Do we need to protect uncooperative readers?

goofy recently posted at bradshaw of the future about momentarily and some strange advice Grammar Girl sent out about it. Her advice:

“Don’t use momentarily to mean “in a moment”; you may confuse people. If you mean in a moment, say or write that. There’s no need to use momentarily in such cases, and doing so will irritate language purists.”

A quick note first: both the “in a moment” and “for a moment” meanings of momentarily have been around for 140 years, so the purists are completely unjustified in their complaint. Also, sure, there’s no need to use momentarily here, but then, there’s no need to ever use any given word. You can always paraphrase or re-write the sentence.

But the real question is two-fold: whether the benefits of using a questionable word outweighs its costs, and whether there’s a better word. You might think of this as a satisficing condition and an optimization condition.* And I suspect — although I don’t know if anyone’s studying this, or what they’ve found — that there’s some sort of a switch-off between the two methods depending on what production task you’re doing. When speed is one’s primary concern, presumably it’s sufficient to check that the word is beneficial; only when one has the luxury of time does full optimization kick in.

So is momentarily costly — i.e., will it confuse readers? goofy makes a good point about the potential confusion:

“If it’s more common for people to use momentarily to mean ‘in a moment’, then why advise people not to use it that way? It seems that Grammar Girl is essentially saying ‘don’t speak like everyone else in your speech community speaks.’ This seems counterproductive. […] it might confuse people – but if most people already use it that way, why should it be confusing?”

He gives the example of a pilot saying “we’ll land momentarily”, and notes that no one except for an uncooperative speaker will think “that means ‘for a moment’!” But one might harbor doubts. Maybe no one will end up with that interpretation, but maybe they’ll be distracted by it during interpretation. Yeah, that’s certainly possible — but listeners are more adept at ignoring irrelevant ambiguities that we tend to give them credit for.

The famous example from introductory linguistics classes of this is Time flies like an arrow. The first time someone sees this sentence, it just sounds like a standard aphorism, and the only meaning they’re likely to seriously consider is “time moves in a swift manner, akin to an arrow”. But this sentence is ambiguous, of course, as almost all sentences are. Many of the words have different senses and different parts of speech that they can take on.

If we switch from a Noun-Verb-Preposition reading of time flies like to an Noun-Noun-Verb one, we get: “‘Time flies’ (as opposed to houseflies or gadflies) appreciate an arrow”. There’s also a Verb-Noun-Preposition reading, yielding an imperative: “as though you were an arrow, record the time the flies take to complete a task”. There are other interpretations, too, but none of these is likely enough, given our world-knowledge and parsing probabilities, to register in our minds. We can reasonably expect that Time flies like an arrow will be correctly understood, without time lost to alternative interpretations, by any audience that isn’t actively looking for implausible interpretations.

So too should we expect momentarily to be correctly understood; claiming to have difficulty with it marks the complainer, not the speaker, as the one who doesn’t understand language. As an editor, one generally ought to foolproof writing, looking for and eliminating potential (even if fairly unlikely) misinterpretations. But there’s a difference between editing to protect fools from ambiguity and editing to protect uncooperative readers from ambiguity. The former is difficult, but generally doable. The latter is often simple, but generally worthless.**

Let me conclude with a good question from Jonathon Owen in the comments on goofy’s post:

“And if the problem is simply that purists will be annoyed, why not direct our efforts to teaching the purists not to be annoyed rather than teaching everyone else to avoid offending this very small but very vocal set of peevers?”

—

*: “Satisificing” is an idea I’m fond of, though one that doesn’t get talked about much outside of human decision-making tasks. In the familiar optimization strategy, you’re trying to find the best of all possible options, whereas a satisficing strategy is just looking for any option that’s better than some threshold. For instance, if you go to the store with two dollars and need to buy milk, you can optimize by comparing multiple sub-$2 cartons before picking the best of that lot, or you can employ a satisifice by buying the first carton that costs less than two dollars.

Satisificing is generally faster and, if I remember my undergrad psych classes correctly, is common in human decision-making processes, especially when time is of the essence.

**: One exception, presumably, is in legal writing/contracts.

Can we all admit none of us know what we’re talking about?

First off, if you haven’t already heard, the AP Stylebook finally dropped its objection to sentential hopefully (i.e., the “it is hoped” meaning), thanks in no small part to John McIntyre’s agitation. Another shibboleth bites the dust, hooray.

If you’re harboring any doubt about the wisdom of this move, cast it to sea. Living with sentential hopefully isn’t giving into modern ignorance; it’s giving in to traditional usage. Emily Brewster points out this 1999 article from Fred Shapiro in American Speech. Smack on its first page, we’re given a quote from Cotton Mather in 1702:

“Chronical Diseases, which evidently threaten his Life, might hopefully be relieved by his removal.”

In previous work, Shapiro traced it back to 1851, and here’s an example I found in Google Books from 1813. So it’s not some new and insidious usage, though this is often claimed.

And it’s not like sentential adverbs are inherently bad, either; witness well-regarded members of our lexicon such as frankly, happily, thankfully, or luckily, each of which can be used at the start of a sentence with nary an eyelash batting. The truth is that accepting sentential hopefully is not giving in to a tide of misusage but rectifying an objection that should never have been raised.

Mary Elizabeth Williams doesn’t see it that way. In a piece at Salon, she views the AP’s leniency on hopefully as capitulation. She thinks the AP’s giving in to the uneducated masses instead of remaining the guiding and educating light it ought to be. It’s another sign that no one knows about language anymore, and no one cares about it, not even its presumed defenders. She closes with this regret:

“Language keeps evolving, and that’s fine and natural. Yet as it does, I’ll still gaze hopefully toward a world in which we battle over our words and our rules because we know them so well, and love them so much.”

Hey, you and me both. But here’s the thing: it’s not just everyone else who doesn’t know what they’re talking about. Williams needs some work to get into her own dream world as well. While she lectures us who see nothing wrong with sentential hopefully about how we ought to have a better language arts education, she explains her disgust with it by exasperatedly pointing out:

“‘Hopefully’ is an adverb. An adverb, I tells ya […]”

Ok, cool, but I’m with the red-headed guy here:

She’s really stressing the hopefully-is-an-adverb point, which is fine, but no one’s saying it’s not. The sentential usage is an adverbial usage. If you think that people think that hopefully can be used in a non-adverbial context, then you’re not in a position to be disparaging anyone’s knowledge of English.

So it’s strange that Williams is complaining about people who don’t know enough about English causing the acceptance of sentential hopefully, since the people opposing sentential hopefully apparently don’t know English either. A person who really knew about the history of usage in English would know that sentential hopefully is a member of a large and grammatical class of sentential adverbs, that it’s been around for centuries, that, in short, there’s nothing wrong with it. It engenders some distaste from the uninformed and it’s perhaps a bit informal, but there’s no reason why it should be so despised. Many of the people who condemn the rabble for not knowing the rules or history of English don’t know them themselves.

Let me cast the mote out of my own eye first: I don’t either. I was gobsmacked by the Brewster/Shapiro/Mather finding; in an earlier post talking about sentential hopefully, I only had it going back to 1932. There is a lot that any one person won’t know about a language. But one key difference between people who claim to care about how language works and those who actually do is that the latter category will investigate a usage before accusing it of being bad grammar.

So yes, it’s a shame that so many people don’t care about language. But the problem isn’t that alone; it’s also that too many who do care about language care about it wrong. They’re not interested in the actual data; they’re interested in what they decided the language ought to be. They argue their points in a world apart from actual usage, based on the logic that they presume underlies language. When they do cite usage, it’s with a heavy confirmation bias. And their complaints are run through with this strange — and to me, infuriating — willingness to grant themselves pardons from their otherwise zero-tolerance policy. Williams groans at people who use nauseous for “nauseated” (standard since the 19th century, BTW) or who write gonna, but then gladly admits that she uses stabby and rapey*.

This isn’t caring about language; it’s caring about feeling superior.

—
*: Which, by the way, seriously?

On feeling good, well, bad, or badly

Last post, I argued that “I’m good” is a perfectly acceptable response to “How are you?”, because the linking verb am takes an adjective, and good is an adjective. “I’m well” is a fine response as well, although I personally prefer “I’m good” as it seems to refer more to my state of mind than my state of health.

In this post, I’d like to take a little extension into the related response, “I’m feeling good”. I’m also going to talk about its darker cousin, “I’m feeling bad” and explain the difference between bad and badly in this context. Originally, I’d hoped to talk about “I’m doing good” as well, but I’m going to shunt that off to yet another post.

Let’s start with “I’m feeling good”. In short, it’s fine, and here’s why. Feel is like be in that it can be a linking verb, taking a predicate adjective that modifies the subject. When one says “I’m feeling good”, good modifies I, not feeling. Some might argue that “I’m feeling good” is ambiguous (“content” vs. “moral”), but I just don’t see it, and anyway, I’m going to show that that ambiguity isn’t a big problem.

Going deeper on the adjective point, it might seem a bit weird to say that good describes how you are, not how you feel. Let’s compare the use of a somewhat clearer adjective/adverb pair: crazy/crazily. (1a) means that I’m a bit weird, whereas (1b) is itself a bit weird:

(1a) I’m feeling crazy. [adjective, modifies me]
(1b) ?I’m feeling crazily. [adverb, modifies feel]

What makes “I’m feeling X” a more grammatically interesting structure than last post’s “I’m X” is that, unlike with be, you can modify the verb feel with an adverb. (1b) isn’t ungrammatical; it’s just uncommon. Suppose, for instance, you’ve lost your autographed Harvey Haddix baseball in a ball pit at Chuck E. Cheese’s. You dive in and start feeling around for it, but blinded by fear of losing the ball, you’re feeling crazily amongst the balls.

When you adverbially modify feel, it’s a different sense of the verb from the one that takes a predicate adjective. In “I’m feeling crazy”, the verb refers to one’s sense of interoception, one’s perception of oneself. In “I’m feeling crazily”, the verb refers to one’s sense of exteroception, one’s perception of the outside world.* (The exteroception verb sense can also take a predicate adjective, as in “gelatin feels squishy”.)

So if you’re concerned that “I’m feeling good” is ambiguous (i.e., are you moral or at ease?), you ought to fear ambiguity in “I’m feeling well” (i.e., are you talking about your intero- or exteroception?) just the same. Ambiguity is pretty unavoidable sometimes. As a result, there’s no grammatical reason not to say “I’m feeling good”, although it might not be standard depending on your age and dialect. And “I’m feeling well” is fine, too.

Moving on to “I’m feeling bad(ly)”, the standard contemporary form is I feel bad. However, I feel badly, though non-standard, is pretty common and pretty robust. In fact, at various times it has been argued to be the standard form; Charles Dod wrote a nice article in 1875 arguing that I feel badly was being used as the standard form, in order to avoid the supposedly ambiguous I feel bad. He continued by arguing that here badly was functioning as an adjective even though it looked like an adverb. His discussion raises some important points, many going beyond the mere adjective/adverb distinction:

The expression [feel badly] is needed; hence it is correct. We must allow the speaker to explain what he means, and not let the grammarians force upon him a meaning which he rejects. Let us then review our grammatical principles; if we cannot adjust the phrase to our principles, we must adjust the principles to our phrase. It is a fact, that respectable and well-educated people do say, “I feel badly.” Now let us explain the fact. We may have to widen our generalization to let this fact in; but being a fact, we cannot leave it out of view in any theory we may form. We may be sure that we have overlooked something in our analysis of the phrase, “I feel badly.”

Dod argues that feel badly is necessary because feel bad can’t mean what we want it to mean. I disagree with him there, at least in current American English, but the truth is that many people — even the educated — do use feel badly where feel bad would seem to be prescribed. This is a fact that cries out for an explanation, and merely claiming that its users don’t know their English isn’t a very good one.

Let me offer a proposal. As with “I’m feeling crazily”, feel is very strongly biased against adverbial modification in this usage, where one is reporting one’s own feelings. For instance, if I hear someone say “I feel badly”, it’s so unlikely to me that they’re complaining about the incompetence of their exteroception that I find it difficult to get that interpretation of the sentence. Pretty much the only interpretation that comes to mind is “I feel bad”, unless the context suggests that the incompetent exteroception meaning is likely. So even if you hear “I feel badly”, it’s hard to misinterpret it.

As a result, I feel badly exists in a sort of weird situation, where there is very little to suggest that it is erroneous, and a decent amount to suggest it is correct. The meaning is biased against misintrepretation, the verb can take either an adjective or adverb in different situations, bad is ambiguous (or so Dod says), and there’s a general bias to err on the side of adverbs over adjectives. Add those up, and it’s easy to expect I feel badly to be an especially persistent non-standard usage. That said, if you want to follow Educated American English, I feel bad is the better form.

The “I’m good” outrage is nonsense

This is such a common complaint that I’m only going to offer a single example of it, and leave it up to you whether you want to waste part of your life looking up other examples. From the BBC’s idiotic list of “Americanisms”:

16. “I’m good” for “I’m well”. That’ll do for a start.

There is nothing wrong with “I’m good”. And yet, this is the sort of grammatical myth that not merely persists, but pervades. One of my best friends in college ragged on me for it. One of my current friends (an English teacher no less) subtly corrects me for it regularly.

There are a few reasons why people might think that I’m good is incorrect. The most prominent, the one I’m often given as justification, is that good is an adjective and well an adverb. That’s all well and good, but am is a conjugated form of to be. To be is a linking verb here, which means that it takes a predicative adjective, not an adverb. We say things like I’m hungry, not I’m hungrily. An adjective is what you need here, without question.

Of course, well isn’t only an adverb; it can be an adjective as well. That leads to the next argument against I’m good: that good is an adjective, but it’s the wrong adjective. For instance:

When you ask an American: “How are you today?”, they say: “I’m good” (Meaning: I’m a good person) when they should use “I’m well” (Meaning: I’m fine or healthy or something like that).

But to get the “I’m a good person” meaning out of I’m good, you have to try to misinterpret it. Sure, saying I’m good can be interpreted as “I’m not evil”, but that’s far from the only possible meaning, and it’s hardly the most reasonable. I don’t want to be condescending, but even a non-native speaker of English is aware that good has a lot of possible meanings. Here are two from the Oxford English Dictionary:

1. Of persons, as a term of indefinite commendation.
2. Such as should be desired or approved, right, satisfactory; sound, unimpaired; not depressed or dejected.

Those senses of good, which date to 1154 and 1175, respectively*, are more likely intentions when responding to “How are you?” than an unsolicited assertion that one is a moral human being. To say that the “moral” meaning is either the only acceptable one or the most reasonable one in this context is to say that you do not have a good grasp of the English language.

So I think that that establishes why I’m good is acceptable, and really does mean “I’m fine”. But perhaps I’m well is more acceptable? Hey, maybe for you it is, and if it is, godspeed. But for me, the two forms have significantly different meanings, and in general I mean to say that I am good when I say I’m good.

I’m well means that I am healthy, which I almost always am if I’m wandering around talking to people. When people ask, “How are you?”, they’re not, in general, inquiring about your state of health but rather your state of mind. Thus I respond that I am good, in that second definition above, feeling right, satisfactory, unimpaired, and neither depressed nor dejected. I do not respond that I am well, because I think that’s pretty obvious, and if it’s not obvious I’m well, it’s likely because I am unwell.

I think that most people feel the same; when my friends tell me that they are good, they tend to follow up with something like “I got a new video game” or “I’ve been enjoying this weather”, indications not of good health but of good feelings. That’s not to say that you shouldn’t say that you’re well; you are welcome to. This is only why I don’t say I’m well.

I want to talk about two similar situations — I’m feeling good and I’m doing good — as well as whether I’m good is too vague, but I’ve gone on long enough. I’ve put together a second post discussing I’m feeling good and I’m feeling bad, and hope to finish off with one on I’m doing good in the future.

Summary: I’m good is correct, because am is a linking verb, taking an adjectival predicate, and good is that adjective. I’m good means that one is fine, in good spirits, etc. I’m well is fine too, but I find it to focus more on one’s health than general state of being.

—
*: And, of course, they’re attested through the modern day.