Archive for May, 2012

What Back Pain and Grammatical Errors Share in Common

When you complain to your doctor of back pain, your doctor has the option of approaching the problem from many angles. They could approach it as a structural problem, in which case they order an MRI or physical therapy. They could approach it as a manifestation of a high-stress lifestyle: exercise more and work less. Is it as an issue of tissue inflammation? Take some ibuprofen and apply heat/ice.

When teachers find a grammatical error in student writing, the stakes are lower and the relationship less clinical, but we face a similar sort of decision. How we categorize the errors in student essays is neither trivial nor straightforward. It caries important implications for how we understand it and how a student addresses it.

To see this, try a simple experiment—read the following sentence, which closely resembles a sentence in a student essay I recently read, and then say what type of error it contains:

1. I don’t have much time to surf the internet like my little cousin, who’s on it for 10 or more hours per day, I have work and school to take care of.

Most people will classify this error as a comma splice between “day” and “I.” This is how I was inclined to mark it. But marking it as a comma splice is mainly the result of tradition.

If we leave traditional typologies of error aside, the error could have been labeled as:

  • A failure to insert a conjunction. Perhaps the student wanted to join multiple clauses together, but didn’t know how. This categorization would make sense in the context of a Christensen-inspired sentence-combining pedagogy.
  • An error with the punctuation of a non-restrictive clause. The phrase “who’s on it for 10 or more hours per day” is used non-restrictively, and students are often instructed to set these off with commas. Perhaps the student over-generalized that principle.
  • An error due to an oral-sounding sentence. When students’ speech influences their writing, they might sprinkle in commas where they “sound” right or mix up one type of punctuation with another (commas and periods). After all, in speech, commas and periods sound the same—like silence.

In another view, the error in #1 exists in our imagination only. Edgar Shuster notes that what are traditionally categorized as comma-splice errors do appear intentionally in professional writing.[1] Anecdotally, I have seen an increased incidence of comma-splices in the edited writing of the English teachers that I work with. Punctuation standards change. Twenty years from now, who knows if teachers will judge sentences like #1 as error-free?

There’s no “right” answer on how to categorize this error. Just as back pain is influenced by many factors and can be addressed in many ways, the same holds with student errors. With English grammar and the human spine, many complicated subsystems interact, and a student error often results when several things go wrong at once. Before a doctor chooses how to approach the problem, they need to learn more about the whole person. Likewise, the instructor needs to know something about the student’s background and the pattern of errors in the rest of the essay.

Many grammar errors might better be understood as the surface manifestations of “deeper” problems students are facing. Just as the back and neck muscles tense up painfully under psychological stress, a student who otherwise writes error-free prose slips into clumsy, ungrammatical sentences when they push beyond their zones of comfort, think through complex issues, or try to sound “academic” to their teacher—a point which has been made by Mike Rose[2], David Bartholomae[3] and Ian Barnard[4]. Students’ language processing centers overload when they try—all at once—to think through complex ideas, weave together elaborate syntaxes, and deal with all the other demands of writing and being human.

How you categorize an error strongly suggests to the student how to address it. Labelling the error in #1 a comma-splice suggests the student should change the comma to a period. Labelling it an error with a missing conjunction suggests the student should review the ways to combine sentences. And labeling it a speech-based error suggest that the student needs to more generally become more attuned to the differences between written academic English and the version English they speak at home and with friends.

Finally, let’s consider three less-directive alternatives to the categorizations of the error discussed above:

  • Maybe we don’t even need to bother with categorizing the error. Maybe the student already knows how to fix it. Sometimes students just need to be reminded to proofread.
  • It could be categorized simply as an “error,” with no finer-grained distinctions made. Smart teachers know when it’s futile to try to understand what was going on inside a student’s head. In this case, the student is left to figure out how to fix it. This implies a non-directive pedagogy where students take the initiative in improving their writing.
  • Or the teacher could have ignored the error and said nothing, which suggests to student that in the universe of issues to address in their essay, this one ranks low.

Doctors often respond this way to short-term back pain, essentially brushing it off. Like many student errors, it often disappears on its own with little to no intervention. Patients’ health faces a bigger threat from unintended consequences when the doctor over-treats the initial problem. And I’d argue that a parallel principle holds when teachers have to decide how to respond to a wide swath grammatical errors.


Everything you Know about Subordination and Coordination is Wrong

When textbooks (and teachers) talk about the subordinating and coordinating conjunctions, they often do so in a way that inaccurately describes how English works and that confuses students. The problem stems from the way that textbooks read too deeply into the traditional names of these terms, and they muddle the crucial distinction between syntax and semantics.

Consider the way that these conjunctions are defined by two writing textbooks. Below I will summarize from Writing Worth Reading[1] and A Writer’s Reference[2]—whose definitions typify the genre. These textbooks list the conjunctions that belong to each category. The coordinating conjunctions include:

  • and
  • but
  • or
  • etc.

While the subordinating conjunctions include words like:

  • after
  • because
  • when
  • since
  • etc.

In addition to listing the members of each category, these textbooks define these conjunctions in terms of their rhetorical and semantic properties. Coordinating conjunctions are said to join two elements of equal “weight” or “importance,” while subordinating conjunctions join elements where one is more important or demands heavier emphasis, with the less important or less emphatic element following the subordinating conjunction. The discussion then jumps to giving students rhetorical advice on how to combine sentences and clauses using both types of conjunctions, in order to embellish the relative levels of “weight” or “importance” of the items being conjoined.

Before we continue, let’s consider what decades of linguistic research into the structure of English tells us about coordination and subordination. In most theories, a coordinating conjunction is said to conjoin two (or more) grammatical constituents at the same syntactic level. Usually the conjoined elements are words/phrases that are syntactically like. Any place where you can put a phrase of syntactic type X, you can place two conjoined phrases of type X.

With subordinating conjunctions, the conjunction joins just two clauses/phrases, and the subordinating conjunction puts the two on separate syntactic levels. Usually, the constituent that begins with the subordinating conjunction is said to adjoin to the edge of the independent clause.

Linguists have shown that subordination and coordination differ in a number of other ways, but I’ve covered some of the most crucial. Crucially, linguists do not assume that the names “coordination” and “subordination” have any special significance. Nor do they assume that the syntax of a coordinated or subordinated construction automatically influences the semantics. The relation is syntactic, though individual conjunctions of course differ in meanings.

The textbooks, however, read much deeper into the names that happen to be traditionally applied to these two types of conjunctions. At the same time, they take a generalization about the syntax of two types of conjunctions and assume that a property of the syntax somehow affects the meaning and rhetorical structure of the sentence. Below I will show that why their claim is empirically inaccurate:

The authors of Writing Worth Reading give the following example sentences to illustrate how main ideas should appear in the main clause and less important ideas should appear in the subordinate clause:

1. When I finally left the casino, I had lost my last dollar.

2. After I lost my last dollar, I left the casino (260).[3]

As I interpret these sentences, it does indeed seem like the ideas in the main clauses are being emphasized, but because of a different reason.

It has nothing to do with the syntactic relation of subordination or coordination, and everything to do with the linear order of the two clauses relative to one another. As Joseph Williams points out (in a discussion that has nothing to do with conjunctions), a heavier prosodic stress falls near the ends of sentences, which has a way of subtly emphasizing the endings.[4] We could think of this as the default prosody, all else being equal. Speakers tend to provide information that is new to the discourse at the ends, after familiar information.[5] Therefore, the novelty and unexpectedness of this information can impact the reader, often catching them off guard. And finally, information presented just before some sort of ending point (period, paragraph ending, etc.) will hit harder to the extent that it is the information that the reader is left dwelling on. Note that the prior patterns describe strong tendencies, rather than hard rules. This means that it is not always a reliable property of endings to be more emphatic or important. It does mean, however, that skilled writers can exploit this tendency—which we can think of as a default way of interpreting a sentence—for rhetorical purposes.

We can easily perform a test to prove that the relevant variable is what’s at the ending, rather than what is or isn’t in a subordinate clause. All we have to do is flip the order of the two clauses in #1 and #2 and read them again:

3. I had lost my last dollar, when I finally left the ca-SIN-o.

4. I left the casino, after I lost my last DOL-lar.[6]

As I read these two sentences, the more emphatic part is what comes at the end, rather than what’s in the main clause. Furthermore, this pattern has nothing to do with subordinating conjunctions. It manifests just the same when the two clauses are conjoined with “and.” To illustrate this, I have slightly reworded #3 and #4 to say essentially the same thing, but I have inserted “and” in place of the subordinating conjunction:

5. I had lost my last dollar, and I finally left the ca-SIN-o then.

6. I left the casino, and I had just lost my last DOL-lar.

As I read #5 and #6, the more emphatic part still comes at the end. And we can flip the order around but still leave out the subordinating conjunction:

7. I finally left the casino, and I had just lost my last DOL-lar.

8. I left the casino, and I lost my last DOL-lar.

Again, the more emphatic part comes at the end of both #7 and #8. The effect is subtle, but if people agree with my intuitions, then it strongly suggests that the linguistic distinction between subordinating and coordinating conjunctions is independent of any notion of relative “importance.”

Though I focus in this posting on subordination and coordination, I’m making a point that’s much more general: in textbooks and writing classes, many definitions of grammatical terminology are shaped more by tradition than by what contemporary linguistic research has discovered about how English works. As such, it’s no surprise when students zone out in the face of grammar instruction and come to see it as an illogical activity that depends more on the teacher’s authority than the reality of the language that they live and breath. And it will remain that way until our teaching of grammar becomes more informed by linguistic research.

Here’s an immodest question: do students even need to be taught about the distinction between coordination and subordination?

In some classrooms, it makes sense; in others, it doesn’t. It depends on the larger goals of your pedagogy.

Often, subordination and coordination are taught as part of a sentence-combining pedagogy, where they represent two of the options students can use to join short sentences together and effect a more sophisticated writing style.

In such pedagogy, students can often get by without being explicitly taught the needless complication of the subordination/coordination distinction. When I teach my students sentence-combining, for instance, I minimize focus on the coordination/subordination distinction (which is tough, given that every textbook finds it so important to dwell on). I present students with a representative list of conjunctions (both types), and encourage them to use a variety. In particular, I encourage students to select the conjunction that most precisely expresses the meaning relation between the clauses being joined. This pushes students beyond the kind of monotonous writing where sentences are lamely combined with “and” or “but,” if at all. As a natural consequence, students will be using both coordination and subordination, which adds syntactic variety to their writing.

Students might need to be taught the coordination/subordination distinction for a separate reason—fixing punctuation errors. Coordinating conjunctions and subordinating pattern differently in terms of comma usage, and if students have trouble with these punctuation patterns, they might benefit from being explicitly taught which conjunctions pattern which way.  But crucially, I teach this distinction as one that focuses on word order and comma usage rather than on semantics, and I caution students against reading too much into the names “subordination” and “coordination.”

How do you deal with the issues of subordination and coordination in your teaching? Send me a message. I’d like to hear what other teachers do.

[1] p. 256 – 261 & 489

[2] p. 116 – 120

[3] In each example, I have underlined the main clause, which the authors claim is more important.

[4] p. 122 – 124. Style: Lessons in Clarity and Grace. 6th ed.

[6] In each example, I have marked the heaviest prosodic stress in the entire sentence with capitals.