The Science of...Writing?

Teaching requires lots of experimenting, but that doesn't make it a "science."

There’s something I need to get off my chest.

There is no such thing as the “science of writing.”

There isn’t such a thing as the “science of reading” either. When you see the construction the “science of…” as applied to education you are looking at a marketing term, not a reflection of something real when it comes what most of us think of when it comes to scientific inquiry and the standards of proof for claims.

Before I get dragged into the reading wars—from which I am a conscientious objector—I am not denying the evidence as to the importance of phonics instruction for many students. We have a robust evidence base that suggests somewhere around 60% of students benefit from explicit phonics instruction in order to acquire the skill of decoding words, a necessary precursor to being able to read.

But idea that phonics instruction is somehow a science of reading falls apart at the start given that 40% of people learn to read without any explicit phonics instruction.

More importantly, decoding phonemes is not “reading.” Even when I say “reading” we have to decide whether we’re talking about a skill or a behavior. There are many different types of reading skill we may need to acquire and apply over the course of our lives. One of the unfortunately negative consequences of an emphasis on the narrow acquisition of reading skills in school and the standardized tests that run alongside these initiatives has been to drive out reading as a behavior, as covered recently by Dan Kois at Slate looking at a report at how voluntary book reading has decline precipitously, and not just because of students turn to their screens.

It seems clear that we need more and better phonics instruction in schools to make sure children who cannot decode learn that skill, but we are already in danger of overshooting the mark, as the UK did when the mandated phonics instruction in schools and then a decade later found out they’d wasted significant instructional time on a discrete skill many students already had.

At my

The Biblioracle Recommends

newsletter I wrote about the need to “make more readers” meaning not more people who know how to read, but more people who want to read. We need to look at reading as a “practice” that we want to be able to help students cultivate. Within that frame, decoding is the point of entry, but it is just barely the start of the journey. If we don’t have structures and incentives that invite students into the rest of that journey, we’re not going to have students who are truly able to “read.”

If the only measurement that counts is the percent of students who can decode phonemes, we are courting a very bad case of Campbell’s Law in which we substitute the measurement (decoding) for the actual social outcome we desire (reading).

Anyway, I’m off track, I wanted to try to nip this notion of the “science of writing” in the bud.

The whole concept of a science of writing is nonsensical. Let me count the ways:

1. We have centuries of evidence demonstrating that people can and do write effectively despite using wildly divergent processes in their writing.

2. A science of writing suggests there is a measurable objective standard for quality writing. There isn’t. This is one of the things that is awesome about writing.

3. A science of writing suggests there is a causal sequence which leads to (apparently objectively measurable) writing skill. This is not true. We definitely learn to write by writing, but what, how, and when we learn is variable.

There is no such thing as the science of writing, the same way there is no such thing as a science of making music, or science of loving your dog.

All that said, the fact that we cannot reduce writing to a “science” does not mean that writing or teaching writing is “just vibes.” We have significant evidence about different aspects of teaching writing and their relative effectiveness.

For example, we know that structured practice in a variety of genres, and the more writing students do overall, the better. We know that students benefit from writing to audiences and working as part of a community. We know that writing benefits from a process.

We also know some stuff that doesn’t seem to impact improvement as writers. For example, diagramming sentences is an interesting way to visualize the structures of a language’s syntax, but it does not translate to improved writing.

Above all, we have excellent evidence as to the best learning conditions for teaching writing. The NCTE (National Council of Teachers of English) recommends in post-secondary education that instructors be limited to three sections of students of fifteen students each, for 45 total students. These are conditions I experienced as an instructor exactly zero times in 20 years of teaching. Some semesters I had triple or close to quadruple that number of students.

When outside reformers start invoking science and writing, if they want me to listen to their ideas on writing pedagogy, they can start by first establishing the proper classroom conditions.

I’ve always run my classrooms as extended, iterative experiments of changing variables and collecting data. I try something, I see how students respond, I measure the results in both quantitative (e.g., the number of words written by students) and qualitative (student attitudes towards the experience) ways, and then re-adjust before experimenting again the next semester.

This is how I ultimately developed my approach of “the writer’s practice.” It truly is the byproduct of long and rigorous experimentation.

But that doesn’t make the results “scientific.”

When someone or something is touting a “science of writing” as a marketing term. what they are really describing is a method for passing a standardized assessment, an assessment which is likely too narrow, is gamed by employing prescriptive rules, and is quite likely an entirely alienating experience for all parties involved, particularly students.

“Science of…” is meaningless. When we talk about how to teach writing, the conversation is not about the absolute effectiveness of a pedagogical approach or specific curriculum.

It’s a discussion about the values we bring to the work. I’m a big fan of my own approach of “the writer’s practice” but if you don’t dig the values embedded in that approach - student agency, writing outside of academic contexts, emphasis on formative reflection over summative grading - you are not going to find my approach particularly attractive.

That’s fine! It’s a big world.

Let’s not get sucked into a pointless discussion when there’s more important stuff to talk about, and an endless number of experiments we should be doing.