Tag Archives: English grammar

I only wanted to explain this

This week is a bit of a double header for me: two articles published in different places at the same time. Yesterday I posted a link to my latest article on TheWeek.com; today I’m posting (in its entirely) my latest article on the Editors’ Association of Canada’s national blog, The Editors’ Weekly.

Adverbs are a problematic and much-maligned class of words. Linguists often have trouble explaining exactly why they go where they go. Some sorts of adverbs are baselessly despised (hopefully, people will eventually get over those hangups, but I’m not hopeful). Some people think adverbs should be excised from writing altogether.

I’d like to cover all the misconceptions relating to adverbs, but that would make for very long reading. So today, I’m only going to talk about the placement of one specific adverb.

Those of you who read that and thought, “That’s bad grammar — it should be ‘I’m going to talk only about’ or ‘I’m going to talk about only,’ ” raise your hands.

Hands raised? Keep them there. As long as they’re raised they won’t be causing any trouble.

Oh, there’s no mistake in thinking that careful placement of only has much to recommend it. The mistake is in thinking that putting it in the default position, right before the main verb, is an error unless it’s limiting the main verb specifically.

Let’s look at an example sentence:

I only wanted to bake three cakes.

Without context, all you know is that baking three cakes is the full scope of what I wanted. You don’t know if the thrust of it is “I was not happy about baking the fourth one” or “I didn’t want to have to decorate them” or “I had no desire to bake three pies too.”

But do you know what the supposed only correct interpretation of that is? It’s that I only wanted to bake the cakes — I didn’t actually do it.

Does that seem odd? It should. It’s a made-up rule with no correspondence to reality. As Matt Gordon recently said on Twitter, quoting a student paper he was marking, “If ‘this grammatical distinction has confused writers for centuries,’ maybe it’s those trying to impose the distinction who are confused.”

In truth, the only way you can make that only about wanted is to emphasize wanted, or to phrase it differently. That position is the default position for only regardless of what aspect of the action it is limiting. But the rule-thumpers insist that you must move the only right next to what it limits:

I wanted only to bake three cakes.

I wanted to bake only three cakes.

I wanted to bake three only cakes.

Ah, wait, the last one isn’t even usable. You have to do this:

I wanted to bake only three cakes.

But that blows the “rule” right away. If you can do that, you can do this:

I only wanted to bake three cakes.

You can do likewise for the other restrictions:

I only wanted to bake three cakes.

I only wanted to bake three cakes.

The fact that we can — and do — shift the emphasis like that without moving the only is proof that this is a standard feature of English grammar, and that the word feels natural in the default position and can work there quite well. Anyone who tells you that “I only wanted to bake three cakes” is wrong and must be “I wanted only to bake three cakes” has not correctly analyzed the syntax of the sentence. He or she also has a tin ear, and yet thinks himself or herself a better writer than the many respected authors throughout the history of English who have used only in the “wrong” way.

This is not to say that you can only put only before the main verb — or, if you prefer, it’s not to say that you can put only only before the main verb. Its mobility gives you a very good tool for clarifying the meaning. But the availability of the default position gives you a tool for adjusting the rhythm and the naturalness of the sentence when the meaning is clear anyway. Why limit your toolkit unnecessarily?


Today’s word tasting note is a guest post by Vancouver editorial genius Iva Cheung.

“Oh yeah? Well you guys may have won this time, but my team’s still the winningest!”

That’s how I imagine the first utterance of winningest, and the fact that the speaker, in my mind, could just as well have been a four-year-old as a drunken sports fan in a pub is telling. Winningest has a decidedly juvenile and unsophisticated ring to it, and, judging by the comments on Merriam-Webster’s entry for the word, a lot of people hate it, calling it a “made-up word” and a “lazy degradation of modern language.”

So why is it so objectionable?

Well, first, It hasn’t been around for all that long. Although the Online Etymology Dictionary claims winningest appeared in the written record by 1804, without seeing a reliable example, I’m more inclined to believe the Oxford English Dictionary and Merriam-Webster, which both trace the word back to the early 1970s. According to Webster’s in 1974 it appeared in The State newspaper in Columbia, South Carolina, which described the Maryland-Eastern Shore as the “winningest college basketball team in the nation.” Since the word was coined, it’s been used almost exclusively in (North) American sports pages.

Newness alone isn’t a good reason to find a word immature, though; after all, plenty of much newer words have seamlessly slipped into everyday usage. The resistance to accepting winningest as a “real word” probably comes from its weird morphology and incongruous semantics. We don’t usually slap the superlative -est morpheme onto the -ing morpheme—it seems like a mistake a young child might make—and winningest may be the only English word to feature this odd combination. Sure, present participle forms can serve beautifully as adjectives, but even words like charming and stunning—which we probably use more often as adjectives than as verbs—are compared using more and most rather than the -er and -est suffixes.

And speaking of which, if winningest exists, why doesn’t winninger? This lack of a comparative counterpart to winningest contributes to its oddness. What’s more, you could easily argue that the superlative doesn’t even make sense. We tend to think of winning as half of a dichotomy, not the end of a spectrum. Winning isn’t really a gradable adjective; if you’re not winning, you’re losing.

Yet this semantic mismatch is why winningest has found a niche as a functional word, despite the many reasons it shouldn’t exist. It doesn’t actually mean “most winning,” does it? To be the winningest means to have the most wins or victories, or to have the most success, typically in a sports context. And in that context, it’s unambiguous, succinct. You may wrinkle your nose at a phrase like “the winningest team in the league,” but you’re unlikely to be confused by it, and any other way of expressing the same concept would simply take more syllables.

Despite the objections to this “non-word,” it seems to be slowly seeping out from its sports confines into the rest of the world; even law firms are seizing the opportunity to declare themselves “the winningest.” The question is whether it’ll remain a one-off anomaly or spawn a new, productive way of affixing. Will we one day read about the eatingest competitor at the hot dog contest or the flyingest pilot in the air force?

I myself think people get too bothered about it

My latest article for TheWeek.com is about something some people get quite bent out of shape about: use of myself for purposes other than the reflexive. See what I have to say, and why:

Myself, I don’t see a problem


Topics, we front them

Featured on The Editors’ Weekly, blog of the Editors’ Association of Canada

English is normally a subject-verb-object kind of language, but there are some interesting exceptions, especially in casual contexts. Consider examples such as the following:

Poodles, we walk them. Labradors, they walk us.

Chickens we have; roads, not so much.

Not one of the clauses above follows standard English sentence order, and yet we understand them. Of the four clauses, the first is object, subject-verb-object (pronoun); the second is subject, subject (pronoun)-verb-object; the third is object-subject-verb; and the fourth is object, adverb – no subject or verb overtly expressed at all. All four are examples of topic fronting, sometimes called left dislocation. (Note that we normally use a comma to set off the topic when it is followed by a complete clause or an elliptical statement, but not usually when the sentence is not syntactically complete without it.)

So why do we depart from our usual syntactic structure? We do it to maintain our preferred information structure. When we communicate information, we usually prefer to introduce a topic and then comment on it with new information. This can be especially useful when we are contrasting two topics. We don’t have to do it; we could rewrite the above sentences as follows:

We walk poodles. Labradors walk us.

We have chickens; we don’t have roads to nearly the same extent.

The first example works well as rewritten, though it loses its folksy feel; it also loses the parallelism of topics, but it highlights the inversion, which has its own effect. The second really, um, fails to cross the road. And even in a structure such as the first, you get poorer results if you can’t make a useful inversion:

Shirts, we mend them. Shorts, we toss them.

We mend shirts. We toss shorts.

You can see it loses some of the contrast effect. In speech you can emphasize the topic structure using intonation; in print that option is not as available.

Now look at the preceding sentence (“In speech…”): it contrasts the adverbial prepositional phrases thematically by moving them from the end to the beginning of their clauses – and no one would object to it. So why does it seem somehow incorrect to do it with nouns?

It’s not because it’s some new error, or a structure borrowed from another language (though some languages do normally introduce topics first, regardless of syntactic role). In fact, as Mark Liberman has discussed on Language Log, left dislocation has existed in English as long as there has been an English for it to exist in.

What has happened is that it has fallen out of use in recent centuries. Like some other formerly standard things, such as double negatives, double superlatives, use of ain’t, pronunciation of –ing as –in, and use of ’em in place of them, it has come to be seen as nonstandard, especially when there is also a pronoun filling the same role – we can sometimes get away with it as poetic when there is no pronoun:

Parsley we put on the plate, sage we leave on the plain, rosemary and thyme we drop in the pot.

The nonstandard air of left dislocation gives a useful means of making your text seem casual or colloquial – as well as keeping a nice clear parallelism. On the other hand, if you need to seem more formally correct, you still have a means of putting it in front acceptably: just turn the parallel nouns into parallel prepositional phrases.

With poodles, we walk them; with Labradors, they walk us.

Between you and I, could you take a picture of my friends and I?

My latest article for TheWeek.com deals with a popular issue: pronouns in compound objects (the things in my title, above, that may have your teeth grinding). I talk about not just the rule but why so many people find it so hard to stick to it. The article is

‘You and I’ vs. ‘You and me’

In the past couple of days, I’ve also added a couple of longer posts on grammar. One of them tears to bits a web page of grammar advice: Why it’s best to leave grammar advice to experts. The other does a detailed dismantling and analysis of a potentially confusing sentence from a recent award-winning book: A little Hellgoing sentence mechanical deconstruction.

A little Hellgoing sentence mechanical deconstruction

In Lynn Coady’s Giller-Prize-winning book Hellgoing, one of my editorial colleagues has spotted the following sentence:

She tried not to stare at Marco while he spoke to whomever he was speaking to, but she wasn’t succeeding.

Does that look quite right? Are you not quite sure? It’s the sort of sentence that you might not think much about unless you stop and look at it, but if you do stop and look at it it might start to drive you a little crazy.

So let’s go into this little Hell and take apart this sentence and see how it works.

Core subjects and verbs:

She tried . . . but she wasn’t succeeding.

Tried what?

…tried not to stare…

Not to stare at whom?

…not to stare at Marco…

I’ll abbreviate “She tried not to stare at Marco” as STNSM. It’s syntactically fine, I think we can agree.

Now: when?

STNSM while he spoke…

OK, spoke to whom?

…he spoke to X

where whoever X was, he was speaking to him.

So X was the person he was speaking to. Whoever that was.

Now here is how that plays out in that bit of the sentence. We need a relativizer:

…he spoke to {the person} {[relativizer] he was speaking to him}

What happens is that {the person} is replaced by the whole {[relativizer] he was speaking to him}.

The relativizer is “whoever” or, as the case may be, “whomever”:

…he spoke to {whoever {he was speaking to [him]}}

The “him” at the end moves up and merges with the relativizer, giving it accusative case (i.e., making it the object) (see the bottom of this post for more on this):

…he spoke to {whoever [+him] he was speaking to}

…he spoke to {whomever he was speaking to}

That’s where the confusion happens. The raising and merging into “whomever” is something that can confuse just about anyone until they learn about the underlying movements.

It could have been

…he spoke to the person to whom he was speaking

Then “the person” stays put and it happens this way:

…he spoke to {the person} [relativizer] he was speaking to {him}

The relativizer would become just “whom”, moved up from the end:

…he spoke to {the person} [whom] he was speaking to

But the “to” typically follows it up in this case:

…he spoke to {the person} [to whom] he was speaking

Notice there are two “to”s in both versions.

So let’s look at the whole sentence again and match the parts:

She tried not to stare at Marco while he spoke to whomever he was speaking to, but she wasn’t succeeding.

What are the verbs?

tried … (not) to stare … spoke (to) … was … speaking (to) … wasn’t … succeeding

The conjugated verbs have subjects:

She tried … he spoke … he was … she wasn’t

Let’s add the complements:

She tried not to stare

to stare at Marco

he spoke to whomever

whomever he was speaking to [him]

she wasn’t succeeding

There are also the conjunctions “while” and “but”, and that makes the whole thing.

She tried not to stare at Marco while he spoke to whomever he was speaking to, but she wasn’t succeeding.

So there are actually no surplus words. Each verb “speaking” requires a “to”: “spoke to” and “was speaking to”. Formal English often frowns on stranding the preposition at the end, but it’s always been an available feature of English; indeed, it requires less syntactic movement. Raising the preposition with its complement is a funny thing to do from a syntactic perspective, and linguists call it “pied piping” because it’s as though the object is a pied piper getting the preposition to come dancing along with it.

Most of that is of course rather complex and more than the average person is inclined to want to know. But editors aren’t average persons, so I have put it here for your enjoyment.

Now, here’s the bit more about the “whoever” becoming “whomever”: The reason “whoever” is “whomever” is because it’s merged with the “him”. The entire relative clause is the object, and the case doesn’t penetrate inside it. Here’s proof:

He looked at whoever was speaking to him.

It would not be correct to say “whomever was speaking to him” because the “was” requires a subject, and that is “whoever.”

As it happens, I talk about how case assignment doesn’t automatically percolate into phrases in my latest article on TheWeek.com, “‘You and I’ vs. ‘You and me’.”

Why it’s best to leave grammar advice to experts

A company called ePly Online Event Registration has, on its website, a page on common errors of grammar and word choice that people make when creating web forms: “Are You Making These Common Wording Errors on Event Websites and Registration Forms?

It’s such a pity they didn’t get someone who had any expertise on the subject to write it. You see, some of their recommendations are good, but it’s all couched in a muck of ignorance and rubbish. It does no favours to the company – nor, for that matter, to the reputations of those who give grammar advice.

Let’s have a look at what I’m talking about, point by point.

First they tell people not to use insure where they should use ensure. I actually have no factual disagreement with this point; it’s a correction I make all the time. They give an example of correct use: “Ensure you register on time.” This is indeed grammatically correct. However, it’s also a bit stilted. Depending on the tone you want, “Make sure you register on time” would be better; giving the exact time might be even better (e.g., “Make sure to register by 11:59 pm on December 23”).

Next they hop onto the which/that distinction:

“That” refers to the noun in the sentence and gives essential information about the noun. “Which” introduces a qualifier that is non-essential.

Hm. First of all, the restriction of which to nonrestrictive clauses is not a grammatical law; it is a stylistic recommendation and does not have to be followed, even in North America (let alone in Britain). Secondly, their grammatical explanation is no good: “refers to the noun in the sentence” – most sentences have more than one noun; the subordinate clause beginning with that refers to the noun right before it… when it refers to a noun. It can also be the complement of a verb: “I think that you’re wrong.”

Then there’s their use of “essential” and “non-essential.” This is not really a clear way to put it. If I say “My car, which is older than I am, is not yet 50,” it is an essential trait of my car that it is not yet 50, though the “which is” clause is not essential to the sentence structure. The real difference is that the which-clause is nonrestrictive: that is to say, it’s not further specifying which car I’m talking about; I have only one car. Were I to say “My car that is older than I am is not yet 50,” it would imply that I have more than one car, and I am restricting the scope to the one older than I. But it is actually the commas, not the which or that, that make the difference. I could say “My car which is older than I am is not yet 50” and that would in fact be grammatically correct – though against a common North American practice.

They next pick on whether versus if:

“Whether” is not interchangeable with “if.” “Whether” expresses a condition where there are two or more alternatives.

Hmm. I don’t know if that’s true. Which is to say I don’t know whether that’s true. Which is to demonstrate that I actually know it’s not altogether true: although use of if in place of whether is considered less formal, it is possible in many (though not all) places.

One example of where it’s possible is their example. This is their example of correct use: “Indicate if you are attending the dinner.” Yes, that’s right, they seem to be saying that “Indicate whether you are attending the dinner” is incorrect. This is actually gobsmackingly off-base. The formal standard would actually be “Indicate whether you will attend the dinner.” The problem with the if version is not simply its reduced formality – which can be a good thing – but the fact that in some cases it may be misread: “Do X if you will do Y” can be taken as “If you will do Y, do X” and by implication “No need to do X if you are not doing Y” – an unlikely misconstrual in their specific example, but a possible one in some cases.

Next they dive into the less/fewer matter. Now, I’ve covered this in “When an ‘error’ isn’t”; the use of less for countable items dates back a millennium and a half and has been seen even in journals of science and philology. It’s true that if you use less to refer to countables you will provoke the ire of the peevers, and thus you do well to stick with fewer in contexts with any formality. But it’s not the iron-clad rule they make it to be. And what’s really bad is their explanation for it:

“Less” is for hypothetical amounts, whereas “fewer” or “few” refers to a number that is quantifiable.

OK, now I know for sure that this was written by some underpaid drudge with no real knowledge of the subject who looked up some stuff on grammar advice on the web and didn’t really process it well. For the record, less refers to mass objects: things that can’t be counted – be they real or hypothetical. Fewer refers only to countables. (Few is not a comparative and should not even be in that sentence unless we also bring in the non-comparative words for mass objects: little, not much, and a few others.) “A number that is quantifiable” is a head-desk phrase. All numbers are quantifiable. Oh, ePly, would you hire a non-expert to write about any other thing? Perhaps they would. I don’t know. Well, what they have here is rubbish. Browbeating, grammar-peeving rubbish that has some facepalming mistakes in it.

Just to prove further that their author is desperately ignorant in matters of English words and grammar, they move on next to this:

Impactful simply isn’t a word – “The keynote speaker will give an impactful presentation…” is grammatically incorrect. Sorry to take that one away from you.

First: impactful is a word, and has been for a long time. You don’t have to like it, but if you think it’s not a word, you have no understanding of what is and isn’t a word. Take a linguistics course, for heaven’s sake. One intro course in linguistics and you would stop making all of these dreadful errors.

That includes the error of saying that using a non-word is grammatically incorrect. No, a sentence such as “The vulks spanged the gromple” is not grammatically incorrect, it just uses lexical items for nouns and verb that are not attested in the lexicon and have no agreed sense. On the other hand, “The clowns spanks the dog” is grammatically incorrect because we can see it has an improper conjugation, and “Spanks the dog clown the” is grammatically incorrect because the word order is all wrong.

Next they talk about affect versus effect. This is in fact an important distinction to make. The advice they give is also generally true:

Remember, “affect” is a verb while “effect” is a noun.

However, they should effect one little change there: add a word such as normally or usually after each is. There is a noun affect referring to emotion, but it is uncommon; there is a verb effect, which is a silk-shirt way to say “cause to happen.” Still, this is not as bad as most of their slip-ups.

Next is the it’s/its distinction, also one worth maintaining. The only mistake they make is to say this:

“It’s” refer to a verb, whereas “its” is a possessive.

First, refer is conjugated improperly; it should be refers. Second, refer is not the right word either. It’s doesn’t refer to a verb; it contains a verb – one of two possibles: is and has. They should just say “It’s stands for it is or it has, whereas its is a possessive like his.”

Next they talk about the difference between then and than. They actually pretty much don’t say anything wrong about this. For once.

They follow this, however, with another example of a stylistic recommendation presented as an absolute law:

“Farther” refers to measurable distance, “further” should be used for an abstract length.

This is a distinction you can make but don’t have to. You can talk of going farther in a relationship and further down a road if you want; it’s just more common to do the converse. On the other hand, you can go far but not fur, and you can further an aim but you can’t farther it.

Oh, yes, by the way, their sentence has a comma splice in it. These people who tell you to proofread your forms (in their tips below their grammar advice) haven’t managed to catch a comma splice. Now, many nice people make comma splices – but I would recommend against making one in a screed of grammatical prescriptivism. Or, if you do, hire a proofreader to catch it.

They inveigh next against “misconnecting verbs”:

Wrong: You should try and register before the price goes up. Right: You should try to register before the price goes up.

The “wrong” version is informal – I wouldn’t recommend it in formal businesslike prose. They don’t really explain what’s up (the first is two imperatives presented in a colloquial idiom; the second is an imperative with an infinitive complement – but that’s more technical than most people would understand), but I’ll give them a pass on this one.

They cap off with this injunction:

“Cannot” should always be written as one word. Not “can not.”

In many cases, it’s actually better to write can’t, but that does depend on the formality of the document: many web forms suffer from excessive formality: “Upon completion of the registration process” rather than “When you have finished registering,” for instance. With cannot, though, they’re making too hard a rule again. Consider the distinction between can not do it (“am able not to do it”) and cannot do it (“am not able to do it”).

It’s not that there’s no value in reminding people of some common usage errors and some things that may seem excessively informal. It’s just that it can be done without overstating cases, saying inaccurate things, and making errors of one’s own.

How does one do that? Get an actual expert to write about it. There are many available, often at shamefully low rates.

100% of these usages is wrong

I have just seen an infographic (heaven help us, yes, an infographic – generally now not actual charts but just text tarted up) with the following statements:

46% of all U.S. workers claims that they are less productive without coffee.

61% of the workers who need coffee to get through their day drinks 2 cups or more each day.

49% admits to needing coffee while on the job in the Northeast where the workday coffee ritual is the strongest.

Let’s ignore all the other issues in those sentences and just focus on the most egregious, unnatural usages: 49% of workers claims; 61% of the workers drinks; 49% admits. Ick. Just ick.

This is a classic overthink error. I see it mainly in newspapers and similar places where the writers are trying to enforce their understanding of “proper” grammar and are going against their normal speech instincts in doing so.

Percentages can apply to unitary or mass entities and they can apply to populations of entities. When you’re talking about mass or unitary entities, it’s right to use the singular: “50% of this cake is chocolate”; “50% of this collection is action figures.” Moreover, when you’re talking about average (or consistent) percentage of each individual in a set, you may use the singular, though it can sometimes be awkward to phrase it thus: “40% of her cupcakes is sugar.”

But when you’re talking about the portion of the individuals in a set of individual entities, percents are plural quantifiers. You don’t say “46% of the people here drinks coffee” unless somehow each employee has a body 46% of which (perhaps on average) drinks coffee and the other 54% of which abstains. Would you say “Half of the employees here drinks coffee”? How about “A lot of the people here drinks coffee”? Hey, a lot is singular, you know!

Which is just the point. A lot may be singular, and 46% may be a discrete quantity, but their effect on the nouns they describe is a plural quantification. Remember that a dozen is also a singular construction and a discrete quantity, and a hundred likewise, and yet you don’t invariably conjugate verbs in the singular after them: “A dozen people is coming over”? No. (But you can say “A dozen eggs sits in the basket” because you know that’s a carton.) You can say “A bunch of flowers sits by the window” because in that case a bunch is a unitary object; if you say “A bunch of people sit by the window” it means that the people may or may not be together as a unit, but there are a fair few of them in any event. (And “A bunch of people sits by the window” is an almost amusing image of a set of people so together in their grouping that they even sit as a single unit.)

It’s easy enough to see how people can get confused. Many of these things can take singular or plural depending on what are sometimes very fine nuances of meaning. I can say “100% of these usages is wrong” and mean that each usage is 100% wrong, and I can say “100% of these usages are wrong” and mean that every last one of them is wrong. But there are cases where your ear just screams: “46% of workers claims”?! No. Just no. A percentage of a population of individuals is a plural.

And really, if your analysis of grammar leads you to write something that sounds staggeringly wrong, stop and reconsider your analysis.

Famous quotes that break “rules”

I expected my latest article for TheWeek.com to generate some reaction in the comments, and I was not disappointed. Not that I wrote it just to troll people, but when you venture into certain territory…

The idea behind the article was to look at some famous quotes – sayings that are well known and often said – that break rules that are often learned in schools at about the same time as the quotes are. And then, of course, to look at whether those rules are really rules or not. But I didn’t explain that in an introduction. I just dove right in (or, if you’re a hoary prescriptivist, dived right in). Which may not have been the best idea, since – in combination with an eye-catching but slightly misleading headline (I don’t write the headlines, by the way, but I do get to see them in advance and could always suggest a change) – this approach provoked a variety of reactions in the comments section.

Here, for better or worse, is a link to the article:

9 famous quotes that are (technically) grammatically incorrect

And feel free to tell me what you think!

Incomplete sentences? Sure! Why not?

My latest article for TheWeek.com is up, and it’s on the oft-maligned “sentence fragments”:

It’s totally okay to write incomplete sentences

A few readers have pointed out, as I rather thought someone might, that Shakespeare isn’t really the best example. This is true, but I needed an example that I could be confident readers would be familiar with and would not dismiss as too modern, and I also had a length limit. So there it is. The compromises always get you in the end.

You can also see this article on Salon.com, and I don’t even know where else.