How to teach hedging and generalising

Summary: How to teach and practise language to stop students overgeneralising and making stronger or more vague statements than they mean to

Although hedging language like “It could be said that…” and generalising language like “Most people…” are often taught as separate language points, they both share the function of stopping students making bold overgeneralisations like “Japanese people like…” or “People in my hometown work as…” I therefore tend to teach these two functions in the same lessons, and in many more than the Academic Writing classes that hedging and generalising are most often get covered in. As well as IELTS and FCE Speaking exam preparation (where the two real quotes above come from), these functions are vital for talking about cultural differences, and can be a good way of expanding on the adverbs of frequency, modals of probability/ possibility, certain and uncertain language, concession expressions, conditionals etc that are taught much more often. Hedging/ Generalising is also easy to combine with tips for things that they might have to do in English like taking part in debates, and almost any topic (families, mental health, intelligence, gender, etc) that students might want to generalise about.

What students need to know about hedging and generalising

Students almost always know some adverbs of frequency, and a good way into hedging and generalising is to expand their range to include other ways of answering “How often?” like “almost always”, “very rarely”, “generally”, etc. In approximate order, a fairly complete list would be:

  • (absolutely) always
  • almost always
  • usually/ generally/ most of the time/ normally/ typically
  • often
  • fairly often
  • sometimes
  • occasionally/ once in a while
  • quite rarely
  • rarely/ seldom
  • hardly ever
  • very rarely
  • almost never
  • never

“Very often” is also useful, but it’s not clear where it would come in the list with respect to “usually” above.

Expanding on their adverbs of frequency can lead into similar expressions like “tend to”, “in general”, “as a general rule” and “to generalise”, or I tend to next work on answers to “How many/ much?” like “A great number”, “A very large amount” and “A small minority”. It’s amazing how many students don’t know that “one or two”, “a couple”, “a few” and “some” are in that order by size! A fuller ranking could include:

  • an absolutely huge amount/ number
  • a huge/ massive/ vast amount
  • so many/ so much
  • very many/ very much/ a very large number/ a very large amount
  • many/ much/ a large number/ a large amount
  • a considerable/ reasonable/ reasonably large/ sizeable/ substantial amount
  • quite a lot/ a fair number/ a good amount
  • some
  • a fairly small number/ a limited amount/ not so many/ not so much
  • a few/ several/ a small number
  • a very limited number/ amount
  • a tiny amount/ number
  • an absolutely tiny amount/ number
  • a couple
  • one or two

There are also ones which only suggest a fraction/ percentage/ ratio rather than an actual number, making them more like answers to the question “What proportion?” Examples include:

  • (absolutely) all/ every/ everywhere/ everyone/ everybody/ everything
  • very nearly/ virtually all
  • almost/ basically/ nearly all
  • the vast majority/ a huge proportion
  • a very large majority
  • a majority/ most/ mainly
  • a substantial portion of
  • a considerable minority
  • few/ a small (but significant) fraction/ section
  • a tiny minority

Other useful hedging/ generalising expressions when talking about “How many/ much?” include ones meaning “approximately” (“a ballpark figure is…”, “more or less”, “around”, “in the area of”, “in the neighbourhood of”, “or thereabouts”, “plus or minus tens percent”, “with a margin of error of 20,000”, “give or take a thousand”, etc), “(just) under” (“almost”, “nearly”, “(slightly/ far) fewer/ less than” etc) and “(just) over”. There are also less useful vague expressions like “some”, “by no means all” and “nowhere near all”.

The next obvious thing to cover is “How likely?”, with expressions like “probably”, “almost certainly” and “perhaps”, plus verbs like “really must”. The easiest ones to rank are:

  • certainly/ definitely/ (really) must/ assuredly
  • almost certainly
  • very probably/ there is a great chance/ all too likely/ there is every chance
  • there is a very good chance
  • probably/ likely/ there is a good chance that/ the odds are
  • perhaps/ maybe/ might/ may
  • could/ possibly/ potentially
  • could possibly/ could potentially
  • conceivably
  • there is little chance
  • there is very little chance
  • there is next to no chance
  • there is no chance/ can’t
  • there is absolutely no chance/ really can’t

We can expand this to include certain and uncertain phrases that are more difficult to rank like:

Certain (positive)

  • evidently/ obviously/ clearly
  • undeniable/ undeniably/ there is no doubt that
  • full consensus about…
  • conclusively proven/ has been established
  • confidently predict
  • well-documented
  • we can safely assume
  • convincing argument/ convincingly argues
  • there is (very) little doubt that


  • hypothetically/ theoretically
  • arguable
  • not necessarily
  • controversial
  • debatable
  • unreliable
  • on shaky foundations
  • claim
  • apparently
  • supposedly 

Certain (negative)

  • unjustified assumption
  • unfounded rumours
  • we can reject…
  • invalid
  • demonstrably false
  • inconceivable

There is obviously a lot of overlap between the categories above – another good reason for teaching them together as I’ve suggested here!

A different but very useful (for this and more generally) point is concessions like “…but there are (obvious) exceptions such as…”, “apart from” and “except for”. Others include:

  • … but it depends on/ there are also other options/ you can make an exception if…
  • While it’s generally true that…/ Although it’s often the case that…

There are also several similar conditionals and related forms such as:

  • If/ When/ Where (it is) (at all) possible/ practical
  • Only if/ when/ where
  • Providing/ As long as
  • As much as (is) possible/ practical
  • Unless…
  • In most cases/ In the majority of cases
  • In normal circumstances

It can also be useful to teach similar expressions which are used almost like an apology at the beginning of a generalisation such as “Among the people I know”, “This is a bit of an oversimplification but…”, “This is a bit of a stereotype but…” and “To overgeneralise/ simplify somewhat…” You could also introduce:

  • It’s impossible to be precise, but…
  • To make a somewhat sweeping statement
  • If we ignore... (for a moment),…
  • This is just ballpark figure, but…

Some time could also be spent on changing the forms of the words above, e.g. adding prefixes like “over-“ and “un-“, or adding “–ation” to “generalise” to make the verb.

Typical student problems with hedging and generalising

There are many collocations involved with teaching this language point, as seen in many of the example sentences above. Some like “vast + majority” and “could + possibly” make sense, but I’m not sure it is possible to explain away why we often say “a good + number” but not “a good + percentage”.

Other useful collocations include, in alphabetical order:

  • a + certain/ considerable/ huge/ large/ small/ substantial/ tiny/ very large + proportion
  • a + certain/ considerable/ substantial + percentage
  • a + certain/ small + section
  • a + certain/ small/ tiny/ very small + fraction
  • a + couple/ few + of + the
  • a large/ the vast + majority
  • a/ an + absolutely huge/ absolutely tiny/ considerable/ fair/ fairly limited/ fairly small/ good/ huge/ large/ limited/ massive/ reasonable/ reasonably large/ sizeable/ small/ substantial/ tiny/ vast/ very large/ very limited + amount/ number
  • a/ an + considerable/ insignificant/ large/ small (but significant) / substantial/ tiny + minority
  • absolutely/ almost/ basically/ by no means/ / nowhere near/ practically/ virtually + all
  • almost + all/ always/ certainly/ every/ never/ no
  • certainly/ definitely + not
  • common + misconception
  • complete/ full + consensus
  • conclusively + proven
  • could + perhaps/ possibly/ potentially
  • definitive + proof
  • expert + opinion + suggests
  • extremely/ fairly/ very + common/ large/ likely/ often/ small/ unlikely
  • fairly + certain/ common/ large/ likely/ small/ sure/ uncommon
  • far/ many/ much + fewer/ less + than
  • few/ many/ much/ some + of + the
  • give + or + take
  • in + general
  • in + principle
  • just + over/ under
  • leading + authorities
  • not so/ so/ very + many/ much
  • notable/ obvious + exception(s)
  • on + average
  • on + some + occasions
  • once + in + a + while
  • only + rarely/ some/ sometimes/ very few
  • or + thereabouts
  • over + simplification
  • plus + or + minus
  • quite + rarely
  • really + must
  • safely + reject
  • tend + to
  • there is + absolutely no/ a good/ a great/ a very good/ every/ little/ next to no/ no/ very little + chance
  • there is + no/ (very) little + doubt that
  • to + over- + generalise/ simplify + somewhat
  • un + common/ likely
  • unconvincing + claim
  • under + some/ special + circumstances
  • unfounded + assumptions/ rumours
  • very + common/ few/ likely/ little/ many/ occasionally/ often/ probably/ rarely/ seldom/ typically/ unlikely/ widespread
  • with + a + margin + of + error + of

These expressions can also bring up some grammar, e.g. determiners and adverbs with limiting/ extreme adjectives.

As well as the problem of overgeneralising that teaching hedging is supposed to cure, students can also suffer from the opposite problem, throwing in random expressions from the lists above in an attempt to make the sentence sound more academic and justifiable. This can produce sentences like “It could perhaps be said that some young people sometimes think…” which could be used about any topic in the world and so give no information about either young people or the views of the person writing/ speaking. Native-speaking teenagers are also prone to answering every question with “It depends”, but I find it helps to tell students that the purpose of hedging/ generalising language is to give more precise information about number, probability etc, rather than to obscure what they think.

More specific language problems include using “almost people” to mean “almost all people” and “I almost fall asleep on the train” to mean “I almost always fall asleep on the train”.

Some expressions cause confusion due to having two meanings. For example, “a fraction” could include “234/27” in the field of maths, but when generalising in conversation it is only really used to talk about an amount under 50%, e.g. “a small fraction” like 1/7. Also, “a majority” in voting could include 50.5%, but “a majority of my friends” usually means around 70%. “A considerable/ substantial amount/ number” could also be said to have two meanings, literally meaning less than “many/ much/ a lot” but in fact often used as a softer or more academic-sounding way of saying exactly that.

“Will” and “must” can cause problems. “Must” sounds strong, but because it is a logical assumption rather than a fact it could be said to be weaker than “will”. There are then the even stronger forms “really must” and “definitely will”. All four forms are 100% if you put figures on them, though it can help a little to write the last two forms as “100%!” There are similar problems with “surely”, which is more of a hope or opinion than a really confident expression like “certainly”.

As well as the well-known fact that “a few” is positive and “few” has a negative connotation, this topic can often bring up the fact that “a few” is a number, while “few” can also be a proportion. For example, “a few people” is still just three or four people even if you’re talking about the whole world, whereas “few people in the world” could be millions as long as it is no more than a few percent of the world population.

How to present the language of hedging and generalising

In order to point out the problem of overgeneralising and to get students to use the good amount of language they already know which could be used to avoid it, I tend to use a TTT (Test Teach Test) approach with hedging and generalising. I give students ten statements which no one could agree with as they stand such as “Spanish people love partying till early in the morning every weekend”. Their first task is to work out what is wrong in general with all ten statements, something that can take a while! When they finally twig that they are all hideous oversimplifications, they try to use the language they already know such as “perhaps” and “usually” to modify the statements to make more defendable ones like “Many young Spanish people love partying till early in the morning most weekends”. They are given more expressions to use in that task, then compare their completed statements with another group. They could also compare their statements with texts on the topic, e.g. one from the internet on “How true are the stereotypes about...?” This task can be adapted for almost any topic, e.g. the business world, media, tourism, academic writing, or presentation tips.

This task can also be done starting with students’ own generalisations, for example by asking them to write ten statements about teenagers for homework then getting them to express doubt about their partners’ statements with expressions like “How common is that?” and “What makes you so sure/ unsure?” (see below for more possible questions).They can then brainstorm suitable words to respond to those questions/ challenges, and then put more useful expressions that you give them into the same categories. Suitable topics for eliciting stereotypes and other oversimplifications include talking about people of a particular age, regions, parts of a town, free time activities, families, gender, education, work, health, and people with certain jobs such as taxi drivers.

Another good way of eliciting suitable hedging/ generalising language is to give them good and bad tips on using this kind of language such as “Think carefully about how sure you are of your statement and use expressions to show that”, “Avoid very vague expressions like ‘some’ and ‘sometimes’”, “Add ‘It could be said’ to as many sentences in your academic writing as you can”, and “The aim of this kind of language is to obscure your real opinion”. After crossing off the bad tips, students brainstorm language to achieve the good things (plus maybe language to avoid the bad things, depending on whether that works with the tips that you gave them).

Another way of making brainstorming easier is to give them gaps into which they should brainstorm. For example, if you give them “It is usually true that… but _______________________” they can try to come up with “it depends on…”, “it can also be said that…”, “there are exceptions, such as…” and “it is occasionally the case that…”. Other suitable gaps include “According to _______many people/ the majority of experts/ the (leading) authority on this topic_____________,…”, “The majority of ____________ results/ papers/ people/ under 24s in this country_______________”, and simpler things like “a _______small/ considerable/ large______________ number of people…” You can then give them expressions to match to the gapped sentences. If the stems you gave them had useful language in them, an alternative next stage is to give them the lists of words which could go in the gaps and test their memory of the gapped sentences, for example seeing if they can remember “number of people” to add to “small/ considerable/ large”.

Another good active way to introduce the language is to get students to classify expressions by which question they are answering (“How likely?”) etc, then put them in order, like the lists above. This is obviously difficult to do with expressions they haven’t come across before, but can be made possible by adding expressions like “tiny minority” and “very probably” that are both easy to slot in themselves and make the meaning of their component parts (“minority” and “probably”) on their own clearer.

An easier way of getting them actively involved in the language right away is to make a collection of language that you could broadly divide into two categories, e.g. strong/ weak or sure/ unsure. Students listen to the phrases and hold up one of the two cards that they have been given depending on what meaning they think each one has, e.g. holding up the “Weak” card when they hear “It could be true in some cases”. They then look at the lists you were reading from and write the categories next to the phrases. To make guessing right from the first stage possible, you’ll need to make sure most of the expressions are guessable without having come across them before (“It could possibly be said”, “a tiny minority”, etc). You will probably also want to put the expressions in groups of three or four on the worksheet so that students can guess without having to know each and every example. After checking their answers, students can test each other in pairs in the same way as the in the first stage, and then help each other remember many phrases with the two meanings.

Classroom practice activities for hedging and generalising

Because of the TTT approach used, all the activities given above can also be used in the practice stage. For example, I’ve used the modifying overgeneralisations activity many times at this stage.

My recent favourite practice activity is giving students questions and other expressions to challenge each other’s statements with. Other useful phrases for that stage include:

How many/ How much/ How often

  • How common do you think that is?
  • I think it’s (slightly/ quite a lot/ a lot/ far) more/ less common than that.
  • Is it really so much/ many/ little/ few/ often/ seldom?

How sure

  • Can you really be so sure?
  • How sure are you about that?
  • I don’t think you can be so sure about that.
  • I think you can be more certain about that.
  • Is it really so likely/ unlikely?
  • What makes you so sure/ unsure?

More general ones

  • Before I can agree with that statement,…/ I could agree with that statement if you changed it to (something more like)…/ I’d change that statement (slightly) to say…
  • But isn’t it more like…?/ I’d say (it’s more like)…
  • Do you (really) mean (to say)…?
  • Do you have any figures on that?
  • I don't think it's (really) fair to say that …
  • I think that’s (a bit of) an overgeneralisation./ Isn’t that (a bit of) an oversimplification?
  • I think you’re exaggerating (somewhat).
  • I wouldn’t go so far (myself).
  • I’d go (even) further and say…
  • Isn’t that (a bit) out of date?
  • That’s (very) nearly true./ That’s almost true.
  • That’s a bit of a stereotype, isn’t it?
  • That’s not a very accurate statement.
  • That's a bit vague./ I think you could be (a bit) more precise.
  • What makes you think that(…)?/ What evidence is there (that…)?

Another game I’ve been using a lot recently is giving them a pack of cards of key words from hedging/ generalising phrases to use during such a debate, e.g. one student using the word “general” on one card to suggest that their partner’s statement should be “in general” rather than “sometimes”. If you make that pack of cards from just the words given in the list of collocations above, in the next stage they can then try to put those cards together to make longer expressions. You can also do a similar debate with cards that just define the kind of hedging/ generalising that they should do, e.g. cards saying “how often”, “how sure” and “apologise”. In all those games, the person who can use the most cards is the winner.

You can also give students hedging/ generalising phrases which they should try to use to write something they both agree with, e.g. getting them to make a statement that they both think there is “definitive proof” of. They might need suggested topics or even possible statements (such as something on manmade global warming for the “definitive proof” one), to help them.

A more fun activity is for students to make statements longer and longer, e.g. changing “Humans like music” to “Humans like music, although some more than others” and then “All humans fundamentally like music, although some like it more than others and some believe it to be evil”. To stop them drifting into the common problem of randomly throwing these kinds of expressions in, tell them to reject any longer sentences from their partner that they think are not actually true. Each successful extension and/ or the last person to successfully extend it gets points.

A way of focusing even more on the actual meanings of hedging/ generalising expressions like those in the lists above is to give students A and B different real figures on how likely, how many, how often etc things really are, e.g. “There is a 70% chance of a large earthquake in the Tokyo region in the next 70 years”. They change that into a generalised statement, e.g. “There will probably be a large earthquake in Tokyo in the next 70 years”. Their partner uses that clue to guess the precise figure behind it, with extra help from hints like “(The real figure is) much higher/ quite a lot lower”.

A similar activity is for students to make generalisations on topics they will later see statistics about, then decide which of their statements they think match the real statistics when they see them, with one point for each correct statement.

An obvious topic after generalisations is giving evidence, and both points can be practised with a bluffing game. One student gives a generalisation based on the topics or statements given (e.g. “Taking the underground is generally better for the environment than taking a bus”), and then their partner asks them for what evidence they have to support that generalisation. Perhaps using a list of suggested kinds of evidence to help, they then give true or made up personal experience, statistics, quotes, etc. Their partner then guesses if that evidence is true or not.

For freer practice, the topics that I suggested above for eliciting overgeneralisations like health can also of course be used to practise good generalising and hedging. Other things you can ask them to come up and discuss with include:

  • A justified generalisation
  • An unjustified generalisation
  • A common generalisation
  • A generalisation that might have been true but no longer is
  • An overgeneralisation
  • A generalisation that your parent(s)/ colleague(s)/ teacher(s)/ neighbour(s)/ classmate(s) make(s)
  • A generalisation that is often made by politicians/ old people/ middle-aged people/ rich people/ young people
  • A generalisation that you have recently heard/ read/ read about/ heard about/ challenged/ reconsidered/ thought of/ said
  • A generalisation about the past/ present/ future

Copyright © 2014

Written by Alex Case for

Enjoyed this article?

Please help us spread the word: