You are currently browsing the category archive for the ‘prepositions’ category.

We’re all Rolling Stones fans here, right? I mean, we’re all here on a grammar blog, so I don’t think I’m jumping to too wild a conclusion to assume that we’re almost all oldsters, whether in actual age or personality. So let’s talk about the classic “Get Off of My Cloud”:

As it turns out, the Stones weren’t terribly fond of this song; they felt it was a rushed follow-up to the runaway success of “Satisfaction”. But some grammar peevers dislike it for an unrelated reason:

“‘Off of’ is no way to talk. It IS really, really bad English.”

Hatred of off of is widespread. It pops up commonly in peeve lists. Some professional grammar commentators share this complaint: the quote above is from Patricia O’Conner of Grammarphobia*, and Grammar Girl tersely dismisses it with “You jump off the pier, not off OF the pier”.

So what’s supposedly wrong with off of? The main problem seems to be that the of is unnecessary, but another common one is that since it’s on and not on of, it must be off and not off of. I also see complaints that it’s dialectal or informal or American, that one can’t put two prepositions next to each other, or that it ought to be from. And worse, given all of these problems, the phrase is supposedly spreading.

Let’s take these in reverse order. First, I’m unconvinced that it’s spreading, unless you’re talking about a very recent (last 20 years) spread. Here’re the Google Books counts, and you’ll note that modern off of usage is still below its peak in 1910. The Corpus of Historical American English has a slightly different picture, with more-or-less stable usage from 1900 to the 1980s, and then a jolt up in the 90s and 2000s. Maybe it’s spreading, maybe not. But let’s talk about why it’s not bad either way.

I’ll start with the easiest objections. No, it shouldn’t just be from. Consider:

(1a) The numbers station is broadcasting from a shed off of Route 395.
(1b) *The numbers station is broadcasting from a shed from Route 395.

And yes, you can put two prepositions next to each other, as in this unobjectionable example:

(2) I pulled a coat out of the closet.

Going on to a somewhat more complex objection, antonymic phrases do not have to share structures or prepositions. The fact that you get on and not on of a train doesn’t mean that you have to get off and not off of it. Consider:

(3a) I put the sandwiches into the picnic basket, but someone has pulled them out of it.
(3b) One velociraptor was in front of Muldoon, the other next to him.

And now on to the involved discussions. One question is whether off is always sufficient, and off of thus always unnecessarily wordy. And the answer, I think, depends on that of a second question: whether off of is dialectal.

In my idiolect, off of is perfectly standard. I was probably in my twenties before I heard someone object to it. That’s not to say I can’t use off without of. To the contrary, I prefer (4) without of, though both forms are acceptable to me:

(4) The leaves fell off the tree.

That said, of is not always superfluous to me. A few examples where I find removing of to make the sentence noticeably worse:

(5a) It’s a way of profiting off of something you expect to drop in value.
(5b) My new invention will knock the socks off of the scientific community.
(5c) I broke your statue by knocking the top off of it.

You may not agree, even if you come from an off of idiolect, that these forms are better, but that’s not important. The key point is only that sometimes, to some people, off of is distinctly more mellifluous than off. Dismissing off of out of hand as superfluous is valid only in dialects that already don’t allow it.

Let me elaborate this “necessity depends on dialect” point by proposing an insane argument. I’ve mentioned before that, being from Pittsburgh, I am perfectly content to say The car needs washed instead of The car needs to be washed. Within my dialect, to be is often superfluous, and there are some sentences that I find greatly improved by omitting it. Thus, I could see arguing that to be is, at least sometimes, unnecessary. But if I argued this to someone speaking a “standard” dialect of English, I would sound crazy. Saying that of in off of is across-the-board unnecessary sounds equally crazy to me.**

So is off of dialectal and/or informal? The answer would seem to be yes to both. The Oxford English Dictionary calls it “only colloq. (nonstandard) and regional” in current use. The Merriam-Webster Dictionary of English Usage says it’s “primarily a form used in speech”. The Columbia Guide to Standard American English says it’s avoided at “Planned and Oratorical levels and in Semiformal and Formal writing.”

Those sources are generally pretty trustworthy with their opinions, and given the amount of people who find off of unacceptable, I’m inclined to believe that it really is dialectal. When that’s coupled with its primarily spoken usage patterns, it’s no surprise that it would feel informal, especially to people from other dialects. And using the Corpus of Contemporary American English as a measuring stick of informality, off of occurs in speech twice as often as in written fiction, about four times as often as in newspapers/magazines, and almost ten times as often as in academic writing. The more formal the style, the less likely you’ll see off of.

All that said, its informality doesn’t mean it’s an illiteracy. Off of used to be standard in English; the MWDEU starts off with a Shakespearean usage [1592] and continues with Pepys [1668] and Bunyan [1678]. In the last century, they show it used by Hemingway, Faulkner, and Harry Truman, among others. So if it is making a comeback, it’s no harbinger of linguistic doom, just a return to form.

Summary: There is nothing linguistically or grammatically wrong with off of. It’s nonstandard in some dialects and informal in most, so you should probably avoid it if you’re concerned about your writing seeming formal. But when formality isn’t a concern, use it as you see fit.

*: This is a surprising stance, because it comes from Patricia O’Conner of Grammarphobia, who’s normally a lot less judgmental about such things. In fact, three years later, she softened her stance, although she remains against off of. I included her original opinion because her reconsideration shows that even hard-line opinions can (and should) be altered in the face of evidence, so long as the commentator is reasonable.

**: In fact, I and others within my dialect seem to have strong intuitions about times when the to be can and can’t be felicitiously dropped, in the same way as I see off of. It’s not a matter of necessity but of felicity.


If English words were Norse gods, perhaps than would be the best candidate to play the role of Loki, the trickster god.* It causes confusion not only due to its similarity with then, but also by raising the question of what case the noun phrase it governs should have. Of course, there’s no confusion if you are a brilliant grammaticaster, as in this example from a list of peeves:

12. I/Me: We had several different takes on this, with one correspondent nailing it thus: “The correct choice can be seen when you finish the truncated sentence: He’s bigger than I am. ‘He’s bigger than me am’ actually sounds ridiculous and obviates the mistake.”**

Now, there’re two questions one should be asking of this explanation. First, can you just fill in the blank? By which I mean, does a sentence with ellipsis (the omission of words that are normally syntactically necessary but understood by context) necessarily have the same structure as a non-elided sentence? There are many different types of ellipsis, so this is a more complex question than I want to get into right now, but the short answer is no, and here’s a question-and-answer example:

Who drank my secret stash of ginger-grapefruit soda?
(1a) He/*Him drank it.
(1b) Him/*He.

(The asterisks indicate ungrammatical forms.) The second question to ask here is whether there even is an elision. We know that “He eats faster than I do” is a valid sentence, but does than necessarily trigger a clause after it? Could it be that “He eats faster than me” is not an incomplete version of the above structure but rather a different and complete structure?

This boils down to the question of whether than is strictly a conjunction (in which cases the conjoined things should be equivalent, i.e., both clauses) or can function as a preposition as well (in which case the following element is just a noun phrase, with accusative case from the preposition). It’s pretty easy to see that than behaves prepositionally in some circumstances, pointed out in the Cambridge Grammar of the English Language, via Language Log:

(2a) He’s inviting more people than just us.
(2b) I saw no one other than Bob.

But this prepositional usage has become frowned upon, save for instances like (2a) & (2b) where it’s unavoidable. Why? Well, it’s an interesting tale involving the love of Latin and the discoverer of dephlogisticated air. It all starts (according to the MWDEU) with Robert Lowth, who in 1762 claimed, under the influence of Latin, that than is a conjunction and noun phrases following it carry an understood verb. The case marking (e.g., I vs. me) would then be assigned based on the noun phrase’s role in the implied clause. So Lowth’s grammar allows (3a) and (3b) but blocks (3c):

(3a) He laughed much louder than I (did).
(3b) He hit you harder than (he hit) me.

(3c) *He laughed much louder than me.

Lowth’s explanation was not without its wrinkles; he accepted than whom as standard (possibly due to Milton’s use of it in Paradise Lost), and worked out a post hoc explanation for why this prepositional usage was acceptable and others like (3c) weren’t.

Lowth’s explanation was also not without its dissenters. Joseph Priestley, a discoverer of oxygen, argued against it in his Rudiments of English Grammar (1772 edition). Priestley noted that the combination of a comparative adjective and than behaved prepositionally, and that good writers used it as such. Regarding the Lowthian argument against it, he wrote:

“It appears to me, that the chief objection our grammarians have to both these forms, is that they are not agreeable to the idiom of the Latin tongue, which is certainly an argument of little weight, as that language is fundamentally different from ours”

Another pro-prepositionist was William Ward, whose 1765 Essay on Grammar argues that than in phrases like to stand higher than or to stand lower than is akin to prepositions like above or below, and thus that the prepositional usage of than should be allowed. That’s not much of an argument, because semantically equivalent words and phrases don’t have to have the same syntax — consider I gave him it vs. *I donated him it. But then again, Lowth is basing his argument on a completely separate language, so this is a slight improvement.

The real key, of course, is usage, and the MWDEU notes that Priestey and Ward are backed by standard usage at the time. Visser’s Historical English Syntax strengths the case with examples of prepositional than from a variety of estimable sources, including the Geneva Bible (1560), Shakespeare (1601), Samuel Johnson (1751, 1759), and Byron (1804). And of course prepositional than continues in modern usage — why else would we be having this argument?

Despite the irrelevance of his argument, Lowth’s opinion has stuck through to the present day, reinvigorated by new voices repeating the same old line, unwilling to concede something’s right just because it’s never been wrong. It’s the MWDEU, not the commenter mentioned at the top, who nails it:

“Ward’s explanation [that both conjunctive and prepositional uses are correct] covered actual usage perfectly, but it was probably too common-sensical—not sufficiently absolutist—to prevail.”

So in the end, we’re left with this. Than I and than me are both correct, in most cases. Than I is often regarded as more formal, but interestingly it’s the only one that can be clearly inappropriate.*** Using the nominative case blocks the noun phrase from being the object of the verb. I can’t write The wind chilled him more than I to mean that the wind chilled me less than it chilled him.

Sometimes this can be used to disambiguate; I love you more than him is ambiguous while I love you more than he isn’t. This only matters for clauses with objects, and in general, these tend not to be ambiguous given context, so the benefit of disambiguation must be weighed against the potentially over-formal tone of the nominative case. (I, for instance, think the second sentence above sounds less convincing, even though it’s clearer, because who talks about love so stiltedly?)

I’m branching a little off topic here, but I’d like to conclude with a general point from Priestley, a few paragraphs after the quote above:

“In several cases, as in those above-mentioned, the principles of our language are vague, and unsettled. The custom of speaking draws one way, and an attention to arbitrary and artificial rules another. Which will prevail at last, it is impossible to say. It is not the authority of any one person, or of a few, be they ever so eminent, that can establish one form of speech in preference to another. Nothing but the general practice of good writers, and good speakers can do it.”

Summary: Than can work as a conjunction or a preposition, meaning that than I/he/she/they and than me/him/her/them are both correct in most situations. The latter version is attested from the 16th century to the present day, by good writers in formal and informal settings. The belief that it is unacceptable appears to be a holdover from Latin-based grammars of English.

*: Interestingly, and counter to the mythology I learned from the Jim Carrey movie The Mask, Wikipedia suggests that Loki may not be so easily summarized as the god of mischief.

**: I’m pretty sure the use of obviate is a mistake here. I can just barely get a reading where it isn’t, if thinking of the full version of the sentence causes you to avoid the supposed mistake. But I suspect instead that our correspondent believes obviate means “make obvious”, in which case, ha ha, Muprhy’s Law

***: Is there a case where than me is undeniably incorrect? I can’t think of one.

Hiding from my dissertation in a little alcove under the stairs on the bottom floor of the library, I was scanning through a book of grammar gripes. One of them was the common objection to transitive usage of the verb graduate. For instance, people will sometimes say:

(1) Now that they’ve graduated high school they can set their goals on college.

Those of an older bent will be more familiar with an intransitive usage where the graduated institution appears in an ablative* prepositional phrase:

(2) Yesterday the heir to the Notorious B.I.G. throne, young Tyanna graduated from high school at an undisclosed location.

And, you may be thinking, darn right! It’s graduated from, and it’s always been, and the kids are screwing up the language again. And it’s true that the transitive form in (1) is newer and seems to be gaining in popularity.** But it turns out that graduate from isn’t the original form, either. It used to be graduated at, as in this 1871 example:

(3) He graduated at Williams College in 1810, and studied theology with the Rev. Samuel Austin, DD, of Worcester, Mass.

So already, just going back 140 years, we’ve seen transitions from graduated at to graduated from to the plain graduated. But there’s an even more substantial change in the history of graduate. Graduating used to be something a school did to its students, not something the students did to the school. One was graduated at some school — witness this 1827 list of folks that Harvard graduated, such as:

(4) Jabez Chickering, Esq., son of Rev. Jabez Chickering, was graduated at Harvard University, in 1804; and settled in the profession of law in this town.

I’ve put together a Google Books N-grams graph illustrating the changes over time:

[The history of graduate]

Interestingly, it looks like the forms in (3) and (4) were both in use throughout 19th century American English. That’s a bit surprising because the two forms assign different roles to their subjects, but it just goes to show that grammatical ambiguity is tolerable when there’s no chance of confusing the roles. (It’s always clear that the person is getting the degree, and the university issuing it.) We see was graduated at start dropping off in the second half of the 19th century, graduated at remaining strong until the early 20th century, and graduated from taking off from there.

So while I graduated high school may not yet be standard, it will be, and there’s nothing wrong with it. It just isn’t what people used to say. For whatever reason, the younger generation likes to change how graduation works. There’s no reason to fret over it; it’ll change, and life will go on, and our kids will be just as grumpy as us when their kids re-reinvent the word’s usage.

*: Ablative is one of a set of words describing the cases that can be marked in a language. Ablative in particular indicates motion away from something; Wikipedia has a list of these, including such fun ones as illative and inessive. (Valid only for certain definitions of “fun”.)

**: I’m a little surprised, but I don’t see any clear evidence in Google Books N-grams or the Corpus of Historical English of the transitive usage growing faster than the ablative intransitive. I suspect this is due to a strong avoidance of the transitive usage in writing, which both of these corpora are based on.

Let’s continue the S-Series by talking about beside and besides. I’ve heard a lot of people kick up a fuss over these two, but having thought through their usage, I’m rather surprised. I don’t think a lot of native English speakers really confuse the two forms anymore. The two used to be pretty interchangeable, like toward and towards, but beside generally ceded its non-literal meanings to besides sometime in the 19th century.

Unfortunately, it’s always difficult to get good statistics on the prevalence of different meanings of a word, so I’m basing what I say here what the Oxford English Dictionary, the Merriam-Webster Dictionary of English Usage, and (to the least extent possible) my brain have to say on the matter. I’ve added corpus statistics where possible.

Next to: beside. The most literal meaning is also the primary meaning of beside in modern English. If you’re talking about physical positioning, you definitely want beside; the last attestation of physical-position besides in the OED is from 1440.

(1a) The purple couch beside the road
(1b) I am slightly concerned about the hungry tiger standing beside me.

Idiomatic nearness: beside the point. The “next to” meaning of beside is not limited to physical proximity; there are also idiomatic usages, the most prominent of which is beside the point. The alternative, besides the point, is rarely attested both historically and recently:

Adverb: besides. The OED reports that beside was once a standard adverb, but now it’s become obsolete in most of its adverbial usages and archaic in the rest. So use besides in sentences like:

(2a) Men? Sure, I’ve known lots of them. But I never found one I liked well enough to marry. Besides, I’ve always been busy with my work.
(2b) … lost her social position, job, and husband, and was broke besides. [MWDEU]

If you’re using it as sentence modifier, as in (2a), or as a clear adverb (i.e., without a noun following it (2b)), you probably want besides.

In addition to: besides. Now let’s return to prepositional usages. In modern English, the “in addition to” meaning almost always uses besides:

(3) There was no need to install additional software besides the game itself.

Beside used to be common in this usage, but it seems to have become rare in modern English (although I feel like it may be a common local variant in some places). I’ve found it only rarely in modern American writing, such as “Did Glenn mention anything beside the names I dropped?”, from COCA in 2002.

Other than: besides. A similar usage to the last one, again with besides as the primary modern form. Here I’m talking about using the word to mean something like “except”, as in:

(4a) Having been lost in the forest for days, I began to forget whether I’d ever eaten anything besides acorns.
(4b) Will Los Angeles ever be something besides a “suburban metropolis”?

Summary & caveat

So, in general, beside is used for literal and figurative nearness, and besides takes pretty much everything else (especially all other metaphorical usages).

That said, there is an important point here: each of the suggestions I’ve made above is still a bit fluid, since the distinct usages often didn’t start ossifying until the 19th century. Different usages and different people will vary on how much one form is preferred over the other. Adding an s in (1b) is straight out for me, but dropping the s in (3b) just sounds dialectal. And the further you go back in English, the more the usages will blend together.

Lastly, the promised caveat: it’s very hard to get clear data on the usage patterns of different senses of a word, so while I’m confident that these rules of thumb are accurate for my own idiolect, and fairly confident that they apply to standard American English, I’m not sure how well they reflect non-American Englishes. Use them at your peril.

The S-Series so far:
S-Series I: Anyway(s) [02/03/11]
S-Series II: Backward(s) [06/14/11]
S-Series III: Toward(s) [08/29/11]
S-Series IV: Beside(s) [12/07/11]

Post Categories

The Monthly Archives

About The Blog

A lot of people make claims about what "good English" is. Much of what they say is flim-flam, and this blog aims to set the record straight. Its goal is to explain the motivations behind the real grammar of English and to debunk ill-founded claims about what is grammatical and what isn't. Somehow, this was enough to garner a favorable mention in the Wall Street Journal.

About Me

I'm Gabe Doyle, currently an assistant professor at San Diego State University, in the Department of Linguistics and Asian/Middle Eastern Languages, and a member of the Digital Humanities. Prior to that, I was a postdoctoral scholar in the Language and Cognition Lab at Stanford University. And before that, I got a doctorate in linguistics from UC San Diego and a bachelor's in math from Princeton.

My research and teaching connects language, the mind, and society (in fact, I teach a 500-level class with that title!). I use probabilistic models to understand how people learn, represent, and comprehend language. These models have helped us understand the ways that parents tailor their speech to their child's needs, why sports fans say more or less informative things while watching a game, and why people who disagree politically fight over the meaning of "we".

@MGrammar on twitter

If you like email and you like grammar, feel free to subscribe to Motivated Grammar by email. Enter your address below.

Join 980 other subscribers

Top Rated

%d bloggers like this: