Tuesday 4 March 2008

On non-ambiguities

A correspondent writes that he is worried about the potential ambiguity in such sentences as (the first one from Graham Greene):

(1) He did not dare to go up to where she slept, feeling that if he saw her face again he could not leave her.
where the feeling clause relates back to he, and
(2) He did not dare to go up to where she slept, dreaming of her new life in France.
where the dreaming clause refers back to she.
He wonders whether it is possible to interpret sentence (1) as referring back to she and sentence (2) as referring back to he. 'The participles are in the same place', he observes, so 'how can it be that they go with different subjects in each case'? How does one make it clear who is doing the action, in such cases?

Examples of this kind need to be discussed with reference to the distinction between finite and nonfinite. The above sentences contain participial clauses, which are literally 'not finite', i.e. they do not limit the action of the verb - by contrast with finite clauses, which do limit the action (typically by specifying the number, person, and tense involved). This means that there is always the possibility of grammatical ambiguity whenever a nonfinite clause is used. If I begin a sentence with Walking down the street... it is unclear whether the action is taking place now, or has taken place, or is yet to take place, and who or how many people is/are walking. Once the sentence continues with a finite verb, these matters are resolved: Walking down the street, I saw..., Walking down the street, she sees..., Walking down the street, they will see..., and so on. That is what finite verbs are for, to resolve ambiguity of person, time, and number.

So, faced with examples of the kind mentioned above, the only way to obtain the desired clarity is to go for a finite construction. In (2), for example, it would be more explicit to say: where she was dreaming of... - though the repeated wheres are not particularly elegant. And if it was he the writer had in mind, something like as he was dreaming... would make that clear.

However, such rewriting is hardly required in such cases, for our contextual awareness of the relationship between sleeping and dreaming makes the 'she' interpretation far more likely than the 'he'. Or, putting this the other way round, any writer who wrote sentence (2) wanting to convey the fact that it was the 'he' who was dreaming is doing something perverse - going against our normal contextual expectations. A restructuring of the sentence (such as He was dreaming of her new life in France, so he did not dare...) would be much clearer. Similarly, because we know that when people are asleep they are not 'feeling' in the sense of sentence (1), we are likely to rule out an interpretation in which the feeling refers to the sleeper.

My correspondent worries about how to resolve the question, but the answer is revealed in a telling phrase he uses: 'It is clear (if only from the context) that the participle 'feeling' describes the subject of the sentence'. That parenthensis is rather dismissive, and therein lies the root of the problem. Context is actually everything. Anyone who tries to analyse grammar as if it was somehow separate from the rest of the linguistic world is going to end up in trouble - seeing ambiguities everywhere! Grammar is never separate from meaning (semantics) and use (pragmatics), and all sentences have to be taken wholistically, with reference to these factors. In the vast majority of cases, potential grammatical ambiguities are avoided because our semantic and pragmatic knowledge tells us how to interpret them.

If you want to see this viewpoint developed in my own writing, take a look at Making Sense of Grammar (Longman, 2004), which applies the semantic/pragmatic perspective to English as a whole. I should perhaps add that prescriptive grammars and purist pedants are of no help whatsoever in such cases, as they condemn all so-called 'dangling' participles, regardless of context. So a sentence like Driving in from the airport, the flags were fluttering proudly - an illustrative sentence from John Humphrys' Lost for Words - elicits the comment 'What accomplished flags they must have been'. That is what happens when you let grammar govern your thinking to the exclusion of everything else. You lose all touch with reality!


Andrew Shields said...

I'm a little puzzled by something here. You write:

"Once the sentence continues with a finite verb, these matters are resolved: Walking down the street, I saw..., Walking down the street, she sees..., Walking down the street, they will see..., and so on."

This seems to imply that the subject of the main clause that follows should be read as the "subject" of the participle in an introductory participial clause.

But then you defend your final example ("Driving in from the airport, the flags were fluttering proudly") as unambiguous because of semantic context. I'm willing to accept that (though I wonder about cases where the "grammatical" reading of a sentence is not as easily corrected by a "semantic" reading of a sentence), but it does seem to be at odds with the implications of your first explanation.

I don't really have a question, except perhaps "could you clarify this further?"

DC said...

The 'driving' example is exactly the same as the 'walking' one. In both cases, a nonfinite clause requires a finite clause to resolve the inexplicitness. If the 'driving' example had been Driving in from the airport, I saw the flags fluttering... there would have been no problem, as the grammar alone would have been sufficient to resolve the issue. There is no conflict here with our semantic or situational awareness. But when the following clause begins with flags (or some such item) something different happens. Our grammatical expectations are outflanked, as it were, by the greater interpretive power introduced by our awareness of what the words mean and how the real world works.

We always have a choice, of course. We can insist on maintaining the relevance of the grammatical rule - which is what we do when we decide to make a joke about it. But in most everyday circumstances we allow the priority of the contextual interpretation. And if there are cases where there is a genuine uncertainty as to whether the grammatical or the contextual reading is correct, we have a real ambiguity, and we had better rephrase. That sort of case would include sentences like Standing in the courtyard, the guards shouted at the prisoners.

Hope this helps!

Andrew Shields said...

That explanation was extremely helpful, thanks. It helps me clarify my own approach to teaching the problem.

Anonymous said...

Hi there. Came across this blog in a random search on the internet for ambiguity, participles and finite verbs. Seems to be what I need!

And your post actually touches directly on a question I have. You suggest in the post that ambiguity issues are (generally) resolved once a finite verb is introduced. And I can see how that is so in some cases. But it seems to me that you can have a similar ambiguity (to the Graham Greene one you posted about) with finite verbs. Consider the following, which is based on a sentence I think I read in Ian McEwan (though don't quote me on that!):

"He remembered how she had pulled away, turned and looked angrily at the person opposite."

This is a very rough recollection but it serves my purpose. As you can probably see, the above is ambiguous. Do the finite verbs "turned" and "looked..." describe the actions of the subject of the main clause, i.e. the "He"? Did this "He" remember such and such, and then turn and look (angrily) at the person opposite? Or are those actions being ascribed to the female he is remembering things about? Was it the female who pulled away who turned and looked? Are those actions part of the man's memory?

I suppose my point is that sometimes even the introduction of a finite verb is insufficient to resolve the ambiguities. This may be obvious enough, but it took me a while to realize it when I read your post.

I suppose what I am asking, therefore, is this: how would you resolve the ambiguity in the sentence I just quoted? I can only think of one solution: repeating the personal pronoun: "He remembered how she had pulled away, how SHE had turned..." The only snag is that you then have to repeat "how" and that could be irritating for the reader.

Anyway, I hope you can suggest some solutions to this ambiguity! Thanks

DC said...

I wasn't suggesting that finite clauses are never ambiguous, but only that they are less likely to occur because of the extra explicitness.

In cases where there is an ambiguity, you have to identify the cause before you can suggest a remedy. In the present example, the default analysis is the one which respects proximity. she 'belongs' to pulled etc in the same way as he belongs to remembered. So if you want he to belong to pulled etc you have to do something different to counteract the influence of the she. Apart from a complete restructuring, this can only be done using structural parallelism:

He remembered how she had pulled away, how she had turned, and how she had looked angrily at the person opposite.

If you want the turning and the looking to be part of the same structural unit, you have to restructure a bit more:

He remembered how she had pulled away, and how she had turned and looked angrily at the person opposite.

This sort of sentence repair is well recognized in the usage guides. Fowler, for example, discusses them at length in his Dictionary of Modern English Usage in the entry on parallelism. He maintains that structural balance is important if ambiguity is to be avoided. It produces a longer sentence, but Fowler at least would deny that this is 'irritating'. For him, the elegance of the result is enough justification.

Anonymous said...

thanks for that. I thought the sentence might need to be recast. It's just that in a work of fiction, it can be nice to have punchy, non-repetitive sentences (though anaphora has its uses obviously!).

And "he remembered how she had pulled away, turned and looked at the person opposite" at least has the merit of being full of action and non-repetitive. I take it that what you are saying is that the reader should probably INFER that it is "she" who turned and looked at the person opposite, since "she" is closer to the finite verbs than "he" is. I also believe that the writer of the sentence meant to say that. I think what I am getting at is that he (viz. the writer) took a chance. He realized that his sentence was ambiguous but assumed that his reader would take him as he wanted to be taken, perhaps by following the sort of "proximity principle" that you allude to.

So even if Fowler and others would recommend re-writing the sentence to remove the ambiguity, I think the writer would be happy to KEEP the ambiguity in the interest of vivid prose. Perhaps you can tell me whether you agree with this idea.

DC said...

Absolutely. I bet if one went through a lot of literary writing one would find all kinds of theoretical ambiguities - but they are unimportant because the context resolves them. The result is the kind of stylistic opportunity you mention. Yes, it's a risk - but good writing always takes risks.

Anonymous said...

Thank you very much! I'm so glad you agree with the idea that writers (of fictional prose) often choose vividness over unequivocal grammar and syntax.

It's a fascinating idea (for me at least) and may serve to distinguish fictional from non-fictional prose. If it were a piece of non-fiction I think I'd have some sympathy for those who would follow Fowler and ask for a re-write. But in FICTION it is (to my mind) understandable that a writer should opt for vividness of prose (even if it is purchased at the price of ambiguous syntax!!)

Anyway, thanks for helping me develop the point I had in mind!