logo ELT Concourse teacher training
Concourse 2

Morphology: the building blocks of language



What is morphology?

It is better to start with a definition of its subject matter.
A morpheme is

the smallest meaningful unit of a language.

and the word meaningful is emphasised for a reason.
In this phrase, we have these 10 morphemes:

the + small + est + mean + ing + ful + unit + of + a + language.

The term meaningful here does not refer to meaning standing alone (although it may) but to meaning in context or in combination with other words.
For example, a function word such as it carries no meaning without a context and co-text but in a sentence such as:
    I read the book and loved it
the word book clearly carries meaning.
And in
    John and Mary came to the show
the word and carries a meaning to the hearer / reader relating to combining subjects although the word standing entirely alone has no discernible meaning.
Both it and and are morphemes by this definition and so are I, read, the, book, love, -d, John, Mary, came, to and show.

Morphology, then, is the study of how languages form words.  All languages do this and they do it in a bewildering and fascinating number of ways.

We are concerned here with how English functions in this respect but effective teaching in this area requires at least an outline knowledge of how your learners' language(s) function.  For some information about that, see the guide to teaching word formation, linked below.  Other languages employ morphemes in a variety of ways using, e.g.
    infixes: a rare event in English exemplified by spoonsful or absobloodylutely.
    circumfixes: unknown in English but used extensively in some languages which add affixes in pairs to the beginning and end of a lexeme to derive a single change in meaning or word class.

The guide to word formation, linked below, covers similar ground but extends it into other areas.  This guide is concerned with understanding the theoretical bases for how words are formed in English.


Identifying words

Before we can sensibly look at what words are made of, we need to be clear about what a word actually is.  This may seem a very simple question but it is actually rather difficult to define what a word is in any language.
Here's a definition from Google:

a single distinct meaningful element of speech or writing, used with others (or sometimes alone) to form a sentence and typically shown with a space on either side when written or printed

As we shall see, that is not an adequate definition for our purposes.  The definition is hedged with typically and sometimes and that is a sign that the definition is difficult.

think Task 1: Here are some traditional tests for words in a language.
Click on the table when you have filled in the right-hand column.

(Based on Crystal, 1987:91)

Of course, one ordinary way to identify a word is to look it up in a dictionary.  Unfortunately, that won't work, either.  If you go to an online dictionary for something like inderivability, you are very unlikely to find it but most native speakers of English would be happy to accept that it really is a word.
The other issue with dictionaries is that, in order to save space, they usually list words by the root or lemma.  For example, you will find repeat as a headword or lemma in a dictionary but repeated and repeating will normally not be listed separately unless the words are used with a different significance.  That doesn't mean they aren't words.

So what actually is a word?

For a morphologist, a word is any item which can be derived from the application of morphological rules.  This means that even something like computerhood, which you will certainly not find in a dictionary, is a word because it can be derived from the three morphemes that make it up, following normal conventions in English.  These are:
the verb
    + r to make the noun doer of the verb (or, with words that do not already end in 'e', 'er')
    + hood to make the noun signifying state of being a doer of the verb.


Identifying morphemes

think Task 2: Can you identify the morphemes in this list of lexemes?
Click on the eye open to reveal some comments when you have.

eye open
This is an easy one.  The lexeme is a single morpheme which stands alone.  We can add morphemes to it, of course, to make, e.g., sugary, sugariness etc.
It is a free morpheme which needs no other morpheme to make sense.
eye open
This is a compound formed from two free morphemes, class and room.  Both are nouns and they combine to make a new noun with a new but transparent meaning.  See the guide to compounding, linked below, for more.
eye open
This is an example of a free morpheme combining with a bound morpheme to make a new word class.
We have taken the free morpheme,
ugly, and combined it with a bound morpheme suffix, ness, to make the noun from the adjective.  The suffix is a bound morpheme which cannot stand alone and carry meaning.  Because the morpheme is used to derive a new word, it is called a derivational morpheme.
(Do not worry about the spelling change.  That is simply one of the orthographic conventions in English and makes no difference to meaning.  Do not be concerned, either, with the fact that
ness can be a free morpheme meaning a headland.  That is an example of homonymy.)
eye open
This is an example of the same thing.  A free morpheme, teach, combining with a bound morpheme, er, to make a personal noun from the verbIn other words, to derive a new word class.  There is a second bound morpheme, s, which forms the plural in English, part of how nouns decline.
eye open
This, too, is an example of a free morpheme combining with a bound morpheme (fly and ing, respectively) but in this case we have a grammatical change, converting the base form of the verb into a participle or a gerund, depending on context.  This is an inflexional morpheme, determining how verbs conjugate.
eye open
This is slightly problematic.  Often, the bound morpheme prefix re implies the act of doing something again, as in rewrite, reinforce etc.  Here, however, analysing it that way would result in the morpheme peat which, as a verb, carries no meaning so cannot, by definition, be a morpheme in English.  It is, however, acceptable to analyse it that way because the verb derives from the Latin, via French, from re, [again] + petere [strive after; ask for].  It occurs again in, e.g., petition.
We have another inflexional morpheme here,
ed, which forms the regular past tense in English.
eye open
This, of course, can mean a person from Hamburg and that is an example of a free morpheme, called the root, combining with a bound morpheme to form the name for a person's origin.  Compare Londoner, Parisian etc.
In this case, however, the word probably derives from the name of a beef product from Hamburg and the morpheme
burger has taken on a life of its own as in cheeseburger, beefburger and veggieburger.
eye open
The function of the bound morpheme, dis, as a prefix is clear; it means the negative.  However, hidden in here is another Latin-derived meaning of re.  The word derives from re [back] + specere [look].
eye open
This is an example of how new words are formed from the general rules of a language's morphology.  We simply take the name or philosophy and add ite or ist.  Compare Trotskyite, Marxist, communist, sexist, Thatcherite, Benthamite etc.
eye open
This is a less common way to form new words called a blend.  Two words have been fused to form a third meaning.  Compare smog [from the free morphemes smoke + fog] or motel [from motor + hotel].
eye open
You can be excused for thinking that this is a noun formed from the verb donate by dropping the 'e' and adding the -tion bound morpheme in the conventional way.  However, in fact the noun entered the language much earlier than the verb which was formed later by analogy with more usually formed verb-noun pairs such as inflate-inflation and many more.
This is called a back formation and there is a short list in the guide to word formation of other back forms in English
eye open
This is interesting because the bound morpheme, in, clearly carries a negative meaning.  Unfortunately, the second morpheme, ept, carries no meaning.  The word comes from Latin in- [not] + aptus [apt].
The term for this kind of morpheme is a bound base (sometimes a bound root).  It cannot exist alone but does not act as a prefix or suffix.  Other examples are the ver in verity, the doct in doctor and the dext in dexterity.
Many verbs are formed in this way, derived directly from Latin, Old French or Old English with a bound base which has no independent existence in the modern language.  Examples include
desiccate, modify, indemnify, enlighten and more.
eye open
These are examples of combining forms.  The suffix, -icide does not alter the meaning of the word pest or its word case (as is the case with, e.g., pester).  The suffix instead adds a new layer of meaning to the word and denotes killing agent.
The second example is similar but the combining form (
bio-) is a prefix (the more common role of combining forms).  Again, it does not alter the meaning of chemistry but it does add a new layer of meaning to it.
In these cases, the words
pest and chemistry are free morphemes but in many cases of the use of combining forms in English, the base is not a free morpheme.  Examples are: democracy, ferroconcrete, astrology.
The roots here,
dem, ferro and astr do not stand alone and are describable as bound bases which can be traced to the Latin or Greek roots.
Combining differs from compounding in that the outcome is not a third meaning, it is an additional meaning grafted on to the base form.

All the technical terms are in bold in this table.
Click here to take a test to see if you can remember what they mean.


Derivational morphemes

As we saw above, there are two ways that derivational morphemes added to lexemes can change them.

  1. We can change the word class or category of a word but leave the base meaning unchanged.  For example,
    • entertain (verb) → entertainment (noun)
    • entertain (verb) → entertaining (adjective)
    • elephant (noun) → elephantine (adjective)
  2. We can add morphemes which change the meaning but leave the word class unchanged.  For example,
    1. pleasantunpleasant (adjectives, opposites)
    2. doundo (verbs, reverse action)

Sometimes, we can add morphemes which change both the meaning and the word class.  For example,


Productivity of derivational morphemes

A central concern of morphology is to investigate how productive a derivational morpheme actually is.  This does not apply to inflexional morphemes because their role is generally fixed grammatically
How is it, for example, that the morpheme th can be attached to some adjectives to form the noun (warm → warmth, wide → width etc.) but is almost completely unproductive in forming new words in English?
We would not, for example, take the adjective windy and form a noun such as windith, preferring instead to choose the ness suffix and making windiness.
There are some interesting factors in play.



Some morphemes have an easy-to-see relationship between the form and meaning.  Here are some examples:

  1. -ible and -able
    It is easy to unpack a word like recordable by considering the suffix able and realising that it always produces an adjective from a transitive base verb and that it exists as a free morpheme meaning can.  You can quite easily make new adjectives like this from any number of verbs, providing they take an object.  The fact is that it is the only way English can form a new adjective from a verb.  We have, of course, the ible morpheme in words like flexible, audible, comprehensible and so on but if you try to form an adjective from unusual or phrasal verbs you will almost always opt for able.  Try it with:
        prune, devolve, explode, derange, disturb, put off, turn down
    The morpheme ible is, therefore, unproductive.
    There are still many extant adjectives formed from verb which employ the -ible suffix.  However, removing the suffix does not usually leave a recognisable word so, for example:

    Here, again, we have instances of a bound base or bound root which is no longer able to function alone.  The words aud, ris and corr simply do not exist although they can be traced to Latin.
    In nearly all cases, the -ible forms are more formal and less common so we have formal-informal pairings such as:
        credible - believable
        edible - eatable
        potable - drinkable
        risible - laughable
        illegible - unreadable
        comprehensible - understandable

  2. nouns from adjectives
    Another example is the ways in English that nouns are formed from adjectives.  The two most common are:
        ness: happy → happiness, sad → sadness, great → greatness etc.
        ity: insane → insanity, absurd → absurdity, acid → acidity etc.
    However, if you try to form new nouns from these adjectives:
        blue-green, airy, snowy, freezing, wet
    You will almost always opt for ness, indicating that ness is more productive than ity in this function.
    This is not to say that ity is unproductive in the way that th is, but that it is less productive.  It can be very productive with adjectives ending in the able morpheme.  Try, e.g.,
        openable, disturbable, derivable, pickupable
    and -ability seems to be the form of choice although it is certainly possible to form words like openableness.
  3. doers
    Other derivational morphemes are also less productive than others.  We can, e.g., form the doer of a verb by attaching the morpheme ant as in
        inhabit → inhabitant
        assail → assailant
        claim → claimant

    but this suffix is far less frequently used for this function than the ubiquitous er suffix.  Try forming the doer of the action from these verbs and you'll see how much more productive the er morpheme is:
        disturb, peruse, demand, accommodate
    The suffix ist is even less productive in this sense although typist and telephonist exist.
    Just as we saw for the able/ible distinction, the suffix or is now almost completely unproductive and is frozen into words like actor, transgressor and doctor.
  4. Nonce words
    Speakers of English frequently make up words by the addition of a derivational suffix or a small range of prefixes.  We get, therefore words such as:


    and so on, none of which will appear in a dictionary and all of which are temporary (nonce) introductions.  Nonce words are sometimes referred to as occasionalisms.
  5. Barbarisms
    This term is used by language purists to describe formations which are disparaged because acceptable and derivationally purer forms already exist.  Some will not become popular but many so-called barbarisms have already entered the language.  For example:
        orientate is a verb formation already covered by the verb orient
        preventative is an adjective whose meaning is not distinguishable from preventive
        untactful is an unnecessary addition to a language which already contains tactless
        educationalist is not clearly distinguishable from the current educationist
        metrification does not add anything useful not covered by metrication.  In fact, it adds unnecessary suffixes because there is no verb metrify or metrificate but metricate certainly exists.
    and so on.
    In some cases, these questionable coinages will become the usual forms, in others, they will die out.
    As has been pointed out:
    Ultimately, it is general usage, rather than etymological pedigree, that determines the survival of a word. (Todd & Hancock 1986:71).  (Although they may, of course, mean use rather than usage.)
  6. Combining forms
    When an additional meaning is grafted on to a term by the addition of a suffix it does not usually change the word class so cannot be derivational in this sense.  The guide to word formation contains more on this area and there is a list of such forms linked below.  Words such as

    are examples of the use of combining form suffixes which do not change the word class (even if they are appended to recognisable words at all, which many are not).
    Combining forms are, especially in academic and scientific writing, very productive indeed so recent formulations, such as anthropocene (the current geological age) are coined at will and retain their existence for many years.

For a much more complete list of the semantic functions of suffixes, see the guide to word formation, linked below.


Frequency and usefulness

Some suffixes are simply too constrained in the number of bases they can be attached to to make them very productive.
We can make an adverb from some nouns with the suffixes wise and wards as in, e.g.,
    crabwise, clockwise, northwards, citywards etc.
but the number of possibilities is very limited because of the infrequency of the resulting adverbs and the small usefulness the concepts have.

However, the adverb forming ly is hugely productive in its ability to produce adverbs from adjectives because the results are both frequent and useful.  It can also attach itself to the barely limited number of participle adjectives in English.  So we can form, for example:
    stunningly, interestingly, swingingly, boringly, understandingly, reassuringly
and so on.  We can even make up new adverbs on the spur of the moment and be understood, for example:
    She spoke praisingly.


Other constraints

some morphemes behave along phonological lines.  For example:
The verb forming ize/ise usually attaches to multisyllabic nouns and adjectives if the stress is not on the final syllable.  So we get, e.g.:
    random → randomize
    real → realize
    apology → apologize
Another verb-making suffix, en, attaches itself to single syllable adjectives but only if they end in certain phonemes (stops, and fricatives):
    deep → deepen
    light → lighten
    broad → broaden
    deaf → deafen
but not
    clear → *clearen
    high → *highen
in fact, we insert a 't' to conform with the phonological rule and get heighten.
This applies to prefixes, too, as we see with the choice often of im rather than in before /m/ and /p/ and ir before /r/.
The root of a word will often determine what derivational morpheme is possible.
For example, the adjective-forming suffix ic will not attach to Anglo-Saxon bases but en will so we get:
    vitriolic, metallic, dramatic
    woollen, leaden, ashen
The prefixes un, im, in etc. do not attach to words with negative connotations so we can't have:
    *unvile, *impernicious, *undoleful, *unhelpless
etc. but we can have
    unlovely, unappreciated, improper, unhelpful

Derivational morphemes are dealt with in greater detail in the guide to word formation, linked below.


Teaching implications of productivity and other constraints

think Task 3: Review the sections above on aspects of productivity and the constraints on word formation with derivational morphemes and consider for a moment what the implications are for teaching English.  Consider, too, production vs. comprehension.
Then click here.


Inflexional morphemes

In English we can denote a number of grammatical constructions by using inflexional morphemes.  We usually do this by changing the ending but, irregularly, we can also change the internal characteristic of the lexeme (a process called mutation).  Inflexional morphemes always, in English, follow derivational morphemes.  We get, therefore, e.g.:
Inflexional morphemes include:

  1. Tense:
    We saw above that the -ed/-d morpheme signals past tense but it also signals a past participle for regular verbs:
    and for some irregular verbs the past participle is also marked by a morpheme addition
    Internal changes (mutations) are also morphemic but what usually occurs is that one bound morpheme is altered and becomes a slightly different bound morpheme as in, e.g.
        speak → spoke
        hold → held

    English is unusual among European languages in having only one past-tense form to apply to all persons.  Most other languages in this family use different endings for most of the persons and number.
    It is also unusual in often having the same form for the past tense and the past participle.  Other languages, such as Greek, distinguish the forms.
  2. Aspect:
    English signals progressive and other aspects by changes to the verb:
        go, going, gone
    Again, English is very limited in this respect having only one form of the past participle and only one of the -ing form.  The -ing form, incidentally, has no irregularities.  None.
  3. Person:
    Apart from the truly irregular verb be, English only signals person with the third-person -s morpheme (and then only in the singular):
    By European language standards, this is a very limited range.  In some other European languages, for example, the translation works like this:
    English French German Polish Estonian Italian
    I go
    you go
    she goes
    we go
    they go
    je vais
    tu vas
    elle va
    nous allons
    ils vont
    ich gehe
    du gehst / Sie gehen
    sie geht
    wir gehen
    sie gehen
    ty idź
    ona idzie
    ma lähen
    sa lähed
    ta läheb
    me läheme
    nad lähevad
    io vado
    tu vai
    lei va
    This is not universal and, for example, all three Scandinavian languages (Norwegian, Swedish and Danish) have the same form for all the verbs in all persons (går).
    For this reason, many languages, such as Greek and Italian, routinely drop the pronoun because the person is signalled by the verb ending.  The term for these languages is pro-drop incidentally, and there is more on that in the guide to types of languages, linked below.
  4. Number:
    Although there are some irregularities, English uses the s/es morpheme to signal the plural.
    Almost all nouns function this way in English which is a simple system in comparison to many other languages (and more complex than many which do not signal the difference between singular and plural at all).
  5. Adjectives:
    English uses morphemic addition to change some adjectives from absolute to comparative or superlative:
    The morpheme most can also act as a suffix carrying a similar meaning of towards the extreme in words like nethermost, uppermost, hindmost etc.  The morpheme more can't do that and the suffix most is now unproductive.
    The added complication in English is that some adjectives cannot take the inflexion at all so we do not allow, e.g., boredest or beautifuller.  The rules for when we inflect and when we use the periphrastic form are quite complex, in fact.
    Other languages are often simpler in this respect.  Most Romance languages, French, Spanish, Romanian, Portuguese etc. employ the periphrastic form and some, such as German, usually employ only the inflected form (although the periphrastic form is available), others, Scandinavian languages and some Slavic languages, for example, work a little like English.
    Adjectives in English remain unchanged, whatever case, gender and number they are associated with.  Other languages will often inflect the adjective to show case, gender and number.
  6. Case:
    Many languages, which take advantage of morphemes to signal case (nominative, accusative, dative, genitive, locative, ergative etc.) alter the nouns and other elements to show this.  English generally does not do this with lexical words but has a slightly complex pronoun and determiner system which sometimes radically alters a morpheme to show case and sometimes simply adjusts it slightly:
    Nominative (subject) Accusative (object) Genitive (possessive) Genitive pronoun
    and so on.  A full list is available in the guide to personal pronouns.
    English also deploys the 's morpheme for the genitive:
        John's, the dog's, the ship's, the government's
    The fact that English has two genitive structures (the government's policy vs. the policy of the government, for example) is an added complication which leads to a good deal of unnatural language in learners and some error.

The technical term for the way in which words can have their grammatical function altered by changes to the morphology is accidence, incidentally.

Old English had many more (and much more complex) inflexions than Modern English has retained.  The loss of inflexion is one of the most important changes to have occurred in the language.  (As a matter of simple interest, Old English was made even more complicated by having a separate category gor two people.  So we had ic (I), we (we) and wit (we two) for example and these changed in the accusative, dative and genitive, rather like modern German pronouns alter.)
Many other languages, such as German, Polish, Finnish, and French deploy a much wider range of suffixes to denote number, person, case and gender.  Other languages, such as the Chinese languages are even more limited than English in their use of inflexions and suffixation.


(n. /səˈpliːʃᵊn/)

Occasionally, we come across a form which is clearly not derived from what we would expect.  For example, the words larger and largest are clearly connected to and derived from the base form large but that is not the case with the words worse and worst acting as the comparative and superlative forms of the word bad.
What we are dealing with is suppletion and the result is called a suppletive form.  It refers to the fact that the forms are phonemically and (sometimes) etymologically distinct.  The origin of the word bad is somewhat obscure but the words worse and worst are traceable to Old English forms.
Suppletion may be complete as in the example of bad-worse in which the words share no letters or sounds at all or it may be quite weak as in the case of five-fifth where the words are derived from the same root but the pronunciation has been altered so that the spoken forms may not be recognised as connected at all.
Other common examples of suppletion in English are:

go, went gone
The past tense of this verb is clearly unconnected from the base form and the past participle.  The latter two come from the Old English gan and there are cognates in most Germanic languages, gehen in German, gaan in Dutch and so on.
The past tense of the verb, however, is derived from the irregular past tense of the verb wend and that verb has now taken on the regular past tense forms.
three, third
This is a weaker example because some letters and sounds are shared but hearing the words will not obviously alert people to their connection.
one, first, two, second
are both examples of much stronger suppletion because the word first derives from a different root from the word one (although both are traceable to Old English) and the word second is an import via French from Latin and not connected with the numeral two at all.
far, further, farther, furthest, farthest
are all examples of quite weak suppletion because they all derive from the same Old English root.
be, am, is, are, was, were
is an example of really complex suppletion.  The roots are traceable but all come from a different set of words in Old English and Proto-Indo European.  It is a tangle of semantically connected forms with no discernible spelling and pronunciation commonalities.  The word be derives from the Old English beon (exist), the word am from eom, the word are from the plural of beon and so on.  The past tenses are traceable to the the verb wesan (meaning remain).  The verb has eight distinct parts in Modern English derived from at least two Old English dialects: be, am, is, are, was, were, being, been.
good, better, best
The first of these is traceable to Old English gōd but the two other forms come from a different source (Old English beste).
person, people
The first of these derives from the Old French persone and the latter from Old French peupel (itself from Latin populus).  The plural persons is rare and formal.
cow, cattle
derive from separate roots but the latter is considered a plural collective version of the former.  Other forms such as ox-oxen are referred to as suppletive affixes because they do not follow the normal morphological rules for making plurals.
wreak, wrought
This is a real oddity because wrought was originally the past tense of work (which is now regular in English).  It still exists as a participle adjective in, e.g., wrought iron.  The past forms of wreak are in principle regular (wreaked, wreaked) but it has become common to see wrought used as the past of that verb, perhaps by analogy to teach and seek.



For a much more detailed look at prefixes and what they do, see the guide to word formation.  Here, we will simply list some of the categories into which prefixes fall and the sorts of things the morphemes do.  Nearly all prefixes are bound morphemes.
Free morphemes which may look like prefixes are usually cases of compounding so, for example, the morpheme house in housemaster is not considered here as an example of prefixing.
See the guide to compounding for more on that, linked below.
Prefixes come in three main sorts of which the second is the most productive by far:

  1. derivational prefixes
    are quite rare in English and almost wholly unproductive.  In earlier forms of English they were very productive and many modern words such as become, aforesaid, embolden etc. can be traced to earlier prefixation.
    They include, e.g.,
    1. en- or em- forming verbs as in:
    2. a- forming adjectives such as
    3. be- forming verbs such as
  2. meaning-converting prefixes
    which alter the meaning, but not word class of many words.
    A full(er) list is available from the link below and in the guide to word formation.  The list includes terms denoting:
    1. attitude: pseudoscience, dysfunctional etc.
    2. negative senses: unbalanced, disbelieve, irresponsible etc.
    3. number: monoglot, biannual etc.
    4. reversal: undo, decouple etc.
    5. location: submarine, superstructure etc.
    6. temporal: prewar, proto-language etc.
    7. degree or size: supermarket, overdone etc.
  3. combining forms
    which add a layer of meaning without change the meaning of the base.  These include many scientific forms such as
    Some words consist purely of combining forms with two bound morphemes.  Examples include

Related guides
semantics for more on meaning
types of languages for a guide to how other languages do things differently
prefixes and suffixes a PDF-formatted list of prefixes and suffixes in English
combining forms a PDF formatted list of the most common combining forms in English
compounding for how this type of morpheme manipulation works
word formation for more on how morphemes combine in English with links to PDF lists of affixes and consideration of combining forms
teaching word formation for a guide containing some consideration of how major language groups handle affixation and compounding

There's rather a lot in this area which is a much-researched field and much of it is not relevant to teaching.
Crystal, D, 1987, The Cambridge Encyclopedia of Language, Cambridge: Cambridge University Press
Lieber, R, 2009, Introducing Morphology, Cambridge: Cambridge University Press
Todd, L & Hancock, I, 1986, International English Usage, Beckenham: Croom Helm