What computer programming can tell us about writing

I’ve been doing a fair amount of coding lately—I’m working on an upgrade to GrammarTrainer to make it more user friendly and more informative about student progress.

And as I code my way through javascript and ajax and PHP, I can’t help noticing some similarities between expository writing and computer programming.

Good writing requires logical organization; computer programming, even more so. Even though a disorganized program can, in principle, run properly, it’s awfully hard to debug or upgrade—let alone for a new coder to make sense of.

Good writing requires clarity and precision; computer programming, even more so. The slightest lapse yields code that runs poorly, weirdly, or not at all.

For clarity in writing, it’s often a good idea to break long, complex sentences and paragraphs down into shorter ones; so, too, with programming. You start coding a routine, and, a dozen lines in, you realize that the various if-thens, for-loops, and string manipulations actually comprise several distinct subroutines.

Good writing also involves labeling new concepts with evocative words and phrases; so, too, with those subroutines in programming. It’s a lot easier to keep track of what you’ve done if the subroutines—and the variables—have perspicuous names like “setCursorPosition” and “wordStartPosition,” as opposed to “functionA” and “X”.

Good writing requires an economy of words; even more so with programming, where repetitions can slow down run times.

Finally, computer programming involves a kind of creativity that isn’t so different from the creativity we associate with writing. All writing can (or should) involve some creativity, but here I’m thinking of story writing in particular—and of its distinguishing features of plots and characters. After all, plot-like and character-like entities also figure in computer programs. And, much like the story writer staring at his blank page or screen, the computer programmer, staring at hers, has to figure out what events the program should execute, and in what order—as well as who the various actors are, what they’re going to be called, and how they’re all going to conspire to make everything unfold in just the right way.

Is sequencing sentences no longer a writing skill?

In the following questions, the first sentence of a paragraph is given. Your job is to “unscramble” the rest of the paragraph by putting the next five sentences in the correct order.

QUESTION 1 In the remote mountain country of Nepal, a small band of “honey hunters” carry out a tradition so ancient that it is depicted in drawings dating back 10,000 years.

______ Q. Throughout this entire dangerous practice, the hunter is stung repeatedly.

______ R. To harvest the honey from these combs, a honey hunter climbs above the nest, lowers a bamboo-fiber ladder over the cliff and climbs down.

______ S. The honeybees that colonize the Nepalese mountainsides are among the largest in the world, building huge honeycombs on sheer rock faces that may be hundreds of feet high.

______ T. Only veteran honey hunters, with skin that has been toughened over the years, can return from a hunt without the painful swelling caused by these stings.

______ U. Once he has reached the level of the nest, the hunter uses two sturdy bamboo poles like huge chopsticks to pull it away from the mountainside and into a large basket, which is then lowered to people waiting below.

QUESTION 2 In the 1880s, John Wesley Powell, an explorer of the Grand Canyon and director of the United States Geological Survey, led the development of the first topographical maps of the entire United States.

______ Q. This is because streams cut into the land, so contour lines will turn upstream, cross the waterway and return downstream, creating a V shape, with the “V” pointing upstream.

______ R. Waterways, such as streams, are usually marked in blue on topo maps, but even if they were not, the presence of one could still be identified using contour lines.

______ S. Contour lines indicate the slope of the land as well.

______ T. If the lines are close together, the elevation is changing rapidly and the slope is steep, whereas widely spaced lines depict a gently sloping terrain.

______ U. Also called “topo maps,” these maps differ from others in using thin brown lines, called contour lines, to connect points of equal elevation.

Answers appear in an article in yesterday’s New York Times, which is also my source for these excerpts.

Since 1994, the article reports, questions like these have appeared on the “scrambled paragraphs” section of New York City’s specialized high school entry exam: the sole determiner of whether students are admitted to places like Stuyvesant High or Brooklyn Tech.

But now the scrambled paragraphs are on their way out. Quoting the Times, they’re being “replaced multiple choice questions intended to evaluate writing skills.” As intrigued as I am to see what these new multiple choice questions look like, I’m also baffled by the implication that sentence sequencing isn’t a writing skill.

But that may be beside the point. As the Times explains:

These paragraphs were, to put it politely, unpopular with many students and their families, and were often criticized as having little relation to what students were taught in school, which could give an advantage to those who could afford to pay for test preparation. The city’s Education Department decided to remove them to align the test more closely with classroom instruction.

Yet another indication that sentence-focused instruction is totally passé.

Catherine and I, of course, are trying to bring it back–scrambled paragraphs included. So now we have to wonder: is unscrambling paragraphs really as unpopular with students as the Times implies?

A big help in unscrambling paragraphs, by the way, is awareness of cohesive devices (like those I discuss in my earlier post). Try out these questions, and you’ll see what I mean.

Let’s resurrect sentence-based rhetorics

Part of what’s distinct about Catherine’s and my curriculum is that we’re zeroing in on the basic building blocks of writing–phrases, sentences, and paragraphs.

Many of today’s writing classes instead zoom out to the big picture–communication, argumentation, and audience. Get students thinking what their ideas are and who they’re addressing them to. Have them do lots of open-ended writing. Perhaps toss some peer-editing into the mix. Surely, then, students’ phrasing and paragraphing skills will develop organically, without explicit instruction by teachers or textbooks. And surely, compared with scores of exercises in sentence construction and paragraphing, the holistic, naturalistic approach is more effective and appealing to all concerned.

It certainly appeals to many of classroom writing instructors–as it has for a long time. But, it turns out, sentence-based instruction goes back even further. A while ago, Catherine sent me an article by Robert J Connors (“The Erasure of the Sentence“), and I was surprised to learn that explicit instruction in sentence syntax had been a staple of composition classes (dating back to classical antiquity) until just a few decades ago.

What happened? Here, in a nutshell, is Connors’ thesis:

The usefulness of sentence-based rhetorics was never disproved, but a growing wave of anti-formalism, anti-behaviorism, and anti-empiricism within English-based composition studies after 1980 doomed them to a marginality under which they still exist today. The result of this erasure of sentence pedagogies is a culture of writing instruction that has very little to do with or say about the sentence outside of a purely grammatical discourse.

What are these sentence-based rhetorics that have fallen out of favor? One recent example is Francis Christensen’s “sentence combining.”  In Connors’ words, “Sentence-combining in its simplest form is the process of joining two or more short, simple sentences to make one longer sentence, using embedding, deletion, subordination, and coordination.”

According to Christensen, you could be a good writer if you could learn to write a good sentence. His pedagogy consisted of short base-level sentences to which students were asked to attach increasingly sophisticated systems of initial and final modifying clauses and phrases-what he called “free modifiers.” Effective use of free modifiers would result in effective “cumulative sentences,” and Christensen’s most famous observation about teaching the cumulative sentence was that he wanted to push his students “to level after level, not just two or there, but four, five, or six, even more, as far as the students’ powers of observation will take them. I want them to become sentence acrobats, to dazzle by their syntactic dexterity.”

Another “sentence-based rhetoric” was Edward Corbett’s “imitation exercises.” This involved the “the emulation of the syntax of good prose models.” Students would begin by copying a model sentence word for word. Then came “pattern practice,” in which students construct new sentences that parallel the grammatical type, number, and order of phrases and clauses of the model sentence, perhaps with the help of a syntactic description of the model sentence’s structure. Students might also perform syntactic transformations (informed by Noam Chomsky’s Universal Grammar) on the model sentence. In Corbett’s words, the aim of such imitation exercises was to “achieve an awareness of the variety of sentence structure of which the English language is capable.” Other advocates of imitation exercises noted that student writing “is often stylistically barren because of lack of familiarity with good models of prose style;” the remedy was explicit emulation of good models.

Both Corbett’s and  Christensen’s methods were subject to empirical scrutiny, and studies showed that both methods not only increased the grammatical complexity of student writing, but also improved the overall writing quality (as compared with control groups and as rated by blind raters). In particular, internalizing syntactic structures, even by slavishly copying them, ultimately increased originality and creativity–presumably by giving students a wide repertoire of syntactic tools to choose from and handy ways to play around with them.

But as Connors notes, almost as soon as this sentence-syntax teaching methodology starting showing empirical success, it was shouted down into oblivion by critics who found it philosophically distasteful.  After all, these methods involved:

  1. Textbooks
  2. Mere exercises, devoid of content and real-world application, with (worse yet!) correct and incorrect answers
  3. Rote imitation
  4. An inorganic, narrow, analytical, reductionist approach that stifles creativity
  5. A procedural focus at odds with the authentic writing process in which motivation and communicative intent and self-expression come first and everything else comes along for the ride (including, apparently, grammatically well-formed sentences).

The result of this backlash was that most writing instructors came to believe that “research has shown that sentence combining doesn’t work.”

Communication, argumentation, and audience–these seem like great things for writing instruction to prioritize. But as Catherine and I discuss on our book and will continue to explore here, effective communication and argumentation depend largely on effective phrasing and paragraphing. And, as those studies of Corbett’s and Christensen’s methods suggest, explicit practice with sentence and paragraph construction yields skills that typically don’t emerge from today’s more naturalistic, holistic approaches to writing.

Flow redux

As a writing instructor, I’ve been chronically frustrated by the fact that composition textbooks use words no one has ever defined.

Flow, for instance.

What is it?

Or paragraph.

What is a paragraph apart from a list of sentences separated from two other lists of sentences inside a longer text?

The answer is that a paragraph has a topic (topic?) and the sentences have flow.

But we already said we can’t define flow, so that’s no help.

Then there’s the universal margin-comment: awkward, awk for short.

What is an 18-year old college freshman who writes awkward sentences to make of the word awkward?

Not much, I fear.

How to turn a list of sentences into a paragraph

I came across this passage in a Times story by the mother of a high school student who wrote letters to successful dyslexics asking for advice, and thought it would be fun to see if I could punch it up a bit using principles Katharine and I teach in Europe in the Modern World:

Aidan had started the project in a moment of despair right after getting back his spring grades in ninth grade. They were disappointing. They didn’t reflect how hard he had worked. We were standing in his room at the time. I had pointed to a poster he had tacked up over his desk of successful adults who have dyslexia. “I wonder how they made it?” I had said.

Jay Leno’s Advice for My Dyslexic Son

First thought: personal writing is harder than it looks.

This is a successful piece of writing; it was published in the Times, after all.

Even so, the rhythm needs work. There’s a choppy, start-stop quality to the sentences that makes them read more like a list than a paragraph.

Speaking of: paragraphs are a topic I’ve been contemplating for a while now.

What is a paragraph, exactly?

I’m not sure anyone knows–people don’t seem to study the paragraph per se–but one thing a paragraph is not is a list. To write a proper paragraph, you have to do something to make each sentence feel connected to the one before. You have to create “flow.”

Flow is hard, maybe especially hard in personal writing, where every sentence logically begins with “I” or “he” or “she” or “they” or “we.” Too much of that and you’ve got a list.

So I thought it would be fun to see if I could make the sentences in this paragraph flow.

It wasn’t easy.

I see it’s getting late – back tomorrow with the rest.


How to turn a list of sentences into a paragraph – 9/20/2016
Get me rewrite – 9/24/2016
Why it’s hard for a memoirist to write non-choppy prose and sound like a normal human being – 9/27/2016