Expressing probability in English with modals and adverbs

In a nutshell

  • Total certainty: will, must (deduction), can’t (impossibility)
  • High probability: should, ought to, be likely to
  • Moderate probability: may, might, could
  • Key adverbs: certainly, probably, likely, perhaps, possibly
  • Past: each modal combines with have + past participle for a deduction about the past (must have, might have, can’t have)
  • Trap: must not does not express impossibility but prohibition; it is can’t that says something is impossible or highly unlikely

English has a precise range of tools to express probability: modal verbs that position the speaker’s degree of certainty, adverbs that refine that degree, and lexical constructions that make things even more explicit. The nuance between must (near-logical certainty), should (high probability), may (balanced probability) and might (low probability) is absent from many languages that group these shades behind broader expressions. These modals belong to the family of modal verbs in English, to which can, will or would also belong.

The probability scale: an overview

Before going into detail about each modal, it is useful to picture the whole spectrum, from the most certain to the most unlikely.

Degree Modals / expressions Approximate translation
Certainty (100%) will, must (deduction) it must be, it is certainly
Impossibility (0%) can’t, couldn’t it is impossible, it can’t be
High probability should, ought to, be likely to it should, it is likely that
Moderate probability may it may be that, perhaps
Low probability might, could it might be that, it is not impossible

Will: certainty

Will, in its epistemic function (deduction or prediction), expresses the speaker’s certainty about a present, habitual or future fact. This is not the narrative future but a strong assertion about what is necessarily true or what will inevitably happen.

English Meaning
That will be the postman at the door. It must be the postman at the door. (I am certain)
She’s been driving for twelve hours. She will be exhausted. She has been driving for twelve hours. She must be exhausted.
You will know the answer to this one. You surely know the answer to this one.

Must: near-certain deduction

Must, in its probability usage, expresses a strong logical deduction. The speaker does not know with absolute certainty, but the available evidence convinces them of the statement. It is the modal of deduction par excellence, the one used when everything points to a conclusion and you state it with conviction.

English Meaning
She must be tired. She’s been working since 5 a.m. She must be tired. She has been working since 5 a.m.
You must be hungry after that long walk. You must be hungry after that long walk.
That must be the new director everyone’s been talking about. That must be the new director everyone has been talking about.

Must not does not express impossibility but prohibition. This confusion is one of the most frequent when handling probability in English. Must not (mustn’t) says what must not be done: You mustn’t smoke here = smoking is forbidden here. To say something is impossible or very unlikely, can’t or couldn’t are used: That can’t be right = it is not possible for that to be correct. The page on prohibition in English develops the distinctions between mustn’t, can’t and may not in their prohibitive role.

Can’t and couldn’t: impossibility and strong unlikelihood

Can’t (and its past form couldn’t) expresses negative certainty: the speaker is convinced something is impossible or highly unlikely. It is the symmetrical counterpart of must on the negative side.

English Meaning
She can’t be at home. Her car isn’t there. She cannot be at home. Her car is not there.
That can’t be right. I checked the figures twice. It is not possible for that to be right. I checked the figures twice.
You can’t be serious! It is not possible that you are serious!
He couldn’t have stolen it. He was with me all evening. He cannot have stolen it. He was with me all evening.

Should and ought to: high logical probability

Should, in its probability usage, does not carry judgment or advice: it expresses a logical expectation or a reasoned prognosis, what one expects to be true given the situation. Ought to, less frequent, shares this meaning but belongs to a slightly more formal or literary register.

English Meaning
The train should arrive at 9:15. The train is expected to arrive at 9:15. (according to the timetable)
She should be at the office by now. She is expected to be at the office by now.
That ought to be enough for four people. That should be enough for four people.
The results should be out tomorrow. The results are expected to come out tomorrow.

Should in this sense expresses a prediction based on logic or on a known fact (a timetable, a procedure, a quantity), as opposed to must which relies on a deduction from clues observed in the immediate situation.

May: balanced probability

May expresses a real possibility without tipping the balance one way or the other. The speaker considers the event plausible without pronouncing on its likelihood. It is the most neutral modal of probability.

English Meaning
It may rain this afternoon. It is possible that it will rain this afternoon.
She may be late. The traffic is bad. She might be late. The traffic is heavy.
The project may be delayed. The project could be delayed.
This may not be the best solution. This might not be the best solution.

Might and could: low probability

Might and could both signal a lower probability than may, a possibility considered without much conviction. Might is the past form of may but its present use does not carry a temporal value: it simply marks a higher degree of uncertainty. The nuances between may and might in English are worth knowing in detail, as the two modals share close uses while differing on several points.

English Meaning
He might come to the party, but I’m not sure. He could come to the party, but I’m not sure.
It might be worth checking again. It could be worth checking again.
She could be at the library. She might be at the library.
There could be some truth to what he’s saying. There might be some truth in what he is saying.

May vs might in the negative: may not and might not are not symmetrical to can’t. May not means it is possible that… not (possibility that it is negative), whereas can’t means it is impossible. He may not be at home (he might not be at home) is very different from He can’t be at home (it is impossible that he is at home).

Be likely to and be unlikely to: lexical forms

Be likely to and be unlikely to express probability in a more explicit and precise way than modals, without any possible ambiguity between the sense of probability and another sense (advice for should, permission for may). They are particularly common in a neutral or journalistic register.

English Meaning
She is likely to win the next election. She has a good chance of winning the next election.
This solution is unlikely to work. This solution is not likely to work.
It is likely that prices will rise further. It is likely that prices will rise further.
He is most likely stuck in traffic. He is most likely stuck in traffic.

Adverbs of probability: refining the degree

Adverbs of probability can be used alone or in combination with a modal to reinforce or nuance the degree of uncertainty expressed. Their positioning differs from modals: they are generally placed at the start of the sentence or between the subject and the main verb, never directly before a modal.

Adverb Meaning Degree Example
certainly certainly, without a doubt Very high She will certainly agree.
definitely definitely, assuredly Very high He’s definitely at home.
undoubtedly undoubtedly Very high This is undoubtedly the best option.
probably probably High She’ll probably call later.
likely likely High He’ll most likely be there.
perhaps perhaps Medium Perhaps she forgot.
maybe maybe (more informal) Medium Maybe they changed their minds.
possibly possibly, it is possible that Low to medium He could possibly be right.
conceivably it is conceivable that Low It could conceivably take years.

Perhaps and maybe are interchangeable in the vast majority of cases, with a nuance of register: perhaps is more formal and more written, maybe more oral and informal. Both are preferably placed at the start of the sentence when they apply to the whole sentence, or between the subject and the verb in mid-sentence position.

Probability about the past: modal + have + past participle

To express a deduction or probability about a past event, each modal combines with have + past participle. The degree logic remains identical, only the time reference changes.

English Meaning Degree
She must have missed the train. She must have missed the train. (near-certainty) Very strong
He can’t have said that. He cannot have said that. (impossibility) Impossibility
They should have arrived by now. They should have arrived by now. (unfulfilled logical expectation) Strong
She may have forgotten. She may have forgotten. (balanced possibility) Medium
He might have left already. He might have left already. (low probability) Low
They could have taken the wrong road. They could have taken the wrong road. (low) Low

Should have + PP vs must have + PP: these two constructions are sometimes conflated. Must have + PP expresses a deduction about what happened (She must have left early = she surely left early, I deduce that). Should have + PP expresses an unfulfilled expectation or a reproach (She should have left early = she ought to have left early, but she did not). The meaning is radically different, even though both concern the past.

Summary table: modals and degrees of probability

Modal Present / future Past Degree
will That will be him. That will have been the problem. Certainty
must She must be tired. She must have been tired. Near-certainty (deduction)
can’t It can’t be right. He can’t have done it. Impossibility
should It should work. It should have worked. High probability (prognosis)
may She may be late. She may have been late. Balanced probability
might He might come. He might have come. Low probability
could It could rain. It could have rained. Low probability

Practical exercise

Test your knowledge:

1. Which sentence expresses near-certainty by deduction?




Must expresses a near-certain deduction based on observed clues, here the car parked outside the office.

2. What does “He can’t have stolen it. He was with me all evening.” mean?




Can’t have + past participle expresses the impossibility of a past action: it is impossible that he did it.

3. What is the difference between “must not” and “can’t” for expressing probability?




Must not (mustn’t) = prohibition. Can’t = impossibility or strong unlikelihood. These are two distinct functions that most learners confuse.

4. How do you express a deduction about the past with must?




All modals combine with have + past participle to express a probability about the past: must have left, might have seen, can’t have known.

5. What is the nuance of “should have left” vs “must have left”?




Should have + PP expresses an unfulfilled expectation or a reproach. Must have + PP expresses a deduction about what happened. The meaning is radically different in spite of the similarity in form.

6. Which adverb expresses the lowest degree of probability?




Conceivably expresses that something is theoretically conceivable without being likely, which is the lowest degree of probability on the list.


Your score

You may also like these articles ...

How to say hello in English?

In a nutshell Formal: Good morning, Good afternoon, Good evening (time-based), Hello Common and neutral: Hello, Hi Informal: Hey, How are you?, How’s it going?,

Read More »
30 July 2026 1 July 2026
Free trial lesson 1 to 1 with a native teacher
Privacy Overview

This website uses cookies so that we can provide you with the best user experience possible. Cookie information is stored in your browser and performs functions such as recognising you when you return to our website and helping our team to understand which sections of the website you find most interesting and useful.