In a recent email to colleagues, respected writing teacher and author Tom Newkirk offered the following startling revelation: “In what will be an unprecedented spread of automated essay scoring, the two consortia creating tests for the Common Core State Standards both plan to use machine scoring of student essays.” Yes, you read that correctly: Test developers are actually planning to have students’ writing assessed by machines. I find this frightening. I hope you do, too. I cannot fathom any teacher who loves writing, books, ideas, individuality—and students—not being profoundly disturbed by this.
Computers do some things well. They store and retrieve data. They count. And they do as they are told—and ONLY as they are told. They don’t reason, innovate, surprise us, reflect, connect their own personal history to ours, or seek out new ways of looking at things.
When it comes to writing, they can measure only the most superficial features: e.g., words per sentence, sentences per paragraph or document, letters per word, the number of times a word or phrase is used, the use (or lack thereof) of expressions like first, second, third, finally, in conclusion, to begin with or other “cue” words that too often serve as mechanical impersonators of true organizational structure. They can identify grammatical errors—sometimes. And spelling errors—sometimes. Computers can look for capital letters and periods. But if you think they can tell a complete sentence from a fragment, or offer useful advice on re-crafting a sentence, you haven’t put your so-called “grammar checker” to the test lately. To say nothing of the fact that a computer has NO idea whether anything you’ve written is strategic or accidental. Computers don’t have opinions. They’re not into nuance. They have no ear for beauty, rhythm, or soul. They don’t appreciate humor or ponder mysteries. They don’t get the chills when coming to a passage so touching, so right, so magical that it should be preserved forever.
Imagine the work of people like Hemingway (such short sentences!), Faulkner (such long sentences! and run-ons, too!), e.e. cummings (where are your caps, E.E.?), Mark Twain (why can’t he make these people sound more alike?), or Gary Paulsen (one-sentence paragraphs?? seriously?) being scored by a computer. Computers reduce everyone to the lowest common denominator. They’re incapable of recognizing (much less appreciating) genius. Genius is free-spirited. Genius breaks rules and invents new ones. Do we want our students’ genius labeled as “non-standard”? Worse yet, think about this: Do we want everyone to write in the same way? Sounds like the end of literature as we know it. (Ray Bradbury, who knew we wouldn’t even need the firemen?)
There is a reason behind this insanity, of course. It’s called convenience. Think how simple it is to turn all reading of essays over to machines. No authentic reading required. Whew. What a relief. Unlike human readers, who stop to ponder, reflect, remember, react, tear up, laugh, and feel something, machines simply rip through essays, methodically searching out repetitions, misplacements, incongruities, and omissions, and calculating the numbers of each with lightning speed. And of what value is this information? Why . . . almost none. But collecting it is so easy, so affordable.
We need to ask ourselves an important question: What do we want our students to be capable of doing as writers? We don’t need numbers to answer this. I have only to look at my eight-year-old grandson, who is just beginning to think of himself as a writer. This is a fragile identity, nurtured only through diligent, continuous effort—and easily crushed through careless, hasty, or insensitive assessment. I pray that won’t happen to Jack. I want him to choose his own topics—including those for research. I want him to understand the nature of research, to know that it goes beyond books, and to enjoy exploring concepts such as the discovery of new planets or plans for rescuing endangered sea turtles or managing trash so it doesn’t clog our oceans. I want him to be a reader, so passionate about it that he can’t bear the thought of finding himself without a book—and so in tune with what he reads that he continually tells himself, “Hey—look what that writer did. I could try that.” I want his vocabulary to grow easily and naturally, through travel, discovery, exploration, conversation, and the drive of his own curiosity. I want him to be a thinker, to know what it means to tell a riveting story or share information people actually need, or to fight (through solid argument) for things he truly believes in—not ideas imposed from without. I want him to know the thrill of moving readers, of holding them right in his hand and having them wish for more. Most of all, I want Jack to have a voice. His own voice. Not the pathetic echo of a machine that insists on some cheap imitation of its blatantly nonhuman self. I don’t believe Jack will need to know—ever—how many sentences he averages per paragraph, or how many multi-syllable words per hundred he uses.
My colleagues and I are old enough to recall how large-scale writing assessment looked back in the good old days of the 1980s. We respected students’ voices then. We loved our work. We looked on every paper as a gift, and received them as such. We took our time reading them, savoring every word, and read many aloud to each other—for our own amusement, in part. But also to remind ourselves what good writing sounds like. Along with the voices of people like Maya Angelou, Alice Walker, Nikki Giovanni, Larry McMurtry, Anne Lamott, Dylan Thomas, Isabel Allende, Pablo Neruda, Joyce Carol Oates, Louise Erdich, Sherman Alexie, Garrison Keillor, Kate DiCamillo, Gary Paulsen, and countless others (yes, we actually took time to read aloud to one another in those long-gone times), we shared the voices of student authors who were just beginning their writing adventures. Through their examples—and not from rubrics or standards—we taught ourselves the meaning of riveting voice, precise word choice, fluency that’s downright musical, imagery you can’t get out of your head, and organization that relies on logic, not formula. Students’ voices—each fresh and unique—fed and shaped our six-trait rubrics, not the reverse. It all began with the students.
In one of my favorite essays of all time, an eighth grader wrote about his beloved dog Fox and the pond behind their house where he and Fox reigned supreme, chasing away all intruders and guarding the pond as if it were holy ground. Remarkably, the student wrote the whole piece in dialect, not an easy feat for someone his age. The final page recounts how the writer and his family had to move, and how hard that move was. The essay closes with these words: “Fox and I still visit the pond, but it will never be like them three years when she was mine.” That is a line so perfect that to touch it would be sacrilege. But a computer/assessor will likely look for something more like this: “Fox and I still visit the pond, but it will never be like those three years when the pond was ours.” Oops. There goes the voice. There goes the rhythm. But smart money says the “revision” gets a higher score. And what, for heaven’s sake, matters more than a high score?
For years we’ve been saying that teachers of writing should write—get in the trenches to feel the difficulties of coming up with a topic, getting started, knowing how and when to stop, or how and when to revise. Writers know best whether a piece is working—and if it’s not, what to do about it. Now, apparently, we’re going to abandon that writers-know-best philosophy. Ironically, we’re going to turn the very important job of assessing something as complex as writing over to a machine that cannot think, much less write. Oh, computers can spit back what’s fed in—but that’s hardly the same thing (though it does come dangerously, hideously close to writing by formula).
And so . . . if we let machines that cannot make the music judge the value of the symphony, what happens? What’s the upshot of this decision? First off, we have to bring our capabilities—or more to the point, our students’ capabilities—down to the computer’s level. But that’s not the worst of it—mediocre writing that touches no one’s soul. That would be depressing in and of itself. No, it gets worse. Teachers already complain that many students lack motivation to write. If you think this is a problem now, just wait. How motivated would you be if your only audience were a box that could only recognize you as “input”? And even that’s not the worst of it. By assigning writing assessment to machines, we teach our students that writing is a simplistic, mechanical act, devoid of complexity, design, purpose, and heart. We are asking them to substitute conformity for originality, capitulation for courage. Most insidious of all, we are de-valuing individuality and voice, which is to say, the personal and cultural differences that make each of us unique.
We don’t write, in the end, to dot i’s and cross t’s. Conventional correctness is a means, not an end. We write, as James Baldwin so famously said, to change the world. Writing and reading are a profoundly beautiful dance in which two people work together to create meaning. Why would we not wish to respect something so rare, so extraordinary? If we truly cannot afford to assess writing well, with the time and effort and attention to detail that our young writers deserve, then let’s not do it at all.
Cathy Bernard, Associate Professor of English at the New York Institute of Technology, recently quoted in Marshall Memo 481 (www.marshallmemo.org), put it this way: “On Sunday nights I would rather be anywhere but at my desk facing a stack of student essays. Still, I question the arguments offered in support of automated essay scoring. The chief selling point of electronic assessment is that it gives instant feedback… But writing is not a game in which you click away until you hit on the right answer. Writing is thinking, and revision is a slow process, unpredictable and exploratory. A piece of writing, like a cake taken from the oven, needs some time to cool before the revision process can even begin.”
If you feel strongly about this issue, you may be wondering what you can do. Speak up—to everyone you know. Friends, colleagues, administrators, test developers. Let your voice be heard. To sign a petition protesting machine scoring of student work, please visit http://humanreaders.org
Coming up on Gurus . . .
Look for Part II of Jeff’s review and related classroom work with Paul Fleischman’s amazing book The Matchbox Diary. Next time, Jeff brings us right inside Mr. L’s classroom to show how student writers can take off when inspired to make some pretty cool associations of their own. Remember, for the BEST in writing in-service, featuring traits, standards, writing workshop, process, and literature-based writing lessons, just log in to any computer. WE ARE KIDDING. For the human touch, please phone us at 503-579-3034. Give every child a voice.