Many grammarians go about their days maligning ambiguity. Don’t use while when you mean although, they say, because it’s ambiguous. Don’t use since in place of because either, they say. And so on. If they were right, then everyone would be confused by these two sentences:
(1a) Since I eat the right foods in the right combinations, I’m not focused on calorie restriction.
(1b) The Oscar-winning director tells the story of Venezuela’s “peaceful revolution” since Chavez came to power in 1998 […]
But people aren’t confused, because the clauses readily disambiguate since. (1a) uses the present habitual I eat, which prevents the “ever since” meaning from making sense. (1b) uses the past perfect, which would allow for either meaning, but there’s nothing in the sentence that a “because” clause could attach to, so the “ever since” meaning is the relevant one.
In general, these concerns about ambiguity are actually concerns about potential ambiguity, where someone intentionally misreading the sentence or not paying a lot of attention to it could misread it.* These situations usually don’t result in actual ambiguity for reasonable readers. That’s not to say there are never ambiguities, but only that these ambiguities are usually much less of a problem than prescriptivists claim.
(2) In a second term, Carter might have moved the course of government toward the left, but since Reagan won the election the nation’s political movement has been toward the right instead.
When it is important that the reader gets exactly the meaning you desire, it is important to remove ambiguity, and at those times you’d want to, for instance, replace since in (2) with because or ever since. When the distinction is either obvious or unimportant, there’s no reason to change it. And the problem is that trying to make language completely unambiguous often comes at the cost of readability and comprehension:
(3) Upon such default, and at any time thereafter, Secured Party may declare the entire balance of the indebtedness secured hereby, plus any other sums owed hereunder, immediately due and payable without demand or notice, less any refund due.
That’s legalese; an officiously precise form of the English language that is borderline incomprehensible to those not trained in its tortuous wendings. Although there is little ambiguity in (3), it’s very difficult to extract the meaning, and the sentence seems bloated. But just try shortening or clarifying the above sentence without re-introducing an ambiguity, and you’ll see the difficulty: languages are not built for precision. And, in fact, ambiguity in language is not a bug, but a feature. This is a point nicely summarized by Frederick Newmeyer in a paper that I otherwise disagree with heartily, Grammar is grammar and usage is usage (PDF):
“The transmission rate of human speech is painfully slow […] less than 100 bits per second—compared to the thousands that the personal computer on one’s desk can manage. A consequence is that speakers have to pack as much as they can into as short a time as they can, leading to most utterances being full of grammatical ambiguity […] For that reason, humans have developed complex systems of inference and implicature, conveyed meanings, and so on. […] Stephen Levinson phrased it beautifully: ‘[I]nference is cheap, articulation expensive, and thus the design requirements are for a system that maximizes inference’ (Levinson 2000:29).”
[Emphasis mine.] Ambiguity is useful, as ambiguous sentences can convey the necessary information just like unambiguous sentences, but in fewer words. The reader, listener, or whoever you’re directing your language to is then able to use their knowledge of context and implicatures to determine the appropriate interpretation (this is the “inference” process). A great example of this (again from the Newmeyer paper, but originally from Martin, Church, & Patel 1987) is (4), which has 455 possible parses, many of which yield different meanings.
(4) List the sales of products produced in 1973 with the products produced in 1972.
And yet, given a bit of context, and some knowledge of what one is trying to do in the situation in which this sentence is uttered/written, you are able to pretty quickly figure out which potential meaning is the best. Trying to make the sentence perfectly unambiguous would only drown the reader in words.
Summary: Pick your battles against ambiguity. Where ambiguity is truly detrimental, put forth the effort to clarify, to root out plausible ambiguities and remove them. Where ambiguity is tolerable, it can be better to leave it in to keep from exhausting yourself and your audience.
[If you’re interested in more on potential vs. effective ambiguity, Arnold Zwicky had a post on Language Log from 2008 discussing this topic. Now that I look at his post again, I’ve realized that most of what I said here, he already said there, plus more.]
—
*: I know some of you in the audience are editors, and I’ve had a few editors explain to me that their job consists in part of idiot-proofing writing. This requires you to try to make it as easy on the reader as possible, and to assume that the reader will fall into whatever garden paths and other meaning pitfalls are possible. Removing the potentially ambiguous situations might be seen as a step in this task. That’s a fair counter-point, but it does not compel a change, and the change must be weighed against the considerations. Avoiding ambiguity that requires the reader to wantonly misinterpret is less crucial than avoiding easy-to-fall-into ambiguities.
8 comments
Comments feed for this article
January 21, 2010 at 2:12 pm
Vance
I broadly agree, but there’s another aspect of the speech situation beyond what Newmeyer points out: the chance for interrogation and correction. This means we can more safely risk ambiguity when speaking — we can do “concise first, precise next”. In writing, we don’t have this chance, and I think that tips the scales toward “precision” a bit.
January 22, 2010 at 9:21 pm
mike
If a change from potential ambiguity to no ambiguity has no real cost — e.g., changing “since” to “because” — why not do it?
In my job, one of our important “readers” is machine-translation software that performs a first pass on text before that text is handed over to actual humans to finish the translation. Thus we edit with this dumber-than-normal reader in mind. As is pointed out to us frequently, making things clearer for the lunkhead software also happens to make it clearer (or less likely to be less clear) to humans. (Which, I should point out, also include a substantial number of folks who do not read English as their first language.) Some complain that the result is clunky, and that’s not un-right :-), but as one of my former bosses (herself a technical writer) used to be fond of exhorting us, “We’re not writing literature here!”
All this said, you are of course correct, and there are times when it even takes some work to find the ambiguity, so obvious is the intended meaning. Still, as Vance points out, you only have one shot at it when the reader can’t say “Come again?”, and you have considerably restrained meta-text helpers like intonation to help the reader along.
This is sort of an aside, but I had an example this morning where it took me several readings to even find the ambiguity. But then I couldn’t get it out of my mind, so I had to recast the sentence:
“I have not shown this technique in order to improve the clarity of the example.”
January 22, 2010 at 10:16 pm
Vance
So, mike — why did you show the technique, if not to improve the clarity of the example?
January 23, 2010 at 2:13 am
Gabe
Those are some great points. One of the most frustrating things for me in writing is the inability to interact, and as such I load my writing with qualifications and clarifications that really bog down the prose. Also, ML and second-language readers are certainly good reasons to err on the side of clarity.
The specific because/since example is sort of a bad one, because there isn’t any clear reason not to use “because” in these situations. However, the fact that “since” began to and has continued to serve this purpose suggests that there is some reason to use it in some situations. (I have no idea what this reason would be, but I do notice that “Since I’m going to the store anyway…” is a little more natural to me than “Because I’m going to the store anyway…” This is a weak preference that I expect you might not share.)
I think this post overshoots my true stance a little, as you’ve both noticed. The real trick is to balance the unclarity that results from ambiguity against the unclarity that comes from clunkily unambiguous writing. But then you also have to consider the time it takes to edit out the ambiguity as well, which is more costly in speech and ephemeral writing than in longer-term tasks like technical writing. In technical writing, the cost of reducing the ambiguity is much more easily offset by the payoff of easier reading and translations.
By the way, Vance, I love the “concise first, precise next” line. Is it yours? Mind if I use it too?
And by the way Mike, I barely made it through your sentence before both meanings hit me. I’m sure it was only the context that prevented you from seeing both sides right away. Which was the intended meaning?
January 23, 2010 at 12:04 pm
mike
@Vance, well, given the writer whose document this appeared in, I could have imagined — though I did not believe this uncharitable thought to be his intention — that he didn’t show the technique because he was lazy. :-)
@Gabe, here’s an example where “since” and “because” are, I believe, not interchangeable:
-Since you’re just standing there, why don’t you help me move this table?
-Because you’re just standing there, why don’t you help me move this table?
What’s interesting to me is that “since” here does a have causative, as opposed to chronological, sense to it, yet “because” doesn’t work. (Or is “since” chronologic here after all?)
But again, overall your point is well taken, which is that there are constructs that can, often with some effort, be taken ambiguously, but in practice, the ambiguity doesn’t happen. I’m not sure if you’d consider this still within the scope of your point, but the position of “only” seems to illustrate this:
-The cows only come home at night
vs
-The cows come home only at night
The first does not mean, unless you take ambiguity to an absurd extreme, that cows only come home at night, they don’t go dancing instead. (Tho I do like that image.)
January 23, 2010 at 11:24 pm
Vance
Help yourself to “concise first” — odd nobody should have typed those four words in that sequence before. (You must have configured your blog so comments aren’t visible to Google.)
I struggle with clarity in writing too. I tend to make compact snarls that are hard to unpack, or to revise them into flaccidity. Still, I worry about “potential ambiguity” — I don’t want to waste my reader’s energy by inducing her to lurch toward one interpretation, only to have to stop and turn toward another. Even if there’s no doubt where she’ll end up, I don’t want to lead her down the garden path.
February 2, 2010 at 12:23 pm
Jonathon
As an editor, I’ve always been a little bothered by the idea that the point of editing is to remove any and all potential obstacles, no matter how remote the possibility that anyone is going to be tripped up by them. One problem is that changing it doesn’t have no cost. Every change made takes time, and that’s a cost that’s passed on to the reader. Is it worth it to them to change every causal since to because? I’m just not convinced it is.
August 15, 2013 at 8:05 am
Don’t sweat it: Since and because | Madam Grammar
[…] The reason for this insistence is that “since” might be ambiguous, since it can have either a time sense or a causal sense. But there are only a few instances where “since” may truly be ambiguous (“Since you came over, I feel better” — does “since” here mean “because” or “from the time that”?); in fact, most sentences containing “since” have enough context to make the meaning clear. […]