121
5 Morphology and Word Formation
key concepts
Words and morphemes
Root, derivational, inectional morphemes
Morphemes, allomorphs, morphs
Words
English inectional morphology
English derivational morphology
Compounding
Other sources of words
Registers and words
Internal structure of complex words
Classifying words by their morphology
introduction
is chapter is about words—their relationships, their constituent parts,
and their internal organization. We believe that this information will be of
value to anyone interested in words, for whatever reason; to anyone inter-
ested in dictionaries and how they represent the aspects of words we deal
with here; to anyone involved in developing the vocabularies of native and
non-native speakers of English; to anyone teaching writing across the curric-
ulum who must teach the characteristics of words specic to their discipline;
to anyone teaching writing who must deal with the usage issues created by
the fact that dierent communities of English speakers use dierent word
forms, only one of which may be regarded as standard.
Exercise
1. Divide each of the following words into their smallest meaningful
parts:landholder, smoke-jumper, demagnetizability.
2. Each of the following sentences contains an error made by a non-
native speaker of English. In each, identify and correct the incorrect
word.
a. I am very relax here.
b. I am very boring with this game.
c. I am very satisfactory with my life.
d. Some owers are very attracting to some insects.
e. Many people have very strong believes.
Delahunty and Garvey
122
f. My culture is very dierence from yours.
g. His grades proof that he is a hard worker.
h. The T-shirt that China drawing. (from a T-shirt package from
China)
In general terms, briey discuss what English language learners must
learn in order to avoid such errors.
3. Some native speakers of English use forms such as seen instead
of saw, come instead of came, aks instead of ask, clumb instead of
climbed, drug instead of dragged, growed instead of grew. Are these
errors? If they are, are they the same kinds of errors made by the non-
native speakers of English listed in Exercise 2? If not, what are they?
words and morphemes
In traditional grammar, words are the basic units of analysis. Grammarians
classify words according to their parts of speech and identify and list the
forms that words can show up in. Although the matter is really very com-
plex, for the sake of simplicity we will begin with the assumption that we are
all generally able to distinguish words from other linguistic units. It will be
sucient for our initial purposes if we assume that words are the main units
used for entries in dictionaries. In a later section, we will briey describe
some of their distinctive characteristics.
Words are potentially complex units, composed of even more basic units,
called morphemes. A morpheme is the smallest part of a word that has
grammatical function or meaning (NB not the smallest unit of meaning);
we will designate them in braces—{ }. For example, sawed, sawn, sawing,
and saws can all be analyzed into the morphemes {saw} + {-ed}, {-n}, {-ing},
and {-s}, respectively. None of these last four can be further divided into
meaningful units and each occurs in many other words, such as looked,
mown, coughing, bakes.
{Saw} can occur on its own as a word; it does not have to be attached
to another morpheme. It is a free morpheme. However, none of the other
morphemes listed just above is free. Each must be axed (attached) to some
other unit; each can only occur as a part of a word. Morphemes that must
be attached as word parts are said to be bound.
Exercise
1. Identify the free morphemes in the following words:
123
Morphology and Word Formation
kissed, freedom, stronger, follow, awe, goodness, talkative, teacher,
actor.
2. Use the words above (and any other words that you think are rel-
evant) to answer the following questions:
a. Can a morpheme be represented by a single phoneme? Give ex-
amples. By more than one phoneme? Give examples.
b. Can a free morpheme be more than one syllable in length? Give
examples. Can a bound morpheme? Give examples.
c. Does the same letter or phoneme—or sequence of letters or pho-
nemes—always represent the same morpheme? Why or why not?
(Hint: you must refer to the denition of morpheme to be able to
answer this.)
d. Can the same morpheme be spelled dierently? Give examples.
e. Can dierent morphemes be pronounced identically? Give examples.
f. A morpheme is basically the same as:
i. a letter
ii. a sound
iii. a group of sounds
iv. none of the above
3. The words district and discipline show that the sequence of letters
d-i-s does not always constitute a morpheme. (Analogous examples are
mission, missile, begin, and retrot.) List ve more sequences of let-
ters that are sometimes a morpheme and sometimes not.
4. Just for fun, nd some other pairs like disgruntled / *gruntled and
disgusted / *gusted, where one member of the pair is an actual English
word and the other should be a word, but isn’t.
Axes are classied according to whether they are attached before or
after the form to which they are added. Prexes are attached before and
suxes after. e bound morphemes listed earlier are all suxes; the {re-}
of resaw is a prex. Further examples of prexes and suxes are presented in
Appendix A at the end of this chapter.
Root, derivational, and inectional morphemes
Besides being bound or free, morphemes can also be classied as root, deri-
vational, or inectional. A root morpheme is the basic form to which other
Delahunty and Garvey
124
morphemes are attached. It provides the basic meaning of the word.e
morpheme {saw} is the root of sawers. Derivational morphemes are added
to forms to create separate words: {-er} is a derivational sux whose ad-
dition turns a verb into a noun, usually meaning the person or thing that
performs the action denoted by the verb. For example, {paint}+{-er} creates
painter, one of whose meanings is “someone who paints.
Inectional morphemes do not create separate words. ey merely
modify the word in which they occur in order to indicate grammatical prop-
erties such as plurality, as the {-s} of magazines does, or past tense, as the {ed}
of babecued does. English has eight inectional morphemes, which we will
describe below.
We can regard the root of a word as the morpheme left over when all
the derivational and inectional morphemes have been removed. For example,
in immovability, {im-}, {-abil}, and {-ity} are all derivational morphemes, and
when we remove them we are left with {move}, which cannot be further di-
vided into meaningful pieces, and so must be the word’s root.
We must distinguish between a word’s root and the forms to which af-
xes are attached. In moveable, {-able} is attached to {move}, which weve
determined is the words root. However, {im-} is attached to moveable, not
to {move} (there is no word immove), but moveable is not a root. Expressions
to which axes are attached are called bases. While roots may be bases,
bases are not always roots.
Exercise
1. Can an English word have more than one prex? Give examples. More
than one sux? For example? More than one of each? Give examples.
Divide the examples you collected into their root, derivational, and
inectional morphemes.
2. Check your dictionary to see how it deals with inected and derived
word forms. Does it list all the inections of regular inected words?
Just irregular ones? Does it accord derived forms their own entries or
include them in the entries of the forms from which they are derived?
3. Does your dictionary list bound morphemes? Which kinds?
morph e m e s , a l lo m o r p h s , a n d m o r p h s
e English plural morpheme {-s} can be expressed by three dierent but
125
Morphology and Word Formation
clearly related phonemic forms /@z/ or /z/, /z/, and /s/. ese three have
in common not only their meaning, but also the fact that each contains an
alveolar fricative phoneme, either /s/ or /z/. e three forms are in comple-
mentary distribution, because each occurs where the others cannot, and it is
possible to predict just where each occurs: /Iz/ after sibilants (/s, z, S, Z, tS,
dZ/), /z/ after voiced segments, and /s/ everywhere else. Given the semantic
and phonological similarities between the three forms and the fact that they
are in complementary distribution, it is reasonable to view them as contex-
tual pronunciation variants of a single entity. In parallel with phonology,
we will refer to the entity of which the three are variant representations as a
morpheme, and the variant forms of a given morpheme as its allomorphs.
When we wish to refer to a minimal grammatical form merely as a form,
we will use the term morph. Compare these terms and the concepts behind
them with phoneme, allophone, and phone. (Hint: note the use of / /, [ ],
and { }.)
Exercise
Consult the glossary in the chapter on Phonetics and Phonology and
try to determine the meanings of the morphemes {phone}, {allo-}, and
{-eme}.
(1) /phoneme/
[allophone] [allophone] [allophone] etc.
(2) {morpheme}
/allomorph/ /allomorph/ /allomorph/ etc.
words
Words are notoriously dicult entities to dene, both in universal and in
language specic terms. Like most linguistic entities, they look in two direc-
tions—upward toward larger units of which they are parts (toward phrases),
and downward toward their constituent morphemes. is, however, only
helps us understand words if we already understand how they are combined
into larger units or divided into smaller ones, so we will briey discuss sev-
Delahunty and Garvey
126
eral other criteria that have been proposed for identifying them.
One possible criterion is spelling: in written English text, we tend to
regard as a word any expression that has no spaces within it and is separated
by spaces from other expressions. While this is a very useful criterion, it
does sometimes lead to inconsistent and unsatisfactory results. For instance,
cannot is spelled as one word but might not as two; compounds (words com-
posed of two or more words; see below) are inconsistently divided (cf. inux,
in-laws, goose esh, low income vs. low-income).
Words tend to resist interruption; we cannot freely insert pieces into
words as we do into sentences. For example, we cannot separate the root of
a word from its inectional ending by inserting another word, as in *sock-
blue-s for blue socks. Sentences, in contrast, can be interrupted. We can in-
sert adverbials between subjects and predicates: John quickly erased his n-
gerprints. By denition, we can also insert the traditional interjections: We
will, I believe, have rain later today.
In English, though by no means in all languages, the order of elements
in words is quite xed. English inections, for example, are suxes and
are added after any derivational morphemes in a word. At higher levels in
the language, dierent orders of elements can dier in meaning: compare
John kissed Mary with Mary kissed John. But we do not contrast words with
prexed inections with words with suxed inections. English does not
contrast, for example, piece + s with s + piece.
In English, too, it is specic individual words that select for certain in-
ections. us the word child is pluralized by adding {-ren}, ox by adding
{-en}. So if a form takes the {-en} plural, it must be a word.
So words are units composed of one or more morphemes; they are also
the units of which phrases are composed.
English inectional morphology
Inectional morphemes, as we noted earlier, alter the form of a word in or-
der to indicate certain grammatical properties. English has only eight inec-
tional morphemes, listed in Table 1, along with the properties they indicate.
Except for {-en}, the forms we list in Table 1 are the regular English in-
ections. ey are regular because they are the inections added to the vast
majority of verbs, nouns, adjectives, and adverbs to indicate grammatical
properties such as tense, number, and degree.
ey are also the inections we typically add to new words coming into
the language, for example, we add {-s} to the noun throughput to make it
plural. When we borrow words from other languages, in most cases we add
the regular English inections to them rather than borrow the inections
127
Morphology and Word Formation
they had in their home languages; for example, we pluralize operetta as oper-
ettas rather than as operette as Italian does; similarly, we sing oratorios rather
than oratori. [anks to Paula Malpezzi-Price for help with these examples.]
e regular inections are the default inections that learners tend to use
when they dont know the correct ones (for example, growed rather than
grew).
nouns: {-s} plural (the birds)
noun phrases: {-s} genitive/possessive (the bird’s song)
adjectives/adverbs: {-er} comparative (faster)
{-est} superlative (fastest)
verbs: {-s} 3rd person singular present tense (proves)
{-ed} past tense (proved)
{-ing} progressive/present participle (is proving)
{-en} past participle (has proven)
(was proven)
 :     
[Note: the regular past participle morpheme is {-ed}, identical to the
past tense form {-ed}. We use the irregular past participle form {-en} to
distinguish the two.]
However, because of its long and complex history, English (like all lan-
guages) has many irregular forms, which may be irregular in a variety of
ways. First, irregular words may use dierent inections than regular ones:
for example, the modern past participle inection of a regular verb is {-ed},
but the past participle of freeze is frozen and the past participle of break
is broken. Second, irregular forms may involve internal vowel changes, as
in man/men, woman/women, grow/grew, ring/rang/rung. ird, some forms
derive from historically unrelated forms: went, the past tense of go, histori-
cally was the past tense of a dierent verb, wend. is sort of realignment
is known as suppletion. Other examples of suppletion include good, better,
and best, and bad, worse, and worst. (As an exercise, you might look up be,
am, and is in a dictionary that provides etymological information, such as
the American Heritage.) Fourth, some words show no inectional change:
sheep is both singular and plural; hit is both present and past tense, as well
as past participle. Fifth, many borrowed words, especially nouns, have ir-
regular inected forms: alumnae and cherubim are the plurals of alumna and
Delahunty and Garvey
128
cherub, respectively.
Irregular forms demonstrate the abstract status of morphemes. us the
word men realizes (represents, makes real) the two morphemes {man} and
{plural}; women realizes {woman} and {plural}; went realizes {go} and {past
tense}. Most grammar and writing textbooks contain long lists of these ex-
ceptions.
As a nal issue here we must note that dierent groups of English speak-
ers use dierent inected forms of words, especially of verbs. When this is
the case, the standard variety of the language typically selects one and rejects
the others as non-standard, or, illogically, as “not English,” or worse. For
example, many English speakers use a single form of be in the past tense
(was) regardless of what the subject of its clause is. So they will say, We was
there yesterday. is is an uncontroversial issue: was in instances like this is
universally regarded as non-standard. Other forms are more controversial.
For example, what is the past tense of divedived or dove? How are lie and
lay to be used? How does your dictionary deal with such usage issues?
Exercise
1. Can you think of a reliable way to distinguish the past tense and past
participle of a verb, regardless of whether it is regular or irregular?
(Hint: think of words or classes of words that often occur with these
forms.)
2. Check a reference grammar for further examples of irregular inec-
tions. Also, for an excellent discussion of this and related issues, read
Pinker (1999).
3. From the following words, determine the three distinct pronuncia-
tions or allomorphs of the past tense morpheme {-ed}: towed, sighed,
tapped, tabbed, tossed, buzzed, raided. Specify the phonological envi-
ronment in which each allomorph occurs. (Hints: look at the last sound
of the word to which the morpheme is added and think of the allo-
morphs of the plural morpheme discussed earlier.)
4. Pinker (1999) notes that children learning English as their native
language sometimes produce forms like goed and readed. Why do you
think they do this?
5. Would you expect adult non-native learners of English to produce
129
Morphology and Word Formation
forms similar to those of native speaking children? What further dif-
culties might non-native speakers have that native English-speaking
children might not have? (Hints: think of the frequency of irregular
forms in English and think of your own experience in learning a second
language.)
English derivational morphology
Derivation is the process of creating separate but morphologically related
words. Typically, but not always, it involves one or more changes in form. It
can involve prexing, as in resaw, and suxing, as in sawing, sawer, sawable.
Another type of derivation, while not visible, is at least audible. It in-
volves a change in the position of the primary stress in a word. Compare:
(3) permit (noun) permit (verb)
contact (noun) contact (verb)
perfect (adj.) perfect (verb)
convert (noun) convert (verb)
In some derivationally related word pairs, only a feature of the nal con-
sonant changes, usually its voicing:
(4) advice advise /s/ /z/
belief believe /f/ /v/
mouth mouthe /T/ //
breath breathe /T/ //
In some cases adding a derivational morpheme induces a change in a
stressed vowel:
(5) divine divinity /aI/ /I/
profane profanity /e/ //
serene serenity /i/ /E/
In other cases, the addition of a sux triggers a change in the nal con-
sonant of the root. For example, an alveolar consonant becomes palatal with
the same voicing value:
(6) part partial /t/ /S/
face facial /s/ /S/
Delahunty and Garvey
130
seize seizure /z/ /Z/
remit remission /t/ /S/
In a multi-syllabic word with a stressed tense vowel, the palatalization
may be accompanied by a laxing of that vowel:
(7) collide collision /d/ /Z/ /aI/ /I/
elide elision /d/ /Z/ /aI/ /I/
Sometimes the addition of a derivational ax requires a change in the
stress pattern, with consequential changes in the pronunciations of the vow-
els. In most cases an unstressed vowel is pronounced as schwa:
(8) telegraph telegraphy
regal regalia
tutor tutorial
In still other cases we nd suxing, stress migration with change of vow-
el quality, and change of consonant:
(9) approve approbation /u/ /@/ /v/ /b/
Additionally, English allows us to change a word’s part of speech without
any change of form. As a result, identical forms may belong to dierent
parts of speech, e.g., saw the noun and saw the verb:
(10) a. is saw is too dull. (noun)
b. Dont saw that board. (verb)
Other examples include hit, buy, dust, autograph, brown-bag, which can all
be both verbs and nouns. Change of part of speech without any correspond-
ing formal change is called conversion (also functional shift or zero deriva-
tion). ere is more on this topic in our chapter on Major Parts of Speech.
Exercise
1. Write each of the example words in (3)-(9) in a phonemic notation.
131
Morphology and Word Formation
2. True or False?
a. Every English word contains at least one root.
b. In English, derivational morphemes occur before inectional mor-
phemes.
c. In English, derivational suxes regularly occur before inection-
al suxes.
d. In English, a few inectional morphemes can occur as prexes.
e. Every root in English is a free morpheme (i.e., there is no such
thing as a bound root.) (Hint: consider receive, deceive, con-
ceive, perceive.)
f. In English, some morphemes have both a free and a bound al-
lomorph. (Hint: consider able, ability; France, Franco-.)
3. a. In a broad phonetic (phonemic) transcription, transcribe the
sounds represented by the bolded letters in impossible, ined-
ible, illegible, irresponsible.
b. What meaning do these pairs of letters have in common?
c. What is the rst sound in all four pairs of sounds?
d. What are the second sounds in the pairs of sounds?
e. Why does the second sound vary as it does?
f. How would you analyze this variation in terms of morphemes and
allomorphs?
4. As English readily allows conversion, you should have no trouble com-
piling a list of ten pairs of words with identical forms but dierent
parts of speech. For each pair of words, create a pair of short sentences
that show that the words do belong to dierent parts of speech.
As we’ll see in more detail in the next chapter, words belonging to dier-
ent parts of speech take dierent inections—e.g., {
N
saw} + {
pl
-s}; {
V
saw} +
{-ed}. Because derivationally related forms often belong to dierent parts of
speech and consequently allow dierent inections, and because the mean-
ings of derivationally related pairs are not always as parallel as their forms
are, derived forms may be given their own entries in dictionaries. Websters
New World Dictionary, for instance, has separate entries for generate and
generation and for compete and competence. Look up these words in your own
dictionary and note how the meanings of generation and competence are not
entirely predictable from those of generate + {-ion} and compete + {-ence},
Delahunty and Garvey
132
respectively.
e term word family is often used for a set of words that are related to
each other derivationally or inectionally, though the term is also used to
refer to any set of words that rhyme with each other.
Compounding
e italicized words in (11) are created by combining saw with some other
word, rather than with a bound morpheme.
(11) a. A sawmill is a noisy place.
b. Every workshop should have a chain saw, a table saw, a jig-saw, a
hack saw, and a bucksaw.
c. Sawdust is always a problem in a woodworker’s workshop.
d. Sawing horses are useful and easily made.
Such words are called compounds. ey contain two or more words
(or more accurately, two or more roots, all, one, or none of which may
be bound; cf. blueberry with two free morphemes, and astronaut with two
bound morphemes). Generally, one of the words is the head of the com-
pound and the other(s) its modier(s). In bucksaw, saw is the head, which
is modied by buck. e order is signicant: compare pack rat with rat
pack. Generally, the modier comes before the head.
In ordinary English spelling, compounds are sometimes spelled as single
words, as in sawmill, sawdust; sometimes the parts are connected by a hy-
phen, as in jig-saw; and sometimes they are spelled as two words, as in chain
saw, oil well. (Dictionaries may dier in their spellings.) Nonetheless, we are
justied in classifying all such cases as compound words regardless of their
conventional spelling for a variety of reasons.
First, the stress pattern of the compound word is usually dierent from
the stress pattern in the phrase composed of the same words in the same
order. Compare:
(12) compound phrase
White House white house
funny farm funny farm
blackbird black bird
atcar at car
133
Morphology and Word Formation
Exercise
Very bad teenager joke:
Q: How do you make a cat drink?
A: Put it in a blender.
What are the verbal tricks here?
In the compounds the main stress is on the rst word; in the phrases the
main stress is on the last word. While this pattern does not apply to all com-
pounds, it is so generally true that it provides a very useful test.
Second, the meaning of the compound may dier to a greater or lesser
degree from that of the corresponding phrase. A blackbird is a species of
bird, regardless of its color; a black bird is a bird which is black, regardless
of its species. A trotting-horse is a kind of horse, regardless of its current ac-
tivity; a trotting horse must be a horse that is currently trotting. So, because
the meanings of compounds are not always predictable from the meanings
of their constituents, dictionaries often provide individual entries for them.
ey do not do this for phrases, unless the meaning of the phrase is idi-
omatic and therefore not derivable from the meanings of its parts and how
they are put together, e.g., raining cats and dogs. Generally the meaning of a
phrase is predictable from the meanings of its constituents, and so phrases
need not be listed individually. (Indeed, because the number of possible
phrases in a language is innite, it is in principle impossible to list them all.)
ird, in many compounds, the order of the constituent words is dier-
ent from that in the corresponding phrase:
(13) compound phrase
sawmill mill for sawing
sawing horse horse for sawing
sawdust dust from sawing
Fourth, compound nouns allow no modication to the rst element.
is contrasts with noun phrases, which do allow modication to the modi-
er: compare *a really-blackbird and a really black bird.
ere are a number of ways of approaching the study and classication of
compound words, the most accessible of which is to classify them according
to the part of speech of the compound and then sub-classify them according to the
parts of speech of its constituents. Table 2 is based on discussion in Bauer (1983).
Delahunty and Garvey
134
1. Compound nouns
a. Noun + noun: bath towel; boy-friend; death blow
b. Verb + noun: pickpocket; breakfast
c. Noun +verb: nosebleed; sunshine
d. Verb +verb: make-believe
e. Adjective + noun: deep structure; fast-food
f. Particle + noun: in-crowd; down-town
g. Adverb + noun: now generation
h. Verb + particle: cop-out; drop-out
i. Phrase compounds: son-in-law
2. Compound verbs
a. Noun + verb: sky-dive
b. Adjective + verb: ne-tune
c. Particle + verb: overbook
d. Adjective + noun: brown-bag
3. Compound adjectives
a. Noun + adjective: card-carrying; childproof
b. Verb + adjective: fail safe
c. Adjective + adjective: open-ended
d. Adverb + adjective: cross-modal
e. Particle + adjective: over-qualied
f. Noun + noun: coee-table
g. Verb + noun: roll-neck
h. Adjective + noun: red-brick; blue-collar
i. Particle + noun: in-depth
j. Verb + verb: go-go; make-believe
k. Adjective/Adverb + verb: high-rise;
l. Verb + particle: see-through; tow-away
4. Compound adverbs
uptightly
cross-modally
5. Neo-classical compounds
astro-naut
hydro-electric
mechano-phobe
 :   (, )
An alternative approach is to classify compounds in terms of the seman-
tic relationship between the compound and its head. e head of a com-
135
Morphology and Word Formation
pound is the constituent modied by the compound’s other constituents.
In English, heads of compounds are typically the rightmost constituent (ex-
cluding any derivational and inectional suxes). For example, in trac-cop
the head is cop, which is modied by trac; in line-backer the head is backer,
which is modied by line. Linguists distinguish at least three dierent se-
mantic relations between the head and modier(s) of compounds.
First, the compound represents a subtype of whatever the head repre-
sents. For instance, a trac-cop is a kind of cop; a teapot is a kind of pot; a
fog-lamp is a kind of lamp; a blue-jay is a kind of jay. at is, the head names
the type, and the compound names the subtype. ese are called endocen-
tric compounds.
Second, the compound names a subtype, but the type is not represented
by either the head or the modier in the compound. For example, Dead-
head, redhead, and pickpocket represent types of people by denoting some
distinguishing characteristic. ere is typically another word, not included
in the compound, that represents the type of which the compound repre-
sents the subtype. In the case of Deadhead, redhead, and pickpocket this other
word is person, so a Deadhead is a person who is an enthusiastic fan of the
band e Grateful Dead. ese are called exocentric compounds.
ird, there are compounds in which both elements are heads; each con-
tributes equally to the meaning of the whole and neither is subordinate
to the other, for instance, bitter-sweet. Compounds like these can be para-
phrased as both X and Y, e.g., “bitter and sweet.” Other examples include
teacher-researcher and producer-director. ese can be called coordinative
compounds.
Exercise
For each set of words below, say whether the words are endocentric,
exocentric, or coordinative compounds. Justify your identication.
a. redneck, yellowjacket, cocktail, blackhead
b. armchair, breathtest, rockopera
c. secretary-treasurer, scholar-administrator
As a third (and nal) possible mode of analyzing compounds we briey
consider that used in the series of modern traditional grammars prepared
by Quirk, Greenbaum, Leech and Svartvik (1972, 1985). In this method,
the compounds are analyzed and classied according to the relationships
among their constituents when the meaning of the compound is expressed
Delahunty and Garvey
136
as a phrase or clause. For example:

bee-sting a sting by a bee
blood-test a test of blood
swimming pool a pool for swimming
adding machine a machine for adding
girlfriend a friend who is a girl
killer shark a shark which is a killer
windmill a mill powered by wind
motorcycle a cycle powered by a motor
self-control someone able to control self

sunrise when the sun rises
 :  /   

Exercise
Paraphrase each of the following compounds according to at least one
of the patterns in Table 3.
babysitter, catsh, cry-baby, story-teller, dancing girl, darkroom,
doorknob, taxpayer, security ocer, sleepwalking
Other sources of words
Besides derivation and compounding, languages make use of coining, ab-
breviating, blending, and borrowing to create new words.
Coining is the creation of new words without reference to the existing
morphological resources of the language, that is, solely out of the sounds
of the language. Coining is very rare, but googol [note the spelling] is an
attested example, meaning 10
100
. is word was invented in 1940 by the
nine-year-old nephew of a mathematician (see Compact Edition of the Ox-
ford English Dictionary Vol. III Supplement to the OED Vols. I-IV: 1987
p. 317).
Abbreviation involves the shortening of existing words to create other
words, usually informal versions of the originals. ere are several ways to
abbreviate. We may simply lop o one or more syllables, as in prof for profes-
sor, doc for doctor. Usually the syllable left over provides enough information
137
Morphology and Word Formation
to allow us to identify the word it’s an abbreviation of, though occasion-
ally this is not the case: United Airliness low cost carrier is called Ted. (Go
gure!) Alternatively, we may use the rst letter of each word in a phrase
to create a new expression, an acronym, as in UN, US, or SUV. In these
instances the acronym is pronounced as a sequence of letter names. In other
instances, such as UNICEF from United Nations International Children’s
Emergency Fund, the acronym can be pronounced as an ordinary English
word. Advertisers make prolic use of acronyms and often try to make them
pronounceable as ordinary words.
Blending involves taking two or more words, removing parts of each,
and joining the residues together to create a new word whose form and
meaning are taken from the source words. Smog derives from smoke and
fog and means a combination of these two substances (and probably lots of
others); motel derives from motor and hotel and refers to hotels that are con-
venient in various ways to motorists; Prevacid derives from prevent acid; era-
cism derives from erase and racism and means erase racism or, if read against
the grain, electronic racism (cf. email, ecommerce, E-trade); webinar derives
from (worldwide) web and seminar. In November 2007, an interviewee on
an NPR news item created the blend snolo to refer to playing bike polo in
the snow.
Borrowing involves copying a word that originally belonged in one lan-
guage into another language. For instance, many terms from Mexican cui-
sine, like taco and burrito, have become current in American English and are
spreading to other English dialects. Borrowing requires that the borrowing
language and the source language come in contact with each other. Speak-
ers of the borrowing language must learn at least some minimum of the
source language for the borrowing to take place. Over its 1500 year history
English has borrowed from hundreds of languages, though the main ones
are Latin (homicide), Greek (chorus), French (mutton), Italian (aria), Span-
ish (ranch), German (semester), and the Scandinavian languages (law). From
Native American languages, American English has borrowed place names
(Chicago), river names (Mississippi), animal names (opossum), and plant
names (hickory).
e borrowed word never remains a perfect copy of its original. It is
made to t the phonological, morphological, and syntactic patterns of its
new language. For example, the Spanish pronunciation of burritos is very
dierent from the English pronunciation. At the very least, the two languag-
es use dierent /r/s and /t/s, and the plural marker {-s} is voiced in English
but voiceless in Spanish.
See our chapter on the History of the English Language in Book II for
Delahunty and Garvey
138
more on borrowing.
registers and words
Although most of the words we use every day can be used in almost any
context, many words of the language are restricted to uses in certain elds,
disciplines, professions, or activities, i.e., registers. For example, the word
phoneme is restricted to the linguistic domain. Interestingly, some words
may be used in several domains with a dierent meaning in each, though
these meanings may be a specic version of a more general meaning. For ex-
ample, the word morphology is used in linguistics to refer to the study of the
internal structure of words and their derivational relationships; in botany
to refer to the forms of plants; in geology to refer to rock formations. e
general, abstract meaning underlying these specic meanings is the study
of form.
Besides words that may be used in almost any context and those that are
technical or discipline specic, there are words that play important roles
in academic discourses generally, for example, accuracy; basis; concept and
its related forms, conception, conceptual, conceptualize; decrease; eect; fac-
tor; indicate and its related forms, indication, indicative; and result. As such
words are used across disciplines, generally without local idiosyncrasies of
meaning, they are important words for English learners, both native and
non-native speakers. For a useful overview of the attempts to create lists of
such academic (or subtechnical) words and a new list of them, see Cox-
head (2000) and the references therein (another academic word).
the internal struc t u r e o f c o m p l e x wo r d s
Complex words (those composed of more than one morpheme) are not
merely unstructured sequences of morphemes. For example, the plural {-s}
sux on dropouts must be added to the entire compound dropout, not to out
to which drop is then added. e reason for this is that the plural sux may
be attached to nouns, but not to verbs or particles. Drop and out constitute
a noun only after they have been brought together in the compound.
We can use brackets with subscripts to represent these relations:
[
N
[
N
[
V
drop][
Prt
out]]s]. Alternatively, and equivalently, we can use tree dia-
grams to indicate the parts (constituents) of complex words and their struc-
tural relations:
139
Morphology and Word Formation
(14) N
N Pl
V Prt
drop out s
Consider another example: unreadability. We analyze this word as
[
N
[
Adj
un
1
[
Adj
[
V
read]abil]]ity], represented by the following tree:
(15) N
Adj Sux
Prex Adj
V Sux
un
1
read abil ity
Let’s consider this analysis more closely. e sux {-able} attaches to verbs
to create adjectives. Besides readable we have the adjectives doable, manage-
able, and attachable, which are derived from the verbs read, do, manage, and
attach, respectively. We can represent this part of the word as: [
Adj
[
V
read]
able].
e prex {un
1
-} attaches to adjectives, meaning “not” or “the converse
of.” Compare unwise, unfair, ungrateful, uncomfortable, unmanageable with
unreadable. All can be glossed as not having the quality denoted by the
adjective to which they are attached: “not comfortable,” “not fair,” etc. is
morpheme must be distinguished from the prex {un
2
-} meaning “to reverse
the action,” which can be attached only to verbs (e.g., untie).
{Un
1
-} cannot attach to the verb read; although there is the word unread,
pronounced [@nrEd], not [@nrid], an adjective meaning “not read” and de-
rived from the past participle of read. Consequently, in unreadable, {-able}
must be attached to {read} to create the adjective readable. {Un
1
-} may then
be attached to readable to create unreadable. We will represent this part of
the word as: [
Adj
un
1
[
Adj
[
V
read]able]].
e sux {-ity} attaches to adjectives to create abstract nouns. Conse-
quently it must be attached to the adjective unreadable. e structure of
Delahunty and Garvey
140
the entire word therefore must be: [
N
[
Adj
un
1
[
Adj
[
V
read]able]]ity], as specied
above. In pronunciation the morpheme {-able} will be assigned its allo-
morph /@bIl/ (spelled <abil>, the same allomorph that appears in ability).
Exercise
Provide an analysis tree for each of the following words: retry, sink-
able, thoughtless, meaningfulness, microorganisms.
classifying words by their morphological
properties
Once the morphemes of a language have been identied, their allomorphs
determined, and their distributions specied, we can use our analysis to as-
sign the words of a language to parts of speech. For many words, inections
provide the main basis of this assignment. Refer to Table 1 for the list of
English inections.
Nouns can be identied as those words that can be inected for plural.
Verbs are words that can be inected for 3rd person singular present
tense, past tense, past participle, and progressive. ese forms are often re-
ferred to as the principal parts of the verb.
Short adjectives and adverbs are words that can be inected for compara-
tive and superlative.
Derivational regularities can also be used to classify words. We can, for
example, classify as adverbs words derived from adjectives by the addition of
the sux {-ly}, e.g., quickly.
Classifying words on the basis of their internal morphological structure
works only up to a point. ere are lots of words that are not internally com-
plex and so cannot be classied without recourse to other types of criteria.
For example, the preposition to has no internal morphological structure and
so cannot be assigned to a grammatical class on that basis. Likewise, adverbs
such as hard or fast lack the characteristic {-ly} ending. It becomes necessary
to use other criteria to classify these and many other words. We consider in
detail the principles which have been proposed for assigning words to parts
of speech in the chapters on Major and Minor Parts of Speech in this book.
Exercise
1. Discuss two relatively reliable criteria (don’t use spelling) for distin-
guishing words from morphemes and phrases. Illustrate your discussion
141
Morphology and Word Formation
with appropriate examples.
2. Derivation displays a range of patterns in English. Discuss three dif-
ferent derivational patterns, illustrating your description with appro-
priate examples.
references and resources
Beers, Kylene. 2003. When Kids can’t Read. Portsmouth, NH: Heinemann.
Bauer, Laurie. 1983. English Word-formation. London, UK: Cambridge
University Press.
_____ 1988. Introducing Linguistic Morphology. Edinburgh, UK: Edinburgh
University Press.
Booij, Geert. 2005. e Grammar of Words. Oxford, UK: Oxford University
Press.
Coxhead, Averil. 2000. A new academic word list. TESOL Quarterly 34, 2:
213-238.
Haspelmath, Martin. 2002. Understanding Morphology. London: Arnold.
Huddleston, Rodney and Georey K. Pullum. 2002. e Cambridge Grammar
of the English Language. Cambridge, UK: Cambridge University Press.
Matthews, P.H. 1974. Morphology: An Introduction to the eory of Word
Structure. Cambridge, UK: Cambridge University Press.
Pinker, Steven, 1999. Words and Rules: e Ingredients of Language. New York:
Basic Books.
Quirk, R., S. Greenbaum, G. Leech, and J. Svartvik. 1972. A Grammar of
Contemporary English. New York: Seminar Press.
_____ 1985. Comprehensive Grammar of the English Language. London, UK:
Longman.
Spencer, A. 1991. Morphological eory: An Introduction to Word Structure in
Generative Grammar. Oxford, UK: Blackwell.
glossary
affix: an inectional or derivational morpheme; to attach an inectional or
derivational morpheme to an expression.
allomorph: variant phonological representation of a morpheme.
auxiliary verb: a verb other than the main verb of a clause.
base: part of word to which an ax may be attached; may but need not be
a root morpheme.
bound morpheme: a morpheme that must be attached to another mor-
pheme.
Delahunty and Garvey
142
constituent: a unied part of a construction (e.g., of a word, phrase, or
sentence).
conversion: derivational relationship between two words of dierent parts
of speech but without any formal marking of the dierence.
coordinative compound: a compound word that denotes an entity or
property to which both constituents contribute equally; e.g., bittersweet re-
fers to a quality which is both bitter and sweet.
derivation: process of changing a word from one part of speech to another
or from one subclass to another, typically by making some change in form.
endocentric compound: a compound word that denotes a subtype of
whatever is denoted by the head. Armchair represents a type of chair; breath-
test represents a kind of test.
exocentric compound: a compound word that denotes a subtype of a cate-
gory that is not mentioned within the compound; e.g., pickpocket represents
a kind of person, not a kind of pocket nor a kind of pick.
free morpheme: a morpheme that need not be attached to another mor-
pheme, but can constitute a word on its own.
head: the main constituent of a compound, which may be modied by the
compound’s other constituents.
inflectional morpheme: a bound morpheme that signals a grammatical
function and meaning in a specic sentence, e.g., plural {-s}, past tense
{-ed}, comparative {-er}, superlative {-est}.
morph: a minimal meaningful form, regardless of whether it is a morpheme
or allomorph.
morpheme: the smallest part of a word that has meaning or grammatical
function.
prefix: a bound morpheme attached before a root.
realization: the representation of one or more abstract elements (e.g., mor-
phemes) by concrete elements (e.g., sounds); e.g., women represents the
morphemes {woman} + {}.
root: the basic constituent of a word, to which other morphemes are at-
tached.
suffix: a bound morpheme attached after a root.
suppletion: irregular inectional forms of a word resulting from the combi-
nation of historically dierent sources; e.g., go/went.
appe n d i x a : s o m e e n g l i s h d e r i vational
morph e m e s
(See Beers 2003: Appendixes D and E for other lists of roots and deriva-
tional axes.)
143
Morphology and Word Formation
Prexes
Class/category changing
a-blaze Adj < V
be-calm V < Adj
be-friend V < N
en-tomb V < N
Class maintaining
Nouns
arch-monetarist
mal-nutrition
micro-dot
mini-dress
step-father
Verbs
de-escalate
Adjectives
a-typical
cis-lunar
extra-sensory
Noun or Verb
fore-ground
bacl-ground
mis-fortune
mis-lead
re-arrangement
Noun or Adjective
ex-President
ex-orbital
in-denite
mid-morning
mid-Victorian
Verb or Adjective
Delahunty and Garvey
144
circum-navigate
circum-polar
Noun, Verb, or Adjective
co-author
counter-culture
counter-demonstrate
counter-intuitive
dis-ambiguate
dis-bound
dis-information
inter-mix
sub-conscious
sub-let
Suxes
Creating Nouns
from Nouns
-dom king-dom
-er Birch-er
-ess lion-ess
-ette kitchen-ette
-iana Victor-iana
-hood man-hood
-ism absentee-ism
-let stream-let
-ling duck-ling
-scape sea-scape
-ship kin-ship
from Verbs
-al arriv-al
-ary dispens-ary
-ation (esp. with -ize) categor-iz-ation
-ee blackmail-ee
-er kill-er
-ment manage-ment
-ure clos-ure
145
Morphology and Word Formation
from Adjectives
-ce dependen-ce
-cy excellen-cy
-dom free-dom
-er six-er
-hood false-hood
-ist social-ist
-ity divin-ity
-ness good-ness
-th warm-th
Derived Verbs
from Nouns
-fy metr-ify
-ize Cambodian-ize
from Adjectives
-en short-en
Derived Adjectives
from Nouns
-al education-al (allomorphs/allographs: -ial, -ual:
presidential, habitual)
-ate passion-ate
-en wood-en
-ese Peking-ese
-esque pictur-esque
-ful doubt-ful
-ic algebra-ic
-less clue-less
-ly friend-ly
-ous venom-ous
-y catt-y
from Verbs
-able unbeliev-able
-less count-less
Delahunty and Garvey
146
-ant/-ent absorb-ent
-atory arm-atory
-ful resent-ful
-ive generat-ive
from Adjectives
-ish green-ish
-ly good-ly
Derived Adverbs
-ly slow-ly
-ward(s) in-ward(s)
-wise length-wise
Miscellaneous
down-er
i-y, upp-itty
in-ness, much-ness, such-ness,
there-ness, why-ness
thus-ly