My waggish friend Steven Levine sent me, a little while back, a page from a grade school workbook on writing (I don't know which one, nor do I much care; this page is a not at all remarkable instance of the workbook genre). Here's the text of exercise 125, "Interesting Sentences":
A good sentence should be interesting.
"I have a dog" is not a good sentence with which to begin a story. [Note the very formal fronted preposition; no stranded prepositions! Possibly the writer of this sentence genuinely believes that "preposition at end" is ungrammatical, or maybe the writer is just trying to model "the best grammar" for the kids.] If you are writing a story about your dog that was lost, it would be better to begin the story, "Last week my dog Shep ran away from home."
Can you change the following sentences into interesting sentences? [Note that this is an instruction to change the sentences, not an actual question.]
The sentences are:
1. I have a bicycle.
2. Charlie has a goat.
3. I have a dress.
4. Brother gave me a wagon.
5. I have a pony.
6. My shoes are new.
(and there's a line at the end labeled: My score……………….)
There's a lot that could be said about this exercise, but here are a few observations.
To start with, it's not entirely clear what the task is here. In the example, the uninteresting sentence is considered as the first sentence in a story, but the instructions don't specify that for the sentences in the exercise. In addition, the instructions don't say anything about what makes a sentence interesting; students have to induce that from the one model provided. Then there's the question of what the exercise is supposed to teach.
I can see at least two ways in which an English teacher might judge "I have a dog" to be uninteresting: it's not at all rich in information, and it has a state-denoting verb, have, rather than an event-denoting verb (like run), so that it "lacks vitality" and is "inactive" (as some critics would put it). (It does have virtues: it's brief and very easy to process.)
Take the second objection first. The idea that event-denoting clauses — especially those with agent-denoting NPs as subjects — are intrinsically superior to state-denoting clauses (because they are "more vivid" or the like) has a considerable history, which we have often commented on here at Language Log, especially in connection with injunctions to use active constructions rather than passive ones (and to avoid copular BE wherever possible, and so on), all in the service of an "active style", which is metaphorically associated with directness, forcefulness, power, and even masculinity.
[I'm ashamed to have to say this, again, in yet another formulation, but in criticizing people who are so insistent in their advocacy for an "active style", I am not recommending some asinine opposite style (use the passive wherever possible, use predicative NPs, AdjPs, and PPs wherever possible, use existential-there clauses wherever possible, use nominalizations wherever possible, etc.). In each case, the alternatives have their places and their uses.]
Of course, no one explains such things to kids. It's not even clear how you would explain them, or indeed why you should be trying to explain it.
Part of the problem is that the writing issue is characterized in such vague terms ("(un)interesting"), even coyly. Another part of the problem is that the "uninteresting" constructions have manifold uses: sometimes their function is to provide background (making everything vivid and foregrounded is not at all a service to the reader), and sometimes they can be very vivid indeed, as in this extremely tiny sampling of the possibilities:
[predicative noun] You're an asshole.
[state-denoting HAVE] I have a gun in my pocket.
[existential THERE] There are leeches on your legs!
["classic" passive] My best friend was mauled to death by bears yesterday.
[Surely a child who writes that last sentence as the beginning of a story -- in a, one hopes, imaginative rather than factual report -- gets a Gold Star for interestingness. And for topic management: "Bears mauled my best friend to death yesterday" is another way to do things, but it seems to be about bears, not the death of the writer's friend.]
Yet another part of the problem is that workbook exercises are an absolutely wretched way to teach anyone (child or adult) anything about writing. Exercises in revision (of other people's or the student's own texts) can be tremendously useful, but reworking single sentences out of context is unlikely to help anyone.
There's more to be said on the lack-of-vividness criticism, but I hope this will do for now. On to the lack-of-information criticism.
Here the workbook writer is aiming for information-rich sentences, especially those that introduce information via modifiers or other dependent expressions. So, instead of the slow build-up of
(a) I have a dog. Shep is his name. He ran away yesterday.
(b) I have a dog, who is called Shep, and he ran away yesterday.
students are encouraged to package the propositions into something syntactically denser, like
(c) My dog Shep ran away yesterday.
which conveys (1) that the writer has, or had, a dog, via the determiner my, and conveys (2) that the dog is named Shep, via the postnominal adjunct Shep, while asserting that this dog ran away yesterday. In (a) and (b), (1) and (2) are actually asserted, while in (c), they are conveyed by presupposition.
These are two different styles of writing, and there's no reason to insist that one of them is intrinsically superior to the others, on all occasions. If you look at actual stories, say published short stories in English, you'll find a variety of techniques for introducing information, techniques that achieve different effects. Good writers use a lot of them.
As I see it, the task of workbook writers in this case is to expand the range of kids' writing, and that's entirely admirable. But the tactic of deprecating things like (a) and (b) as "uninteresting" is obtuse and unhelpful.
To start with, things like (a) and (b) are evidence of real learning by the kids who write them.
Any number of people have noted that, though young children in some circumstances, are pretty good at taking the viewpoint of someone else, they often aren't very good at assessing what other people know. As a result, they are inclined to say things that presuppose that their interlocutors know what they know. For instance, they sometimes use deictic expressions (including personal pronouns) that haven't been anchored in the context: "She asked me a question", for instance, said when no discourse referent has been set up — a she pretty much out of the blue.
The problem is vexing enough in ordinary conversation, when the child usually knows their audience, but it's even worse when the child is trying to tell a story to an audience that's likely to contain people who don't know the backstory, and even more perilous when the child is asked to address an unknown audience, as in decontextualized "write a story" school exercises.
In my experience, many kids quickly learn (from being criticized and corrected) that they have to set things up. So they stage things, introducing one proposition at a time. Something like (a) is in fact remarkably respectful of its audience; (b) is a bit more direct, but still brings the background in bit by bit. This is admirable and praiseworthy. Of course, teachers want to move kids on to more complex presentations of information, but surely they can offer tenchniques to urge kids in this direct, rather than labeling the strategies the kids have devised as "uninteresting" (which will translate as "this will get you a bad grade"; remember that the page comes with a score).
Spinning off from this: the workbook, along with other materials on writing for children is recommending a particular discourse form over others (much as usage manuals recommend particular lexical choices and constructions over others) — something like the "plain style" (though there are many understandings of what counts as "plain"). I myself generally opt for some variety of "plainness", but I can do other things, and I don't deprecate other ways of writing or speaking.
My larger point is that the workbooks are trying to inculcate something in between the very paratactic style of (a) — compare fairy-tale beginnings ("Once upon a time there was a queen named …") and epic tales (like the Sanskrit epic episode that begins, in translation, "There was a king, Nala by name, …") — and a much more elaborate style in which referents are introduced and situations are described by presupposition or even more indirect means (for instance, by beginning in medias res), and the syntax is correspondingly complex.
[This story is continued on my blog.]