The Science of Writing: A 36-Year Journey and Counting

Science is not a hammer.

Science is an old-growth forest, each tree an organic thing. Think of a tree as a theory. At any moment that tree (theory) is fully a tree but not the tree it will be.

As a tree grows, it becomes more robust, a stronger trunk, deeper roots. If we inspect that trunk, we find rings detailing the history of how that tree became stronger with age. Theories too are not simply the result of fixed evidence, but an accumulation of evidence, an accumulation that evolves over time.

Science, like that old-growth forest, is never settled, but it is always at any moment the best that it can be in terms of being a forest and in terms of the trees being the tress of that moment. You see, science is also organic, not yet the forest or trees it can and will be.

Old-growth forests are also characterized by being untouched by humans, and while science is the product of humans, science often seeks ways to limit the flaws of that human contact (a lofty and unattainable goal, but one that helps science aspire toward truth and Truth).

Science ultimately is aspirational; it can never be settled, fixed, and anyone using science as a hammer is, in fact, not being scientific.

Science is not a hammer.

Science is an old-growth forest, each tree an organic thing.


Viewing science as a hammer is the fatal flaw of the “science of reading” (SoR) movement that has gained momentum in 2019. Advocates of SoR begin by claiming that this science is settled:

In spite of the current discussions, the science on this instructional issue is settled. Castles, Rastle, & Nation (2018) lay out that there is a clear progression to effective literacy instruction. First and foremost, children need to understand the principles of spelling-sound correspondences and to solidify a store of high-frequency words to read words and phrases fluently. Most children need explicit teaching to build this knowledge. After decoding and high-frequency words are established, more attention can be devoted to comprehension with a focus on making meaning. Castles et al. (2018) offer a logical and research-based model. In spite of this research, educators remain without consensus about what is most important—phonics instruction or a focus on comprehension.

Science is not a hammer, neither is it to be used to bludgeon nor is it a singular tool.

In fact, especially for education as well as teaching literacy, science is a much broader spectrum of evidence than SoR advocates are arguing, steeped as they are in the neurosciences.

The science needed to guide real-world teaching of literacy is an old-growth forest of many types of trees at different stages of growth.

For example, I primarily have been a teacher of writing for 36 years and counting. I have taught students from 9th grade through graduate courses.

As a scholar of teaching writing, I am well versed in the experimental/quasi-experimental research base on teaching writing as well as a huge and complex body of qualitative research.

I also have 36 years of experience with thousands of students.

All of that is at my disposal as I teach any student to write, an act that for me is highly individualized—even when I taught 100-125 high school students five days a week.

The generalizations and controls that result from and govern experimental/quasi-experimental research (which is dominant in neuroscience) are informative (not prescriptive) for me as a teacher, but my work tends to be with many different outliers—humans, that is—who may thrive with practices outside the constraints of narrow types of science.

I don’t use science as a hammer because students are fragile things, and instruction that treats them all as ten-penny nails is unwarranted.


You may be thinking about climate change, evolutionary science, or vaccinations—all of which many people would argue are settled science.

“Settled,” I think, remains a problematic word even in those contexts.

All science based in experimental/quasi-experimental research when properly vetted is compelling, compelling to the point that it feels settled, compelling to the point that we must act in ways that confirm it is settled even as we are aware this tree may grow.

Since all sciences remain in the replication loop, we are best off calling even the largest tree with the most powerful trunk and deepest roots “compelling,” not settled.

In qualitative research, “compelling” is the best we can hope for, but much of that research is compelling, although with caveats about the evidence not reaching standards of generalizability and the conditions of the evidence not bound by controls.


Let me end with an anecdote, what some would call not scientific. It is the story of having taught writing for 36 years and counting, and still being very cautious about my practice and very nervous about the fate of my field of teaching writing.

Actually this is an anecdote about gathering anecdotal evidence, the sort of scientific teaching that John Dewey envisioned for progressive educators.

I always spend the last class of my first-year writing seminars by discussing with students what has worked and what I should do differently in the future.

I also use this class to re-emphasize that my overarching goals for these classes are about fostering in them greater authority and autonomy as students and writers about to run the gauntlet of three-and-a-half additional years (or more) of college.

This fall, students argued for having Essay 1 turned in earlier, allowing more time and class sessions for Essay 3 (the academically cited essay), and moving Essay 4 earlier to leave more time for the revised submission.

We fleshed out these requests against the goals of the course, and ultimately, I found their anecdotal feedback compelling. My schedule for fall 2020 will be revised.

As the professor, as well, I have reflected on how to better encourage students to revise their essays and not simply address what I have marked for them. I discussed this problem with another teacher, and am considering a new policy on how students should resubmit their essays.

In the past, I have required students to resubmit essays in clean Word files, track changes, comments, and highlighting all removed. Part of that requirement was aimed at helping students better use Word as a tool, but I also have trouble with Word files that are busy.

However, as I discussed student revision with a friend who teaches writing, I thought about how students having the track changes visible for their revisions would show them how much, or how little, they actually revised. Visible track changes can be a very effective teaching tool.

So my new policy may be that students submit two Word files, one clean and one with only the track changes of their revisions (with the file including “TC”).

This, then, is a brief anecdote about how I teach scientifically as a professional educator, a writer, and an expert in literacy. I teach with caution, I resist teaching with a hammer.

This means that when some students demonstrate a need for a type of instruction not supported by a narrow type of research, I still provide the student with that instruction. We may even experiment with a range of strategies until the student feels capable on their own.

I am always cautious, but I am also nervous because while the “science of reading” mania is in full stride, I see on the horizon a similar fate for the teaching of writing: Scientific evidence on how to teach writing is slim.

I suspect the mainstream media will discover a field that already exists, has for a century or more. I suspect the allure of “science” will blind that media and those who also feel passionate about the dismal state of student writing.

So somewhat preemptively, I want to offer about the teaching of writing:

Science is not a hammer.

Science is an old-growth forest, each tree an organic thing.