Karen's Linguistics Issues, March 2002 | This Month's Articles | Previous Months


Modals in English Language Teaching

 by Michael Thompson, Italy

If you have ever felt dissatisfied with the way you present modals in your classroom, you are most definitely not alone. As Marianne Celce-Murcia & Diane Larsen-Freeman succinctly put it, “Modal auxiliaries are one of the most difficult structures that you as an ESL/EFL teacher will have to deal with.” (1983: 80).

There are a number of reasons for this, starting with the formal and semantic properties of modals, but a more fundamental factor may be that it is never easy to package complexity into the meaningful chunks of information teachers present to their students. And modals are fascinatingly complex. At the formal level we form modal auxiliaries and determine whether to treat a word as a modal or as another part of speech. At the semantic level, we not only have to deal with the meanings of the various modals, but the different classes of modal found in English. Syntactically, the question is whether all modals are auxiliaries. Since a whole class of modals is used interpersonally, the pragmatics of modal use must be considered. There are dialectal and regional differences, with some speakers preferring to use one modal instead of another. And there is also the question of the relationship between modals and modality. Teachers might well be forgiven for asking themselves how they are supposed to take all of this inherent complexity and translate it into meaningful rules, explanations, and experience that students can take advantage of in the moment.

If you want to reduce complexity and keep it useful, you need to start by understanding what the complexity is. For modals, this means coming to terms with what the modal auxiliaries really are (and aren’t), what the difference is between modal and modality, and what the types of modality are.

What are Modal Verbs?

Most teachers are probably familiar with a list of modals similar to the one offered in Headway Intermediate: can, could, may, might, will, would, shall, should, and ought. Teachers would also probably agree with the explanation given by Headway, namely that these verbs are auxiliaries because “they ‘help’ another verb” (Soars & Soars, 1996: 146). For good measure, Headway tells students that “each modal has at least two meanings” (ibid.: 147). And this, too, jibes well with experience. After all, the must in You must be home by eleven or you’ll be grounded until Hell freezes over is undoubtedly different from the must in You must be Bob. It’s nice to meet you.

On the other hand, what Headway says doesn’t go very far in explaining why only these verbs are considered modals. To start with, what about have to? Have to is not listed as a modal verb, but is discussed without comment as expressing obligation in the same unit. Why would Headway include have to in the list when meaning is discussed, but not offer it as a “modal verb”?

Most students become familiar with the most obvious formal characteristic of modal verbs, namely that there is no -s ending for the third person singular present (no John cans swim). But there are other determining factors on a formal level. The first is that the modal auxiliaries have no non-finite forms (no *I like canning swim or *Sally doesn’t like to must pay her taxes). The second is that you cannot chain modal verbs together in a sentence (no *they will can go to the park tomorrow). There are also a number of formal characteristics that modals share with other auxiliaries. These criteria include 1) inversion with the subject in questions (Can you swim?), 2) forming the negative with -nt ( She can’t play the harmonica), 3) verbal ellipsis (I can sing Swing Low, Sweet Chariot, and so can Joanne) and 4) emphatic stress (but I do like Milli Vanilli; but she can sing Bella Ciao) (from Palmer, 1990). 

Clearly, have to fails the modal verb test. It has an -s ending for the third person singular (John has to swim) It can be used in non-finite forms (Sally doesn’t like having to pay her taxes), and it can be chained with a modal (we will have to go to the park tomorrow). But does it fail the auxiliary test? We cannot invert have to with the subject in questions, at least in current standard English, but is Have I to wash the car? 100% incorrect? The same question can be asked of No, you haven’t to wash the car today-- you can do it tomorrow. And have to can be used for verbal ellipsis (I don’t like paying my taxes, but I have to) and perhaps for emphatic stress as well.

How then, should we treat have to? One traditional solution has been to treat have to as a main verb with a verbal complement, similar to want to. Like want to, have to requires do support, and like have to, want to can be used for verbal ellipsis (Do you want to eat squid? – No, but I have to; Do you have to eat squid? – No, but I want to). But there are problems with this approach. First, have and have to are pronounced differently, the one with a /v/, the other with an /f/. This is something separate from the dialectal differences found in ‘tomayto’ v. ‘tomahto’ and seems to suggest that we have two distinct entries for have and have to in our mental lexicon. The second problem is that have to and want to treat their verbal complements differently. If the sentence I want ice cream is compared to I want to eat ice cream, the meaning of want does not drastically change in the two sentences. The same cannot be said if I have ice cream is compared to I have to eat ice cream.

For some, the meaning of have to is unimportant. “Have to ... though semantically very close to must, has none of the modal properties and is clearly a catenative [verbs with verbal complements, like want to], not a modal.” (Huddleston, 1984: 165) But for others, such as Palmer, the meanings of have to and other forms including to be going to, to be able to, and would rather make them “semi-modals” (Palmer, 1990: 25).

The question of which syntactic role these semi-modals have in a sentence is problematic. Are the semi-modals auxiliaries or main verbs? Palmer does not directly address the question, but in a discussion on be bound to, he does make reference to the main verb being a verb of action, suggesting that he sees be bound to as an auxiliary (Palmer, 1990). Celce-Murcia & Larsen-Freeman (1983: 83) note that periphrastic modals (the semi-modals ending with to) “behave syntactically more like main verbs than do modals”, but a footnote at the bottom of the same page seems to suggest that they still consider periphrastic modals to be auxiliaries. They also note that “Structurally, have to is not truly a periphrastic modal since it requires do support .... In other words, have to looks like a verb and behaves very much like a periphrastic form in many contexts. Thus we have treated it like one.” (1983: 81).

Many of the semi-modals behave oddly. Used to often takes do support (Did you use to live in New York?), while need sometimes acts as a proper modal auxiliary (you needn’t come) and sometimes as a semi-modal requiring do (you don’t need to come). Had better shows the formal characteristics of modal verbs (no -s, no non-finite form, no chaining with other modals), but the presence of better makes treating it as a modal verb problematic, to say the least.

It appears that the semi-modals are hybrid forms, combining characteristics of both main verbs and auxiliary verbs. It also appears that the category is defined by the semantic functions of its members, not their formal qualities. This is important because it suggests that there is no necessary main verb or auxiliary verb characteristic that all semi-modals must share. In other words, students need to calibrate the individual structural characteristics of the semi-modals since each semi-modal has its own combination of main verb and auxiliary verb characteristics. They also need to learn when and how to substitute semi-modals for modal auxiliaries, and to be aware for the subtle changes of meaning these substitutions sometimes indicate.

What is Modality?

These formal qualities do not address the second characteristic found in modals – meaning. The question of what we mean when we say that modals have meaning turns out to be quite a complex one. Palmer notes that modality is concerned with our opinions and attitudes, (1990), and most linguists accept the existence of at least two types of modality, with one more type needed in order to account for the auxiliaries given as a group in Headway. Modal, then, refers to the formal properties of a certain class of words, while modality refers to the meanings of those words (and others).

One type of modality, epistemic, is concerned with the speaker’s judgement of the truth of the proposition embedded in the statement. If I say, for example, “John may go home.”, I am telling you that I am uncertain about the truth of the proposition John is going home. Other examples of epistemic modality are must in “John must be home. Look there’s his car.” and can’t in “John can’t be home yet. His flight doesn’t arrive until 9 o’clock.”

The second primary category of modality is deontic modality. Deontic modality is concerned with “influencing actions, states, or events” (Palmer, 1990: 6); in other words, it is oriented towards performing speech acts – doing things with words, as Austin would say. When I say, “John may go home now.” to give John permission to leave, or when I advise, “Elena should go home. She looks tired.” I am using deontic modality.

These descriptions raise several considerations. The first is that I used the same modal, may, to give John permission to leave (deontic) and to tell my listener that I am not certain if John is leaving (epistemic), which may cause ambiguity (am I being benevolent or hazarding a guess as to John’s next step?). This ambiguity is found throughout the modal system (and not only in English) and is one of the reasons that classroom activities and exercises focusing on modality can be so difficult to develop.

The second consideration is that neither epistemic nor deontic modality accounts for can in John can play the pan flute. You may have noticed that both deontic and epistemic modality refer back to the speaker in some way. In deontic modality, the speaker does something such as giving permission or advice. With epistemic modality, the speaker comments on the probability of the truth of the proposition, perhaps saying that he is certain that it is false (can’t) or that it is reasonable to assume that it is true (should). But the can in John can play the pan flute does not seem to refer to the speaker; rather, it seems to refer to one of John’s abilities. Have to in They have to be in Rome for a meeting tomorrow is also an example of this subject-oriented modality, which is called dynamic modality. Dynamic modality plays an important part in considerations of have to, and its contrast with must.

Epistemically, there seems to be little difference between have to and must (he has to be home; there’s his car/he must be home; there’s his car), with both expressing a sense of certainty. Deontically, must obliges the subject of the sentence to do something (you must be home before 9 o’clock). Have to, on the other hand, does not have a strong deontic role. In dialects where must is rarely used deontically (such as many American dialects) the form for obliging someone to do something is often the imperative. In the realm of dynamic modality, however, the distinctions between must and have to create confusion. When an obligation is clearly speaker-oriented must is a clear choice. When an obligation is clearly external, the choice is have to. But there are many times when a situation is neither “clearly external” nor “clearly speaker-oriented” and here in this “neutral” area many native speakers use must and have to more or less indiscriminately (Palmer, 1990).

Modals in the Classroom

This then is the students’ task. They need to learn the formal properties of the various modal auxiliaries and semi-modals and learn how to manipulate those properties appropriately. Moreover, they not only have to learn three categories of modality (epistemic, deontic, and dynamic), and the uses of the various modals and semi-modals for each, they also need to learn to distinguish between the three categories. Consider the different meanings of can’t in these three sentences: you can’t be Laura’s brother – she’s an only child (epistemic, certainty); I’m sorry, you can’t leave yet (deontic, permission), and you can’t swim (dynamic, ability). (Students also need to learn about the pragmatic considerations of modality, such as the difference between can and may when giving permission, but these are beyond the scope of this article.)

There are a number of ways that teachers can help their students master the complexities of modality. First, particularly for lower-level students, they can design activities that allow students to explore and get comfortable with the formal properties of modals and semi-modals. When designing these activities, teachers should be mindful of the fact that the formal properties of one semi-modal will not necessarily transfer to the next. In fact, many of the mistakes that students make with semi-modals come from over-generalising formal properties.

When developing activities that focus on the meaning of modals and semi-modals, teachers can make sure that there is a strong connection between the words and the meanings. One example of this would be to present students with a sentence containing a modal together with possible meanings: 

You should study Aramaic.

This sentence talks about 1) advice 2) obligation 3) permission 

At higher levels, such an activity could ask students to choose between categories: 

He must be the new Aramaic teacher.

Do you think this sentence talks about 1) certainty 2) obligation 3) advice? 

Or students could be asked to rephrase a sentence using an appropriate modal: 

Suzanne has the ability to speak Aramaic.

Suzanne __________________________ (can speak Aramaic).


Ultimately, “Well, that’s just what we say.” is an unsatisfying answer for student and teacher alike. And the reason it fails to satisfy is that both the student and the teacher can sense the inherent complexity lying beneath the surface. Capturing that complexity and repackaging it in a form that students can use to internalise the complexity into their own language is one of the most interesting, and rewarding, challenges ESL/EFL teachers face.

What we say is, in the end, governed by rules – grammar rules, pragmatic rules, semantic rules, pronunciation rules. There are rules for every facet of the language. If you can deepen your understanding of those rules, then you can articulate to your students why it is that we say what we say. And that is something that rarely leaves people dissatisfied.


Celce-Murcia, M. & Larsen-Freeman, D. (1983), The Grammar Book: An ESL/EFL Teacher’s Course, Newbury House Publishers Inc: Rowley, MA

Huddleston, R. (1984), Introduction to the Grammar of English, Cambridge University Press: Cambridge 

Lyons, J. (1977) Semantics (Volume 2), Cambridge University Press: Cambridge 

McArthur, T., ed. (1992) The Oxford Companion to the English Language, Oxford University Press: New York 

Palmer, F.R. (1990), Modality and the English Modals 2nd Edition, Longman Group Ltd: London 

Soars, L & Soars, J. (1996) Intermediate Students Book New Headway English Course, Oxford University Press: Oxford



©Michael Thompson. All rights reserved.