Word Fails Me #1: This document is smoking

This is an entry in a series of posts I’m calling Word Fails Me, in which I highlight the strange ideas that Microsoft Word has about English grammar. Each post will be a screenshot with little or no comment. The intention of this series is to amuse you and make you wonder where Word is getting its ideas. I’m not trying to be condescending to Word’s grammar checker or the people behind it. Word is a fascinating program and the grammar checker can be a lifesaver, even if it leans prescriptivist sometimes. If I come across interesting research into MS Word’s grammar checker, I’ll share it here. You can find all of the entries under the Word Fails Me tag. Enjoy!

Welcome to Word Fails Me! Let’s start with an example from Microsoft’s own web site about their grammar checker. As you can see, Word thinks the following sentences are A-OK:

  1. The document is sorting that we must review.
  2. The document is smoking that we must review.


To be fair, I think you should definitely review a document if it is smoking.


I, me and Oxford Dictionaries

I’m sure I’ve tweeted about this already, but the Oxford Dictionaries’ advice on the usage of pronouns just came across my interwebs again (they sent out this quiz in their email newsletter). It’s hard to imagine how a dictionary’s website gets this so wrong, but let’s go through it to see what’s up.

In their advice article “‘I’ or ‘me’?”, Oxford Dictionaries claims that in coordinated constructions where a pronoun and a proper name form the subject of a sentence, the pronoun used must be the subjective form of the pronoun (also called the nominative form). What this means is that in a sentence like “John and I went to the GWAR concert”, it is incorrect to use me instead of I. Let’s leave aside the fact that everyone everywhere naturally uses me in sentences like this. Let’s instead think about the advice that Oxford Dictionaries is giving. We’ll use the sentence that they use: Clare and I are going for a coffee. According to Oxford, it’s not just the subjective pronoun I that must be used in this sentence, only subjective pronouns must be used when the pronoun helps form the subject of a sentence. But how does this work? See if any of the sentences below sound odd to you.

  1. Clare and I are going for a coffee
  2. Clare and me are going for a coffee
  3. Clare and you are going for a coffee
  4. Clare and you are going for a coffee
  5. Clare and she are going for a coffee OR Clare and he are going for a coffee
  6. Clare and her are going for a coffee OR Clare and him are going for a coffee
  7. Clare and they are going for a coffee
  8. Clare and them are going for a coffee

If you’re like me, the first four sound fine (obviously, there’s no difference between the subjective and objective form of the 2nd person pronoun, they’re both you). The fifth one, however, sounds a bit stuffy compared to the sixth one (stuffy is a totally legit linguistics term). And the seventh is bordering on unacceptable. Does Oxford really think that Clare and they are going for a coffee is correct, while Clare and them are going for a coffee is not? Maybe? They didn’t use that sentence as an example. They focused instead on the 1st person pronoun – where there is more variation.

This topic boils down to a few things. First, English tends to favor me as the default pronoun in all cases except for when the pronoun stands alone as the subject. There is such a strong tendency to use me in all cases that this form is sometimes referred to as the oblique form, meaning that in addition to being the object, it fulfills other roles in sentences. And so English quite naturally uses the me form in coordinated structures, or phrases where there’s a pronoun and something else joined together with the word and:

John and me went to the GWAR concert.

Me and the bouncer got into an arm wrestling match.

Me and this other guy partied with GWAR after the show.

Second, using the subjective pronoun I in coordinated constructions isn’t wrong. English allows for both constructions and the choice of which one to use usually breaks along formality of the occasion – John and I seems more formal, while John and me seems more informal. But there is evidence of both structures throughout history in many different styles of writing. The John and I form is dictated by prescriptivist grammarians (and apparently some dictionaries), while the John and me form is proscribed, despite being used by everyone. In constructions with the first person singular pronoun, you can’t go grammatically wrong choosing I or me. But notice, however, that me is more versatile in where it can be placed:

Clare and me are going for a coffee

Me and Clare are going for a coffee

Clare and I are going for a coffee

*I and Clare are going for a coffee

As we have seen, in constructions with the 3rd person pronouns, things are potentially more cut and dry. With the 3rd person singular, it seems we should use the subjective forms (him, her) for all but the most formal registers. With the 3rd person plural, however, it seems we should always use the objective form them.

Finally, there is a piece of advice out there that I’ve seen in a lot of places. It goes like this:

In coordinated constructions (noun + pronoun), take out the noun and leave the pronoun. This will show you which case you want.

This advice is dumb. Why would I take something out of a sentence to decide how I should say the rest of the sentence after I put that thing back in the sentence?! This makes no sense at all. This advice is only given with coordinated subjects because it makes it seem like the subjective pronoun is always correct. Here’s Oxford using it at the end of their article:

An easy way of making sure you’ve chosen the right pronoun is to see whether the sentence reads properly if you remove the additional pronoun:

I am going for a coffee. ✗ Me am going for a coffee.

And here’s the Purdue Online Writing Lab:

In compound structures, where there are two pronouns or a noun and a pronoun, drop the other noun for a moment. Then you can see which case you want.

Not: Bob and me travel a good deal.
(Would you say, “me travel”?)

But what happens when I take the pronoun out of the sentence? I’m left with Bob travel a good deal. 😐

Y U NO give better advice, grammer peeple?

Ok, I’m being awful hard on Oxford Dictionaries. The thing is, their advice column could have been cleared up with a line that explained they were talking about Standard English only. Or that outside of standard written and spoken English, people are more likely to come across the form X and me. The X and me construction is so common in informal written and spoken English that using X and I may be out of place. Non-standard and informal English are the default forms of the language, whether they are written or spoken, so users of English will hear/read these forms most often in day to day circumstances. The split in choosing I or me along formal/informal or standard/non-standard lines isn’t a lot of linguistic knowledge for people to understand. They shouldn’t be forced into thinking there is only One True Way to use pronouns in English.

I might post more on this later and include the advice given by other style guides, grammars and dictionaries. If you want to see some of them backing up my claims right now, check out:

  • Merriam Webster’s Dictionary of English Usage, page 778
  • Fowler’s Modern English Usage 4th edition (edited by Butterfield), page 509
  • A Student’s Introduction to English Grammar by Huddleston and Pullum, page 107

A few less countable nouns

While everyone was worrying about whether less or fewer was correct in 10 items or less, another construction has been flying under the radar: a few less. I haven’t seen any style guides make remarks about this phrase, but it is an interesting one. It’s hard to search for online because there’s an Australian movie called A Few Less Men, which dominates the search results. I was able to find a WordReference forum about a few less, but it’s not much help. So let’s go to some corpora to see how a few less is used.

There are 36 hits for a few less in the Corpus of Contemporary American English (COCA), which means it’s not very common (for comparison, there are 4,875 hits for a few more). All of the hits for a few less pre-modify countable nouns.

Year:Genre Concordances – link to search: https://corpus.byu.edu/coca/?c=coca&q=63419241
West Twenties, one step up from a housing project, which meant a few less elevators chronically out of commission
But if we all drove just a few less times in the entire year, that is progress in an automobile-dependent metropolis like Atlanta
Fox: The Five
They may make a few less dollars, and they should do it.
And it could be that those other services continue on – maybe with a few less people, or maybe some people will cross over.
Move family outerwear out and add a few less flimsy hangers inside.
And how does one cure a sequence consisting of ” a few less atoms every day’?
If (Nu) had a few less zeros, only a short-lived miniature universe could exist. No creatures could grow larger
five, over a ten-year period, maybe a few more, maybe a few less, I don’t know, several times.

If you redo the search, it looks like there are 40 hits but the following do not fit the construction:

  • “Some health plans don’t cover Zyban, but a few less than forthcoming smokers have gotten around that by asking doctors to diagnose them with depression”. It’s more a few less-than-forthcoming smokers.
  • “Only a few less accessible villages have so far been spared of tourists”. This is also a case where less is modifying the following adjective and could be rewritten as a few less-accessible villages.
  • “there are always a few less visible non-tariff barriers which arise which will need to be smoothed out.” This again is a few less-visible non-tariff barriers.

There is also the concordance “Twenty years since our first date. A few less than that since I helped her pick out her first grown-up road bike”. In this construction, I would say that less is a noun and few is an adjective.

In the corpus of Global Web-based English (GloWbE), the US, UK and Australia seem to use this construction most often, although the frequency per million words (the PER MIL column) is not that different between the US, Canada, UK, Ireland, Australia, New Zealand, and the Philippines (see the image below). The concordances also appear to show that a few less is a modifier for a countable noun, although I did not go through all of the 328 hits in GloWbE. You can re-do my search on GloWbE by following this link.

GloWbE - a few less

The way I see it, there are two ways to analyze this construction. First, in a few less NOUNs, the words a few make up a non-exact indefinite quantifying determiner and less is an adjective modifying the noun phrase. What you have is this:

A few less NOUNs = a few (indefinite determiner), less (adjective / head of AdjP), NOUNs

Second, I suppose it’s possible to treat few as an adjective too (modifying the adjective less) and leave a to be the single-word determiner. So you would have something like this:

A few less NOUN = a (determiner), few (adjective / modifier), less (adjective / head of AdjP)

But I wouldn’t go for this analysis because the Longman Student Grammar also treats a few as a quantifying determiner which denotes a small amount (p. 75).

The interesting thing about a few less is that it easily – and quite unremarkably – modifies count nouns. People have a problem with ten less items/dollars/miles/people, but no one seems to raise a fuss about a few less items. Of course, there’s nothing wrong with using less with countable nouns, especially ones that are units of measurement and money. But I don’t think people have considered that if less really can’t modify count nouns – and that fewer needs to be used with count nouns – then the construction we would forced to use is a few fewer items. And no one wants that.


Longman Student Grammar of Written and Spoken English (2002) by Douglas Biber, Susan Conrad and Geoffrey Leech.

English Grammar: A University Course (2nd edition, 2006) by Angela Downing and Philip Locke. pp. 428, 433, 481, 492