Just read this exception message in some code:
"You cannot log responses in a session that has completed."
which is slightly ungrammatical to me (I would need to use the passive "has been completed").
It's interesting the difference between "complete", "finish" and "end" in the context of, say a course.
A student can complete a course but a course cannot complete. A student can finish a course and a course can finish. A course can end but it is somewhat borderline to say a student ended a course (although I might be able to say "the student ended the course on a high note", certainly not "the student ended the course").
I've ordered from room service a couple of times here at the Atlanta Marriott Marquis for SBL and every time they end with something like:
that will be approximately less than 30 minutes
A friend once called me asking whether "number of casualties" is singular or plural. Apparently the words had been uttered by a reporter and there had been some debate at the friend's house as to whether the reporter had got the agreement between the verb and subject correct.
My intuition was that the following are both grammatical:
and that a with was and the with were would both be ungrammatical.
Thinking about it some more, I came up with another example pair.
That made things a little clearer to me (although looking back at the original pair, it's now obvious there too). With the indefinite article a, the heads of the phrases are casualties and people (both plural) whereas with the definite article the, the heads of the phrases are both number (singular).
In "a number of casualties were reported", it's the casualties that were reported; in "the number of casualties was reported", it's the number that was reported.
Even more clearly, in "a number of people were high", it's the people that were high; in "the number of people was high", it's the number that was high.
That would all suggest structures along the lines of:
The structure explains the subject-verb agreement and the semantics.
The interesting question in my mind remains: what it is about the article that determines which structure is licensed in each case?
This probably contributes to a lot of people thinking it's Rubix rather than Rubik's.
The phrase "The Rubik's Cube" sounds odd because you can't normally use an article with a pre-nominal genitive if the pre-nominal itself wouldn't normally take an article.
You can say "the paper", "the professor" and "the professor's paper". You can say "David's paper" but not "*the David's paper". (Although note that if talking about the sculpture "the David", you can say things like "the David's left hand". And, because of Donald Trump, you could say "the Donald's hair".)
You can't say "the Rubik" and so "the Rubik's cube" seems ungrammatical if you think about its component parts.
What's happening is, of course, that "Rubik's" isn't acting as a genitive anymore but rather "Rubik's Cube" has been reanalyzed as an opaque compound noun. It's just still written in terms of its components.
But I wanted to respond to zdsmith for starters.
This'll be a stream to capture various linguistic observations.
Start with the sentence "the customer walks away".
It's modified by a participle phrase to become "the customer walks away actually expecting the technicians to do X"
This then becomes a relative clause "what the customer walks away actually expecting the technicians to do" which in turn becomes the object of "enumerating".
In this card zdsmith writes:
I just stumbled into a real linguistic oddity. Take the sentence: 'The customer walked away expecting the technician to do [X]'.
I just wrote '... enumerating what the customer walks away actually expecting the technicians to do.'
What the customer walks away? What is going on here that I can make an indirect question out of that present participle but end up making it the object of 'walks'?
But I don't really see what is odd about it.