Teaching modal auxiliary verbs

There are four language analysis guides that you should have done / do / be familiar with before you tackle this one:
    A general guide to modal auxiliary verbs
    A guide to modal auxiliary verbs one by one
    A guide to semi-modal auxiliary verbs
    A guide to types of modality
They all open in a new tab.

If you are happy with that, read on.

It is surprising that on many initial training courses, novice teachers are frequently asked to teach an aspect of modality. Given the complications of the area (with which you'll be familiar if you have read or followed the guide to types of modality linked above), this is a little like asking a first-year violin student to compose a symphony so we should not be surprised when people get it wrong.

Other languages

The concept of modality is common to all languages. We all have a need to express things like willingness, probability, likelihood, obligation, requirement and so on. Modality may not only be achieved by the use of modal auxiliary verbs, of course. Saying something like:
I'm almost certain he'll be late
instead of
He'll be late
is still using modality in language.
Modal auxiliary verbs are, however, very variable across languages. Here's a brief run-down by major language groupings explaining a little of how it all works. There won't be enough detail here for your particular students and your setting but it is somewhere to start. You can get more on line but beware the unreliability of many sites.
Swan and Smith, 2001, is a usually reliable source albeit frustratingly inconsistent in covering modality.

Languages	What they do
Arabic	Standard Arabic does not have modal auxiliary verbs which correspond exactly to English modal auxiliaries. However, it does have many precise and detailed ways to express modal concepts. For example must, have to, should, might, may, it is possible to, it is impossible to, it is expected that, it is easy to, it is hard to, it's worth mentioning that, it's well-established that, it's most likely that, it's forbidden to, it's permitted to, it's more proper to and a range of other concepts are all expressed through a form of modal construction. The problem, of course, is that these categories do not mirror the modal categories of English so expect a good deal of confusion, especially with modal auxiliary verbs like may and could which have a range of functions. The concept of having a large range of modality in the language will not be mysterious to learners from an Arabic-speaking background. (Source: Arabic learning resources at http://arabic.desert-sky.net/g_modals.html)
Chinese languages	Modern Standard Chinese, too, has a range of modal auxiliary verbs, as one would expect of an isolating language. They are, however, not at all parallel to those in English. There are, for example, three modal auxiliary verbs which perform the functions associated with can (ability, permission, possibility) in English. The most commonly used modal auxiliaries in Chinese languages express want; ask for; wish; desire want to; would like to; feel like (something) should; ought to; must can; be able to; be capable of like; love; prefer; enjoy; be fond of can; may be good at; be skilful in be willing to Conceptually, modality poses no problems but overlapping meanings will make life difficult.
Slavonic languages including Russian, Polish, Ukrainian, Czech, Slovak etc.	Slavonic languages also have modal auxiliary verbs but, as usual, the categories don't exactly match (although they are closer than many languages). A single verb form (roughly translatable as must) serves in Polish for the English verbs must, need to and ought to, for example. In Czech, there is a clear distinction between externally and internally imposed obligations, often rendered in English by must vs. have to, should or ought to. Russian has a simpler modal system than English so you may encounter, e.g., He must not used to mean He needn't / doesn't have to.
Germanic languages	These languages use modality is ways quite similar to English which makes life easier for learners from these language backgrounds. The most important exceptions often concern the negative uses. For example, in German the translation of needn't / don't have to, expressing the lack of obligation either way, would be must not [muss nicht] and that causes confusion. To translate the English sense of must not meaning prohibition, German uses a different modal auxiliary verb [darf nicht], roughly translatable as may not. In Swedish, the verb used for the sense of needn't is cognate with the English expression behove. In these languages, too, there are usually more tense forms available so an equivalent of had to may be rendered as musted. Learners may be confused by a logically constructed tense form in their first language, the past or future of must, for example, not existing in English. The overlap in meanings between cognate modal auxiliary verbs is not precise so expect errors such as It can be* used to mean It might be. There are few serious conceptual difficulties otherwise.
Romance languages including French, Spanish, Italian, Portuguese, Romanian etc.	These languages do not have a separate grammatical category for modal auxiliary verbs and usually render the concepts as a verb followed by an infinitive form, e.g., in French Je peux aller (I can go) or in Spanish Puedo ir. Expect, therefore, errors such as I can to eat You must to enter etc. because many learners perceive the to-form as the infinitive in English. Some modal concepts, e.g., for obligation, are expressed in the passive (It is necessary that ...). There is, in these languages, no equivalent to the use of be to express obligation as in, e.g.: You are to be here at 6 and a modal auxiliary verb (often an impersonal one) will be used instead.
Swedish, Danish and Norwegian	Scandinavian languages have a separate class of modal auxiliary verbs although use is an issue. The verb kan, for example, is used to talk about the possible future where English would use may or might. Expect, therefore, *It can rain tomorrow. Other uses of modal auxiliaries are close parallels with English.
Japanese	Japanese has two classes of modal auxiliary verbs: those which are attached to stems and cannot function independently and those which are ordinary verbs which lose their meanings when acting as auxiliaries. We have, therefore, structures such as miru (to see) leading to mirareru (to be able to see). The concept of modality will not be strange but the forms will cause problems in English.
Other languages	Some languages have a very reduced set of modal auxiliary verbs to call on. Greek, e.g., really has only two (must and can), one of which only exists in the It is necessary / necessarily true that sense in the third person. Added to these problems is the tendency to make past modal auxiliary verbs in completely different ways (often rendering the modal auxiliary in the present with the main verb in the past) so mistakes such as I can saw it (for I was able to see it*) are possible. Turkish speakers will have few problems with the modal system of English as such because the language has many parallel structures.

General problems for all learners

Whatever you students' first language(s), there are certain aspects of the English modal system which will cause problems:

Form and function
If you have followed the analysis guides in the area, you will be alert to the issue that in English one verb can perform a range of functions. The modal could, for example, is used to express:

present possibility	That could be the postman now.	epistemic modality
future possibility	It could rain tomorrow.	epistemic modality
past possibility	He could have seen her. I'm not sure.	epistemic modality
present ability	I could help with that, if you like.	dynamic modality
future ability	I could finish in hour if I get some peace.	dynamic modality
past ability	When I was only two, I could swim pretty well.	dynamic modality
permission	Could I ask you a question?	deontic modality
complaint	You could have warned me!	dynamic modality

This works both ways. Permission, for example, can be realised using may, might, can, could etc.
For more, got to the guide to modal auxiliary verbs one by one, linked below.

The modal system in English exhibits a range of anomalies. For example, the negative of
    It must be him
used for deduction is
    It can't be him
not
    It mustn't be him.
There are numerous other cases of non-parallel structures.
Modal auxiliary verbs are often contracted and, in natural speech, often rendered with weak-form vowels. So we get must pronounced as /məs/ and can as /kən/ or even /kn/ etc. as well as needn't've or couldn't've etc. which are hard to perceive and even harder to produce.
Few languages have verb forms analogous to English semi-modal auxiliaries (dare, need, used to) and the forms will be difficult to master.

Approaches to teaching modal auxiliary verbs

Fundamentally, there are two ways to approach the teaching of modal auxiliary verbs (and all modality expressions for that matter):

Form first: take each modal auxiliary verb as a separate entity and present its various functions logically and appropriately, considering the level of the learners
For example:
Consider the modal auxiliary verb can and then teach the functions of:
    I can swim
    It can get cold in the winter here
    You can't be serious
    He can't have missed the train again
and so on.
There are four examples here, each with a different meaning of the modal auxiliary verb.
Function first: decide on the type of modality you want to tackle and then select the modal auxiliary verbs which express the function appropriately, considering the level of the learners
For example:
    Obligation: teach must, have to, should, ought to, had better etc. according to level and previous learning
    Ability: teach can, could, be able to, could have, was able to etc. according to level and previous learning
    Possibility: teach might, could, must, can't have, couldn't have, will etc. according to level and previous learning
and so on.

To illustrate, you can conceive the relationship in these two ways:

Approach 1	Approach 2

If some of the terms are unfamiliar to you, follow guide to types of modality or simply ignore them. The following will still make sense.

Each has advantages and drawbacks.

If you take Approach 1:
- you risk encouraging your learners to think that there is a one-to-one relationship between the verbs and their functions
- you are going from form to function rather than function to form and that's not how the mind operates
- it will be reassuring because this is the way many courses are structured and many coursebooks written so it's familiar to both learners and teachers
- it allows for bite-sized lessons focusing on a single use of a single verb
If you take Approach 2:
- you need to make intelligent decisions regarding which and how many verbs to present for each type of modal function
- you may find it difficult to fit the approach within a school syllabus because that is not how they are usually constructed
- you will be moving from function to form and that fits with how the mind operates in this area
- you will be giving your learners an appropriate range of ways to realise the functions they need to communicate

Some people may take, of course, Approach 1 at lower levels and move to Approach 2 when the learners have succeeded in mastering a range of simple modal auxiliary verbs expressing a limited range of functions.
Whichever route you choose, the considerations which follow are applicable.

Teaching modal auxiliary verbs

To master any modal-verb use we need to know two things:

the speaker/writer's perception of reality
If a speaker chooses to say
It'll rain
rather than
It might well rain
the intention (arising from the speaker's view of the future) is subtly different
the context in which it is used
It is impossible to interpret something like
    I could leave earlier
without knowing the context. It can mean
    I was given permission to leave earlier
    I was able to leave earlier
or
    I may (in future) leave earlier
We need to supply a good deal of information for learners to be able to interpret the sentence accurately:
- How is the event situated in time (future or past)?
- What is the speaker's relationship in terms of authority with others?
- Is the speaker likely to be talking about ability or permission, given the preceding text?

For example, the statement
She must write to her mother
has a number of possible interpretations. What are they and how does the hearer unpack the meaning? Click here when you have an answer.

She is obliged to write to her mother
She wants very much to write to her mother
She should write to her mother
She ought to / it is her duty to write to her mother

Only knowledge of the context of the utterance and the speaker's relationship to events (and perception of them) can tell us which meaning is intended. We may also draw on our knowledge of the speaker's habitual responses to events and situations.

Handling other issues

The one-form-many-functions and one-function-many-forms problems need careful handling. It is unwise to try to teach any modal expression in isolation. What we need to do is to focus on particular functions of the form independently. Trying to teach may in contexts where it means both possibility and permission:
    It may rain
vs.
    May I ask a question?
is a recipe for disaster. We have seen above that many languages will reserve separate structures for these functions.
A much better approach is to focus on the concepts of types of modality and teach, say, may for epistemic modality:
    That may be true
separately from may for deontic modality:
    You may not leave yet
This way, logical constructs will not be confused. If terms like epistemic and deontic are new to you, follow the guide to types of modality from the link at the end.
Anomalous forms such as the difference between must not and don't have to / needn't also need careful handling in context. The context must make it clear what sort of obligation (if any) is present.
Equally, modal auxiliary verbs of deduction must be presented in a way that makes it clear what sort of deduction can be made with what level of certainty. Here's an example of something commonly used. What are the problems?

Click here when you have detected the two problems with exercises like this.

Problems:

There is a mix of tenses. Are the learners expected to produce the perfect forms (or the perfect progressive) or the present tense forms?
What level of certainty is permitted? Some people are happy to make assertions based on very little evidence; others are more cautious. If you find cosmetics, how sure can you be that the tenant left them (or that the tenant was female)? Was the tenant Spanish or learning Spanish? Given the inexactitude, you can never be sure that the learners' responses are 'correct' in the sense of matching the expressions they select to what their thought processes are.

There are ways to make such exercises work well and be intriguing but it needs a bit more effort:
We need to make the context clearer. One way is to kill off the ex-tenant so deductions can only be in the past!
To tackle issues of certainty (epistemic modality), learners can be asked to express in percentage terms how certain they are and compare their responses with others so that in feedback you can judge whether the statement matches the perception. Additionally, giving a little information about the previous tenant would help.

We need to provide lots of targeted listening and production practice in the weak and contracted forms of modal auxiliary verbs. Drilling is one way but getting learners to select from a multiple-choice list whether they heard, e.g., must've, mustn't've, could've, couldn't've etc. embedded in natural-speed utterances is another way to prepare them for listening in 'the real world'.
Semi-modal auxiliary forms are increasingly rarely produced as central or pure modal auxiliary verbs. For example, dared not is usually now didn't dare, didn't used to is preferred to usedn't to and so on. We need to ask ourselves whether at most levels it is wise to introduce the modal forms at all (except where they may occur and then only for comprehension). This means de-selecting texts which contain them.
The subtle difference between needn't have and didn't need to is another issue we could leave well alone until learners are really competent in using modal auxiliary verbs.

Degrees of likelihood (epistemic modality)

This is an area which deeply confuses many learners for a number of reasons:

Because many languages have only one modal to talk about probability and use adverbials to increase / lessen the sense of how probable something is.
Because English uses at least 4 separate modal auxiliary verbs to express probability: might, may, could and can. It also uses will, incidentally in, e.g.:
That will be the postman at the door.
Because it is often taught in a confusing manner providing false, or at least questionable, information.

There is a school of thought which attempts to represent degrees of probability along some kind of cline from, say, 10% probability up to 90% probability and then arrange the modal auxiliary verbs on the cline so we get something like:

0%	← ← ← ← ← probability → → → → →			100%
can't	could	might	may	must

This is actually misinformation and only serves to confuse the learner, primarily because it isn't true. When you have thought about why this is the case, click here.

Intonation, stress, body language and facial expression all play a crucial role.
For example, if you stress the modal, especially with a rising or falling tone on it, you change, sometimes dramatically, the degree of doubt or uncertainty you are expressing. For example:
    He might↑ come
but I very much doubt it
    Of course, the world could↓ end tomorrow
but I doubt it will
    She may↑ be an engineer
but I have no idea one way or the other and if I draw out the word may, I can often imply that I doubt someone is telling the truth
She must↑ be there by now!
and I’d be astonished if she wasn’t. This example is actually deduction rather than probability per se but the concepts are closely related and the source of more confusion if handled carelessly.
Accompany the statement with a shrug and it usually implies 50-50 certainty whichever modal auxiliary verb you use.
In the negative, something else happens.
    He might↑ not come
but I think he will
    He might not↑ come
but I’m almost certain he will
    He could not come = He wasn’t able to
and isn’t to do with probability
    He may not come =
Either:
    He isn’t allowed to come
or
    He might not come
In written English (if you exclude underlining and bold text) it’s the co-text, adverbial use, topic and tense which matters. For example,
When talking about events far in the future, uncertainty is naturally greater whatever verb is used, e.g.:
    Global warming may / might / could result in serious problems over the coming century.
If we use an adverbial, we can alter the sense dramatically in things such as:
    It could easily rain
    She might conceivably fall
    You may naturally be wondering
    There’s an outside chance that it could arrive today
    Personally, I think it’s almost certain that he won’t have the money but he could
etc.
Not all adverbials can be used with all modal auxiliary verbs:
    *She might easily fall
    *You may easily be wondering etc.
All three of these verbs, may, might and could, incidentally, refer to future possibilities not present ones so the speakers' choice will often be determined by how they feel about the future. Only can is used to talk about present possibility / probability as in things like
    It can be cold here in winter.

The general rule is that these modal auxiliaries do not represent different levels of uncertainty or probability in themselves. What matters are all the other signals in the text or in the speaker’s tone and or gesture / expression. In fact, they can all vary between 1% and 99%. Both might and could are used almost interchangeably in English although in plain written text there are those who say might sounds more improbable. That is arguable. The verb may is almost always more formal.

Genre approaches to teaching modal auxiliary verbs

If you have followed the guide to genre, you will aware that certain text types will use modality in very different ways.
When we teach modal auxiliary verbs, especially to learners who need to write conventional texts, it is worth bearing this in mind.

Instructions sometimes employ deontic or alethic modal uses such as must:
often in passives:
the flame must not be allowed to touch the oil etc.
or as imperatives:
do not over tighten the screw
Discussion texts (arguing both sides of a case) will often contain hedging modal auxiliary verbs:
epistemically: in expressions such as:
It might be argued that ...
or dynamically: in formulations like:
One could consider ...
Recounts and Narratives may employ modal auxiliary verbs:
epistemically for speculation and deduction as well as possibility:
    I don't know what I must have been thinking
    I must have been mad
    It couldn't have been worse
    I can't tell you
etc.
and dynamically:
    I wasn't able to catch the train
    I couldn't get to the station in time
etc.
Information reports contain very little modality normally because they don't usually deal with possibilities and so on, confining themselves to facts.
When modality is employed, it is often alethic:
The eggs must be incubated for 20 days
The solution has to be heated to 150°C
Explanations often involve modal auxiliary verbs:
epistemically of possibility: can cause ... , may result in ... etc.
or dynamically of ability / willingness: can't break, won't be airtight etc.
Expositions (arguing one side of a case) employ
deontic modality to make a case: should be, cannot be, must be, ought to be etc.
and epistemic modality to appear certain: this will surely result in ..., that must lead to ... etc.

In writing lessons, in particular, it is useful to focus on modality because a genre approach, by its nature, points learners towards noticing how the intention of the writer is realised in the use of language structure.

Other guides to do with modality are:
essential guide to modality	a simpler guide in the initial training section
central modal auxiliary verbs	a traditional view taking each modal auxiliary verb in turn and identifying its function
types of modality	for more on types of modality mentioned here: epistemic, deontic, dynamic and alethic
genre	for a guide which considers the types of modality which occur in certain text types
semi-modal auxiliary verbs	which also considers marginal modal auxiliary verbs such as seem, tend, be about to etc.
complex tenses	which also considers complex tenses in relation to modality
modality: tense and aspect	which considers modal auxiliary verbs and perfect and progressive forms and also has some teaching ideas
functions index	modal auxiliary verbs play an important part in expressing many other functions such as expressing (un)certainty, obligation etc.
hedging and modality	for a guide to how these concepts are used in English for Academic Purposes
The modality map	for more choices

References:
Campbell, GL, 1995, Concise Compendium of the World's Languages, London: Routledge
Swan, M and Smith, B (Eds.), 2001, Learner English, 2nd Edition, Cambridge: Cambridge University Press