jutfrank
VIP Member
- Joined
- Mar 5, 2014
- Member Type
- English Teacher
- Native Language
- English
- Home Country
- England
- Current Location
- England
But what if we try to generalize from a single observation in the past? Would that be a valid generalization?
I'd have to say no. By definition, that wouldn't be a generalisation.
Yes, I agree with that but my explanation is to say that there is not one single observation here but many, because all occurrences of bank-robbing, both successful and unsuccessful, must be considered for the generalisation to be made. In other words, the single case only makes sense in relation to the other cases. There just needs to be a minimum of one unsuccessful attempt for the generalisation to be valid.In my last post, I tried to generalize from a single observation when I said, " an event that has happened only once, and there's a good chance that it will/may happen again", and then I went on to discuss a case under it, where I said,
"In the past, John has attempted to rob the bank five times. He was successful four times and got into trouble once: Robbing the bank can get John into trouble."
Here the speaker takes into account five different occurrences of bank-robbing by John, of which one was unsuccessful. Then based on this single unsuccessful occurrence, he makes the general observation: Robbing the bank can get John into trouble.
Now, lets discuss something I'm still confused about.
So far we've talked about generalizations. Now, suppose the speaker is completely unaware of John/one/people having robbed the bank/banks in the past. Thus, no past observations imply no induction, which implies no generalization.
Q): Now if there is no generalization, can we use "can" in the following sentences? If so, what does using "can" mean in those sentences? Do they all refer to the future?
1) Robbing the bank can get John/people into trouble.
2) Robbing a bank/banks can get John/people into trouble.
Well, no. If they aren't meant as generalisations, then what are they? I'm not sure I properly understand your question.
Perhaps what you mean should be expressed with other modal verbs. Look:
Robbing the bank could get John into trouble.
The speaker here is probably thinking about a specific future event. That's what could does, in contrast to can—it places the action in a specific point in space/time. Depending on context, we'd either interpret the future event as a 'real' event where we imagine John really doing it, or as a purely theoretical one, where the whole thing is considered as kind of potentiality that can't be ruled out.
If the future event is imagined clearly to be 'real' in the sense I've outlined above, it would be better to use might in this case instead. Very generally speaking, I'd say the best way to differentiate between could and might is by using this real/theoretical distinction of possibility.
Robbing the bank might get John into trouble.
This is clearly not a generalisation and not a theory. There's a real person, a real bank, and probably a real plan in John's mind.