordless, the text couldn’t be scored for its Language Features, but for every other attribute on the rubric—Meaning, Text Structure and Knowledge Demands—the teachers all decided it was very complex, especially in terms of meaning.

Literary Text Complexity Rubric

If we give equal weight to both the qualitative and quantitative dimensions of this text, we have to say that even with a zero Lexile level, it’s at least moderately complex. And what happens when we add in the Reader and the Task, which sometimes feels like the forgotten step-child in text complexity discussions?

Steve and I used the text for different purposes—Steve to launch a unit on immigration, me for a workshop on text complexity. But we each set up our readersNCTE Logo to engage in critical thinking, which the National Council of Teachers of English defines as “a process which stresses an attitude of suspended judgment, incorporates logical inquiry and problem solving, and leads to an evaluative decision or action.” Both the teachers and students engaged in this process not because they’d had a lesson on suspending judgment or logical inquiry, but because they were curious about what the writer might be trying to show them. And to answer that question, both the students and the teachers automatically and authentically engaged in the work the Common Core’s Reading Standards 1-6.

Unfortunately many of the tasks we set for students aim much lower than that, including some of those found in the Common Core’s Appendix B, such as the following:

Students ask and answer questions regarding the plot of Patricia MacLachlan’s Sarah, Plain and Tall, explicitly referring to the book to form the basis for their answers. (RL.3.1)

Students provide an objective summary of F. Scott Fitzgerald’s The Great Gatsby wherein they analyze how over the course of the text different characters try to escape the worlds they come from, including whose help they get and whether anybody succeeds in escaping. (RL.11-12.2)

Each of these tasks are aimed at a particular standard, and frequently the instruction that supports them (plus the worksheets, graphic organizers and sentence starters) focuses the students’ attention on that single standard, rather than on a more holistic way of reading, which would naturally involve multiple standards. And while the Gatsby task is certainly harder than the third grade one, the prompt takes care of the hardest thinking by handing over a central idea instead of asking students to determine one.

But what if the reading task we set for students in every text they read is to think critically about what the writer is trying to explore or show them, through the details, story elements, word choice, structure—all those words that litter the Standards. Wouldn’t that, in addition to a complex qualitative measure, off-set a high Lexile level, if all three truly held equal weight?

I’ll share more thoughts on the reader and the task in an upcoming post. But for now I can’t stop thinking that if instead of ramping up the complexity of texts, we ramped up the complexity of thinking we aim for—trading in, say, some of the hardness of texts for deeper and more insightful thinking—we might, in fact, prepare students better for colleges, careers and life.

Preparation of Life Quote

It Was the Best of Times, It Was the Worst of Times: Some Reflections on the Year

Illustration from A Tale of Two Cities by Charles Dickens

Between teacher effectiveness rubrics, performance-based tasks and text complexity bands—not to mention testing scandals and the growing concerns about the privatization of public education—it hasn’t exactly been an easy year. Yet when I give myself some time to look back, what strikes me most is how much I learned. And that learning helped to balance out the challenges of the year.

So what did I learn? For one, I learned that I can sometimes be wrong, which is always good to know. In this case, I was wrong about the nonfiction performance-based tasks the New York City Department of Ed required every teacher in the city to implement as part of their drive to bring schools up to speed on the Common Core. As someone who cut her teeth at the Teachers College Writing Project, I’ve always believed that the best writing comes from a process that gives students time to draft and revise with feedback from both teachers and peers. And so I questioned the ‘on demand’ aspect of the tasks. Also, the sample text-sets and tasks, which came to be known as ‘bundles,’ that the DOE posted online seemed a little too test-like to me, with administration guidelines and actual scripts like those found in standardized test packets.

I also worried that yet again the emphasis was being placed on assessment not instruction, which seems problematic to me. But here’s where I was wrong. While some teachers chose to use the DOE ‘bundles,’ many designed their own tasks as a final assessment of a meaningful content unit that was already on their curriculum. They did this by setting aside one last aspect of the unit topic for students to read and write about on their own, without the same level of scaffolding they’d provided throughout the unit. Second graders, for example, who’d been studying plants and learning to write All About Books, were asked to read two final pieces about carnivorous plants then write an information piece on demand to share what they had learned. And two impassioned first grade teachers extended a unit they’d developed that combined a study of social activists with writing reading responses by having students listen to one last book, Wangari’s Trees of Peace  by Jeanette Winter, about the Kenyan environmentalist Wangari Maathai, and then write a response. And, as can be seen, the results were stupendous (though I do think they’re a testament to the thoughtful, well-planned instruction that proceeded the task more than the assessment design):

I also learned much about reading nonfiction, which I dove into deeply this year to help the schools I work with make the first two Instructional Shifts required by the Standards. Of course, I’d ‘done’ nonfiction before. I’d taught students how to use text features to both anticipate the information they’d encounter and locate facts they might want to use for the nonfiction pieces they were writing. And I’d brought in feature articles and creative nonfiction books like Atlantic and Bat Loves the Night for students to study as mentor texts to learn about structure and craft.

© 2012 D. A. Wagner, http://dawagner.com

But I hadn’t thought much about what readers really do to comprehend and understand nonfiction. And so I tried to do what Dorothy Barnhouse and I did when we explored the reading of fiction in What Readers Really Do: “peer into the recesses of our own reader’s mind, attending to the work we do internally that frequently goes unnoticed or that happens so quickly that it feels automatic.” I also studied some of the Standards’ exemplar texts to see what sorts of demands they put on readers in order to better understand what students might need instructionally to read these kinds of texts. And for better or worse, I discovered that much of what passes as conventional wisdom about teaching nonfiction reading, like the practices listed above, don’t always help students move from plucking facts to deeply understanding what they read.

I’ll be sharing more specifics about reading nonfiction over the next few months, along with more of what I learned as I helped teachers implement a second Author Study unit in the age of the Common Core. But I’ll also be taking some time off to recharge my batteries and reconnect with myself as a reader and writer, which may mean not posting quite so frequently. In addition to finally getting to the stack of books sitting on my nightstand, I also plan on spending time reading new children’s and YA books and on joining write Laurie Halse Anderson, author of Speak and Fever 1793in her annual “Write Fifteen Minutes a Day Challenge,” which she hosts in August.

I also want to update this blog to include a list of the wonderful blogs I discovered this year. For this is something else I learned: There are so many smart, dedicated thinkers among us, putting themselves out there week after week, raising questions we all need to consider, sharing their invaluable resources and experiences, and making me, for one, feel less alone. They’ve taught me much in this challenging year that I’ll be mulling over as I sit beneath my own tree that grows in Brooklyn and reap the joys of a literate life.

Illustrations from A Tale of Two Cities by Charles Dickens

All Inferences Are Not Created Equal

Here in New York City, we’ve been thinking a lot about text complexity, especially about what makes a text complex. School networks have traded professional articles. Consultants have helped teachers create rubrics to assess the degree of complexity in a text. Yet oddly enough there’s been far less discussion about how we can instructionally support students to meet the demands of those texts.

My own sense is that, beyond denser print and more complicated syntax, text complexity is directly linked to how a writer conveys information, with complex texts revealing more information—about everything from the characters and themes to shifts in time and setting—indirectly. And this means that in order to access that information readers need to infer.

Given how critical inferring is, you’d think we’d have a boatload of strategies up our sleeves to help students do it. But all too often we rely on a variation of “It Says, I Know, and So,” which asks students to connect something in the text to their own experience or prior knowledge in order to infer what the writer might be saying indirectly. This will, indeed, work some of the time. But it doesn’t always work because inferences aren’t actually all made the same way and no single strategy will do.

To illustrate this, let’s go back to the text I shared in last week’s post, Tony Johnston’s Any Small Goodness: A Novel of the Barrio, which I used with a small group of students who struggled with reading increasingly complex texts. The book is dedicated to “Everyone who gives up a part of himself,” and the first chapter is titled “American Names.” Here’s the passage from that first chapter that gave the students so much trouble:

The passage requires an enormous amount of inferring to simply figure out what’s happened, let alone to start thinking about why. When the students first read it, they were able to infer that Miss Pringle was the teacher by connecting what it said in the text to what they knew from their own lives—i.e., that teachers usually have titles like Miss and often address students as ‘Class’. But the same “It Says, I Know, and So” kind of thinking also led them to infer that Arthur Rodriquez was another boy in the class, not the narrator. And visualizing, which we often teach as another strategy for inferring, wouldn’t have made things better.

Experienced readers, however, make sense of the passage by doing something else: They make connections within the text, not outside of it, connecting one detail to another like the dots in those old Connect the Dots games. Some readers, for instance, might work backwards to connect Miss Pringle’s line to the chapter title, the dedication and the first page, which focuses on the legacy of Arturo’s name. Others might work forward, through the rest of the paragraph, inferring what was done by whom to make things easier, who wasn’t asked about what, and how a person could be erased like a ‘used-up word on a chalkboard,’ in order to figure out what happened. These inferences would require readers to connect those lines to their prior knowledge of a highly specific sort. They’d need to draw on their understanding of how pronouns, sentence fragments and similes work to infer what each line meant. And then they have to connect each of the fragments to Miss Pringle’s statement to arrive at an understanding of what, exactly but indirectly, Miss Pringle did.

In this way, experienced readers infer by a process that could be expressed like this:

Text Detail + Text Detail + Text Detail = Meaning

rather than like this:

Text Detail + Prior Knowledge or Experience = Meaning

The students actually used the former process when, after realizing that Arthur was Arturo by making an intra-text connection, they inferred that Miss Pringle had changed many students’ names and that Alicia wasn’t happy about it. Had they used the latter process instead and connected those dark bruises to their prior knowledge, the chances are good that they would have inferred that Alicia had two black eyes. And had they not been connecting the detail dots to draft and revise their understanding as they read, they might also have been mystified by the exchange between Arturo and Alicia, not only not getting who was talking to whom but why Alice would say Alicia’s gone.

So if we want students to read complex texts—not just for the sake of doing so, but to fully engage in rich reading experiences that can inform and enhance their lives—we need to deepen the way we teach inferring and offer more precise strategies. We need to teach them how readers use their knowledge of pronouns and dialogue to steer through dramatized scenes, how they figure out what figurative language suggests, how they make sense of sentence fragments, and how they might use a title as a lens to interpret some of what follows.

Most importantly, though, we need to teach our students this: While they sometimes can figure out what a writer might be saying indirectly by connecting a detail to their own experience, that strategy alone might not help them know what that detail means in the context of the text. The only way to figure that out is to teach them to connect one detail to another, dot by dot by dot, until they see something they couldn’t see before—and they let out the ‘Oh’ sound of insight.