Modals of Possibility & Probability

Epistemic Modals

Epistemic Modal Verbs can broadly be categorized as Modals of Possibility & Probability as they convey the speaker's assessment of whether the verb phrase is real or how likely it is. All core modal verbs can be used with the Epistemic flavor in some way, as can the marginal modal ought [to] and the pseudo-modals have to and have got to. Subsets of the Epistemic flavor includes Modals of Deduction for past situations and Modals of Expectation for future.

Contents

Meaning

The Epistemic Flavor

The Epistemic flavor of modal verbs expresses the speaker’s speculation regarding the likelihood of the verb phrase.  A flavor is a category of modal verbs wherein the uses of the various modals share commonalities in both their meaning and their structure.  The term epistemic comes from the Greek word for ‘knowledge’.

Modals of Speculation reveal the likelihood of the scenario, according to the speaker.  That conclusion may be based solely on reasoning (where the modal has an internal, subjective source) or it might be backed up with evidence (where the modal has an external, objective source).

Each of the core modals is either proximal or distal.  In the Epistemic flavor, proximal modals (can, may, must, shall, and will) are for real situations, while distal modals (could, might, should, and would) are more often used for unreal situations.

Epistemic Force

The various expressions of modality each align with a different modal force which conveys the level of the speaker’s conviction and determines the overall meaning.  Keep in mind that the exact level of strength isn’t fixed, but can fluctuate depending the the speaker’s intonation.  Modal expressions of the same force are broadly considered synonyms, but there are distinctions among them.

  • POTENTIAL FORCE: the situation is possible.
    • can/could — acknowledging the possibility, whether likely or unlikely
  • WEAK FORCE: the situation is somewhat likely or – the speaker is uncertain
    • may/might — admitting uncompelling likelihood
  • MODERATE FORCE: the situation is very likely or – the speaker is moderately certain.
    • should — subjectively expecting or deducting
    • ought [to] — objectively expecting or deducting
  • COMPELLING FORCE: the speaker is confident that the situation is true.
    • must — confidently claiming likely situation
    • have to — confidently claiming likely situation with bias
    • have got to — confidently claiming likely situation with strong bias
  • ASSERTING FORCE: the speaker is convinced (and won’t consider counterarguments).
    • will/would — asserting a known outcome, usually without evidence
    • shall — formally asserting a known outcome, usually without evidence

The strength of an Epistemic modal may be a reflection of the speaker’s level of certainty, or it may be a reflection of the perceived likelihood.  For instance “That might be true,” could mean that the speaker is confident that the probability is not very high, or it could mean that the speaker isn’t certain one way or the other.

In the Epistemic flavor, modals based on objectivity (like evidence) tend to carry more weight than subjective modals.  Therefore, ought [to] and must are more convincing than should and have to, respectively.

Epistemic Negatives

There are two different ways that negators (like not and never) can affect verb phrases with modals:

  1. Event Negation — the main verb is negated; modality expresses the improbability or impossibility of the verb phrase.
  2. Modal Negation — the modal verb is negated, which often makes the entire clause inert; modality expresses a lack of conviction regarding the likelihood of the verb phrase, like the speaker admitting that they simply aren’t sure.

Most modals enact event negation, including can/could, must, shall, should, will/would, and be.  If you are told, “We must not have left the book on the shelf,” you can surmise that the book is not on the shelf.  May/might and have enact modal negation.  If you are told, “We might not have left the book on the shelf,” you can’t be sure whether the book is still on that shelf or not.

The Epistemic ought [to] and have got to are rarely used for negatives.

Structure

Modal verbs come at the start of their verb strings.  The core modals are true auxiliary verbs.  The marginal modal ought [to] almost qualifies as a modal auxiliary, but falls short in just a couple of ways.  Then there are the pseudo-modals, which take the position of auxiliary verbs, even though the don’t really function the same way auxiliaries do.

  • CORE MODALS: can/could, may/might, must, shall, should, will/would
  • MARGINAL MODAL: ought [to]
  • PSEUDO-MODALS: be, have, have got

Note that the pseudo-modals be and have, which are always followed by a Full Infinitive,
are different from the auxiliary verbs BE and HAVE.

Verb Forms

Details about the form depend on whether a core modal, marginal modal, or pseudo-modal is used.

  • Defectiveness:
    1. Core modals and ought [to] are defective, which means they only come in their base forms (and subject-verb agreement doesn’t change that).
    2. Pseudo-modals take on different forms just like normal verbs do, including taking the -s form following 3rd-person singular subjects.
  • Negation:
    1. For core modals and be, the negator comes after the modal (as in “may not” and “shouldn’t”).
    2. For ought [to], the negator comes after ‘ought’ and either before or after ‘to’, or it can replace ‘to’.
    3. For pseudo-modals (except be), we use DO + a negator before the pseudo-modal (as in “don’t get to” and “doesn’t have to”).
  • the Following Verb:
    1. After the core modals, the next verb should be in its Bare Infinitive form.
    2. After ought [to] and the pseudo-modals, the next verb should be in its Full Infinitive form.

Keep in mind that the verb immediately following a modal is often the main verb, but not always; instead, it might be another auxiliary verb.  In that case, that next auxiliary takes on the necessary infinitive form while the main verb could be in any verb form to communicate any tense or voice.  For instance, we might have “will have taken”, “can be taken”, “may be taking”, etc.

For more details on the forms of modals, visit the Modal Verbs page.

Past Modals

In the Epistemic flavor, past modals are primarily conveyed through Perfect Modals.  A Perfect Modal is the combination of a modal followed by ‘have’.  Any of the core modal verbs can be used in this way, though can have and shall have are rare.

The Perfect ‘have’ should only ever be in its Bare Infinitive form (never ‘has’ or ‘had’), which keeps consistent with the third modal property listed above.  For the standard modals, ‘have’ can take its contracted form and affix to the end of these modals.  The following verb should be in the Perfect Participle form.

  • “I might have locked my keys in the car.”
  • “She could’ve called when my phone was off.”
  • “The intruder should have left some trace of his actions.”
  • “The guests will have eaten all the cake.”

For ought [to] and the pseudo-modal have, the Perfect ‘have’ comes after ‘to’, as in “He ought to have finished his exam by now,” and “The detective has to have figured out the solution.”  Since have got to only works for the present, it has no past form.  Only be takes on a past form other than a Perfect Modal, instead taking on the same past forms as other be verbs (eg: “The presidential address was to air at 6pm.”)

Periphrastic Substitutes

Since modal verbs are defective meaning they do not change form for tense or for subject-verb agreement they can be awkward or otherwise difficult to use in some situations.  Also, it is sometimes unclear which flavor is intended, leading to an ambiguous meaning.  For these reasons, modals can be substituted with periphrastic phrases.  Suitable periphrastic substitutes for the Epistemic flavor include (but are not limited to):

  • be likely to / be supposed to / be expected to
  • be bound to / be certain to / be sure to
  • be about to (near-future expectation)
  • be going to

The pseudo-modal have (to) can also be used as a substitute, at least for must.

These substitute phrases are not bound to the same restrictions as modal verbs, but instead behave like normal verb phrases would in the framework of Verb Patterns.

Apart from using periphrastic expressions, a speaker has alternate methods of communicating Epistemic modality through adverbs like ‘possibly’, ‘probably’, or ‘likely’, or through introductory phrases like “I think that”, “I bet”, etc.

Usage

Deduction

core modals, have to

In their Perfect Modals form, Epistemic modals are used to speculate on the past.  They are often used for deduction in which the speaker uses reasoning to develop some conclusion, perhaps in light of some evidence.  A classic scenario is to play the role of a detective at a crime scene.

  • “The burglar could’ve broken in any time between midnight and 4am.” ← With could, the detective establishes a set of possibilities.
  • “They might’ve had blueprints to the museum.” ← With may or might, the detective states a probable situation, though with low certainty (perhaps there is not enough evidence to confirm).
  • “They must have disabled the alarm system.” ← With must or ‘had to’, the detective states a conclusion with a high level of confidence.
  • “They will have known the guards’ schedule.” ← With will or would, the detective gives a statement as if it were obvious or as if there were no reasonable alternative.

Expectation

core modals, ought [to], be to, have to

Whereas deductions look to the past, expectations look to the future.  In their basic form (not Perfect), any of the Epistemic modals can express a future prediction at varying levels.

  • Noting a possibility: “It could rain next weekend.” ← this doesn’t express whether the speaker believe it will rain, only that they acknowledge the possibility.
  • Noting a probability, from low certainty to high certainty:
    • “It may rain next weekend.” ← the speaker isn’t sure.
    • “It should rain next weekend.” ← the speaker believes it likely.
    • “It must rain next weekend.” ← the speaker has absolute conviction.
  • Noting a given: “It will rain next weekend.” ← the speaker removes their subjective assessment and simply states what they (and others in the know) understand to be unquestionably true.

Hypotheticals

distal core modals, ought [to]

Distal modals (could, might, should, and would) can also be featured in hypotheticals, in which the established context is unreal.  But in some cases, the situation may be theoretically possible.  The Epistemic modals express likelihoods within that unreal context.  Would is especially common here.

  • “James would love Paris.” ← Presumably, James doesn’t currently love Paris only because he hasn’t been their yet.
  • “What’s the worst that could happen?” ← This questions the possible consequences of an action that might not have happened yet.

One subset of hypotheticals is counterfactuals.  Whereas many hypotheticals have the potential to be realized, counterfactuals will never be true.  They are most often used for unreal past scenarios, and thus are usually expressed through Perfect Modals.  Should’ve is very common here.

  • “Clara ought to have seen that coming.” ← It was very likely for Clara to see it coming; but against all odds, she didn’t.
  • “Martin would’ve said ‘yes’ had Tina asked him out.” ← But Martin didn’t say ‘yes’ because Tina hasn’t asked him.

States

core modals, ought [to], have to, have got to

For the real present or for real situations that aren’t time-specific, Epistemic modals most frequently allude to states, rather than actions; the main verb in the verb string is most likely a Stative Verb like be, have, seem, cost, etc.  For example: “My keys must be around here somewhere.”

Alternatives

Introductory Clauses

Another way to express Epistemic modality is by starting with an Introductory Phrase, and then place the subject and event in question into a following Content Clause, as in “I think Linda got lost,” or “It’s possible that Kenan forgot his keys.”  Intros typically come in three varieties:

Subjective Intros, Type I

Structure: I [cognitive verb] that/whether …

These clauses generally reflect the personal assessment or understanding of the speak, and so they have 1st-person subjects.  The introductory verb should be a cognitive verb.  Finally, the linker should be ‘that’ following a judging verb, but ‘whether’ following a skeptical verb.  Here are some verbs we might use for this type of introductory phrase:

  • judging verbs: assume, believe, bet, conclude, deduce, doubt, expect, figure, gather, guess, hypothesize, imagine, infer, presume, reckon, suppose, suspect, theorize, think
  • skeptical verbs: doubt, question, theorize, wonder
Subjective Intros, Type II

Structure: I am/feel [adjective] that …

Again, these clauses generally reflect the personal assessment or understanding of the speak, and so they have 1st-person subjects.  The introductory verb should be a Linking Verb, usually be or feel, and usually in the Present Simple tense.  Next, comes the modal adjective.  Finally, the linker should be ‘that’.  Here are some adjectives we might use for this type of introductory phrase:

  • certain, confident, convinced, doubtful, positive, sure, unconvinced

These can be negated with “I’m not [adjective] that” or “I don’t feel [adjective] that.”

Objective Intros

Structure: It is/was [adjective] that/whether …

These clauses are generally meant to be impersonal and therefore begin with the ___ pronoun it.  The introductory verb should be a Linking Verb: usually be, but sometimes appear, become, feel, look, or seem.  Next comes the modal adjective.  Finally, the linker should again be ‘that’ following a judging verb, but ‘whether’ following a skeptical verb.  Here are some verbs we might use for this type of introductory phrase:

  • judging adjectives: apparent, clear, certain, conceivable, expected, evident, inevitable, likely, obvious, plausible, possible, probable, unavoidable, unlikely
  • skeptical adjectives: ambiguous, debatable, disputable, doubtful, dubious, inconclusive, questionable, uncertain, unclear, unpredictable

Adding a negator in front of judging adjectives negates the modality when they are adjectives of perception (‘never clear’, ‘not obvious’, etc.), but negates the event when they are statistical adjectives (‘not likely’, ‘not possible’, etc.).

Adjectives + Full Infinitives

One more way to express likelihood is with be [adjective] to [verb phrase],” as in “Ginny is expected to attend the gala tomorrow.”  Some adjectives that might work for this include:

bound certain due expected liable likely supposed sure

Advanced Concepts

‘May’ & ‘Might’ for Mitigation

A special case for may and might is that they can be used to acknowledge conditions only to assert their irrelevance or lack of significance.  The phrase “While that may be true,” is often followed by a statement that mitigates the concern.  The sentence “I might not be young anymore, but I can still run a marathon,” shows that the speaker’s loss of youth is of little importance when it comes to the fact that they are still active and fit.

Intrinsic Possibility

A sister category to the Epistemic modal flavor is the Intrinsic Possibility flavor, also called the Alethic flavor.  This flavor communicates likelihood in a very similar way.  But the key difference is that the likelihood reflects the intrinsic nature of the situation instead of reflecting the speaker’s speculation.

Modality expresses the speaker’s conviction regarding the verb phrase.  Intrinsic Possibility is tangential to pure modality because it does not reflect the speaker’s own speculation.  In fact, the speaker’s assessment has nothing to do with it.  Nevertheless, we use modal verbs to convey Intrinsic Possibility.

For example:

  • “We offer a warrantee because we know that accidents can happen.”
  • “This work must have been painted by some artist; we just don’t know whom.”
  • “The sun will rise again tomorrow.”

These statements a unquestionable true and are not up for debate (unlike Epistemic modals; for which a different point of view or new evidence may yield different conclusions).

Teaching Tips

Separate Flavors

Students may recognize that modals are used in different ways.  They are probably first introduced to can as a Modal of Ability, but now they can see its also a Modal of Possibility.  Perhaps they also know it can be a Modal of Permission.  This might seem confusing, so you may want to make it clear that these are three different uses of can.

Point out that there are other words with multiple meanings like ‘right’ (opposite of ‘left’ or opposite of ‘wrong’), ‘bat’ (an animal or baseball equipment), ‘ring’ (on a finger or on the phone), and plenty of others.  Each of the core modals is like that.

Also, it may help to communicate that each modal verb has consistent force across all its meanings.  For example, must is always strong, so it probably wouldn’t be used if the speaker feels ambivalent.  Conversely, may and might are weak, so someone with absolute conviction wouldn’t use either of them to argue their point, no matter which flavor is relevant in that situation.

Bit by Bit

As complex as the application of modal verbs are, it’s probably best to teach them a little bit at a time, going just a bit deeper with each proficiency level that students advance through. For instance, you might want to start with basic applications of the positive form, then later introduce the negations, then later introduce common periphrastic substitutes, then later still cover Perfect Modals and past applications.

Thankfully, most coursebook series use this method across multiple levels of their books.

Sliding Scales

Consider the different strengths of modal force for the Epistemic flavor, and have your students place them in order or otherwise identify the distinctions between them.

  • For audio learners: Have the students state the modal pairs aloud, beginning with uncertain and ending with certain.
  • For visual learners: Have the students write the modal pairs in order from top to bottom.  Show uncertainty at the top and complete confidence at the bottom.
  • For kinesthetic learners: Write each modal pair on a card and have the students arrange the cards from top to bottom.  Show uncertainty at the top and complete confidence at the bottom.

At a higher level, add in the negatives.

Also at higher levels, add in ought [to] and the pseudo-modals.

Further Reading

Encyclopedia Pages