[Grammar] can vs could in the past

[Marta56]

Can you help me please with this issue:

The example is:

I'm sure they weren't happy.

What is the difference between:
1. They couldn't have been happy.
2. They can't have been happy.

Which one should I use?

Thanks very much for any help.

 

[GoesStation]

Hi Marta, and welcome to the forum. :hi:

We don't usually use the sentence 2 version in American English. It looks British to me, but you should wait for a BrE-speaker to confirm that and describe any differences they see.

So for AmE-speakers, the difference is that number 1 is natural and number 2 isn't. We understand number 2, though, and it means the same thing to us.

 

[jutfrank]

We don't usually use the sentence 2 version in American English. It looks British to me, but you should wait for a BrE-speaker to confirm that and describe any differences they see.

So for AmE-speakers, the difference is that number 1 is natural and number 2 isn't. We understand number 2, though, and it means the same thing to us.

Really?! :shock:

2 is preferable logically as it makes a present speculation about a past situation. This is shown by the use of the present tense in I'm sure.

1 is different, logically speaking, as the modal couldn't relates to the past.

They can't have been happy. = It is not possible that they were happy.

I'm still going to go ahead and advise using sentence 2, despite being quite disturbed by post #2.

 

[GoesStation]

Really?! :shock:

....

I'm still going to go ahead and advise using sentence 2, despite being quite disturbed by post #2.

I can't argue with the logic. We just don't often use the construction can't have been, especially in conversation.

I tried to construct ngrams to support this contention, but Google's ngram viewer only finds one result for "can't have been" in its entire English corpus. This surprises me.

 

[emsr2d2]

I can't argue with the logic. We just don't often use the construction can't have been, especially in conversation.

I tried to construct ngrams to support this contention, but Google's ngram viewer only finds one result for "can't have been" in its entire English corpus. This surprises me.

I'm more than surprised. I'm staggered! I tried both "can't have been" and "couldn't have been" (both with and without quotation marks) and got no hits for either. That seems almost impossible!

 

[GoesStation]

I'm more than surprised. I'm staggered! I tried both "can't have been" and "couldn't have been" (both with and without quotation marks) and got no hits for either. That seems almost impossible!

I suspect we've tripped over a bug.

 

[jutfrank]

I'd heavily edited my post #3 by the time I saw the following few posts. Sorry about that.

I've just spent the last 40 minutes seriously scratching my head over this one. I really want to say that sentence 1 is wrong on an analytical level—I just can't make sense of it. However, in the light of GoesStation's remark in post #2, I'm wondering if it's better to advise using it. At least it doesn't sound disastrous to a British ear.

What do you think, Piscean? The different modals obviously show a difference in modality, right?

By the way, ngrams don't like contractions, so you have to separate them into two parts, like this, I think:

don't = do n't
can't = can 't

 

[GoesStation]

By the way, ngrams don't like contractions, so you have to separate them into two parts, like this, I think:

don't = do n't
can't = can 't

That works in the COCA database but it sends the Google result all the way up to zero. :-(

 

[jutfrank]

I can't argue with the logic. We just don't often use the construction can't have been, especially in conversation.

Could you say more? What do you mean by "We don't often"? Do you mean some people do and some don't? Or do you mean it depends on the situation? Or on the register?

Sorry, but I need to know this.

 

[jutfrank]

That works in the COCA database but it sends the Google result all the way up to zero. :-(

Oh, I was wrong. Apparently you need to write the words separately to search Google Books. So can not have been, I think:

I'm not certain that's right, however. Anyway, it's important to remember that this is not a straight comparison of use because the two forms (can) have different meanings. The 'speculation' use (with can) is less common than the 'possibility' use (with could). In fact, there's not much point at all in comparing these two forms unless one goes through every instance of use, in each case checking the form against the use.

 

[jutfrank]

1. They can't have been happy.
2. They couldn't have been happy.



In #1 we have a non-past form of the modal CAN/COULD. The basic meaning, depending on context is is It is not possible to/They are not able to. The perfect form following shows us that what is not possible/they were unable to do happened in past time.

In #2 we have a form of CAN/COULD that is remote in time, directness or probability is It is remotely possible/They were not able to/They would not have been able to/It was not possible to. Once again, the perfect form following shows that what is remotely possible/they were unable to do/it was not possible happened in past time.

I just wanted to check if you understood it in the same way. Thanks.

Where I was struggling here is explaining the distinction between:

a) They could not be happy.
b) They could not have been happy.

There's a significant difference between them, which reveals a difference in modality. They both concern past time, in contrast to the following pair, where sentence c) concerns present time:

c) They could not be happier.
d) They could not have been happier.

 

[GoesStation]

Could you say more? What do you mean by "We don't often"? Do you mean some people do and some don't? Or do you mean it depends on the situation? Or on the register?

Sorry, but I need to know this.

I was hoping Google Ngrams would help. The COCA database shows can't have been to be about one-twentieth as common in American writing as couldn't have been, which is about what I would have guessed. As to why, I don't have a clue. I can only guess that if the two forms ever existed in American usage, their meanings blurred together and couldn't have been emerged as the winning form.

I don't have a sense that can't have been is used more in higher registers but I wouldn't be surprised if it were. Americans who read a lot are probably more exposed to British usage than the general public, so they'd be more likely to internalize it and use it.

Perhaps it was a similar process to the merger of may and might that I've been observing for twenty or thirty years. I can't fathom how people have lost track of the difference which is blindingly obvious to me, but most have.

I'm feeling guilty for having raised the point. But it may be a less stressful thing to worry about than some pressing current events.

 

[jutfrank]

I'm feeling guilty for having raised the point.

Quite right. I'll have you know I shall not be sleeping tonight.