Chapter 2: Words:
The Building Blocks of Grammar
The previous chapter went to some length to show that the aim of grammar is to label
the forms and functions of words, phrases, and clauses. Words are thus the basic building
blocks of grammar. From the point of view of grammar, words can be treated as
indivisible units, and that is what we will do throughout most of this book. It will
nevertheless be useful to us, before taking up grammar directly, to look inside words and
learn something about their internal composition. We will see later that, by doing so, we
will be better able to define parts of speech and to explain exactly how words pattern in
phrases and clauses.
MORPHEMES AND WORDS
At first glance, it might seem that words are not only the most basic units of grammar
but also the most basic units of meaning in language. But it takes little reflection to
realize that we can often identify meaningful parts of words. Notice that the following
words are all composed of two meaningful parts, the first of which is different in each
word and the second of which, spelled s, is the same. It means 'more than one' or 'plural.'
2.1a
boys, bags, tails, trains, hams
2.1b
tacks, bats, cuffs
English therefore has the meaningful word parts boy, bag, tail, train, ham, tack, bat, and
cuff, each of which refers to some different thing and each of which, by the way, can
stand alone as a whole word. In addition, there is another meaningful word part, s,
which, when we attach it to boy, bag, etc., makes us think of more than one boy, bag, etc.
We can call any indivisible meaningful word part a morpheme. The term morpheme
is based on Ancient Greek, where morph meant 'form' or 'shape.' The term morpheme
means 'class of forms or shapes' or 'group of forms or shapes,' i.e., a group of things that
together constitute one meaningful word part. But why would we want to call boy or its s
ending a 'group'? We do so because one morpheme has the possibility of manifesting
itself in two or more (i.e. a group of) spellings or pronunciations. If we add the plural
morpheme to the words bush, watch, and judge, we do not spell it s; we spell it es:
2.2
bushes, watches, judges
And if we add the plural morpheme to ox, or child, we spell it en:
10
WORDS: THE BUILDING BLOCKS OF GRAMMAR
2.3
oxen, children
So the plural morpheme is indeed a group of separately spelled letter sequences, s, es,
and en. But these three different formal sequences of letters all have the same meaning,
'plural,' and they all appear in the same position, i.e., at the end of nouns, and thus are
said to have the same function. The concept of the morpheme always allows the
possibility of two or more different forms, such as s, es, and en, even though most
morphemes, like boy, have only one form. For a set of different forms (sequences of
letters or sounds) to be grouped together as variants of one morpheme, they must have the
same meaning and the same function. To show this unity when explicitly analyzing the
morpheme sequences in a word, we can invent abstract labels for morphemes, such as -es
for the plural morpheme. (The hyphen indicates that this morpheme cannot occur as a
word on its own but must be attached to another morpheme on the side where the hyphen
appears.) We can now represent the morphological makeup of the previously discussed
example words as follows:
2.1a'
boy-es, bag-es, tail-es, train-es, ham-es
2.1b'
tack-es, bat-es, cuff-es
2.2'
bush-es, watch-es, judg-es
2.3'
ox-es, child-es
Notice that it is not just word endings like the plural morpheme that can vary their
form in different words in which they appear. Take a look again at the morpheme
sequence child-es in 2.3'. We have already noted that the -es morpheme is spelled en
when it is attached to the morpheme child. But we have not until now noted explicitly
that the morpheme child also has two spellings; it is spelled child when it occurs alone
but childr (with an added r) when the plural morpheme is attached to it. The variant
forms of a morpheme are called allomorphs. Thus we can say that the morpheme -es has
three allomorphs, spelled s, es, and en, and that the morpheme child has two allomorphs,
spelled child and childr. In this discussion, we are focusing on the varied spellings of
allomorphs, but notice that there is a similar variation in pronunciation. Take another
look at 2.1a and 2.1b, this time reading them aloud and listening to your pronunciation of
the final letter:
2.1a
boys, bags, tails, trains, hams
2.1b
tacks, bats, cuffs
11
WORDS: THE BUILDING BLOCKS OF GRAMMAR
Even though the plural morpheme has only one written allomorph in all the above words,
spelled s, it has two allomorphs in spoken English. Did you notice that, for all of the
words in 2.1a you pronounced the plural morpheme with a 'z' sound, but for all of the
words in 2.1b you pronounced it with an 's' sound.
Some linguists assert that the plural morpheme appears at the end of the words listed
in 2.4, 2.5, and 2.6 even though it is neither spelled nor pronounced. In such cases it is
said to have a 'zero' allomorph:
2.4
sheep, fish, series
2.5
teeth, geese, feet
2.6
data, criteria, alumni
We would thus analyze the morphological composition of the above words as follows:
2.4'
sheep-es, fish-es, series-es
2.5'
tooth-es, goose-es, foot-es
2.6'
datum-es, criterion-es, alumnus-es
The alternative to the above analysis would be to claim that English has two words
spelled sheep, one having just the meaning, '(a) sheep' and the other having the meaning
'more than one sheep.' Similarly, one would need to assert that tooth and teeth are
entirely separate words and that datum and data are separate words. But most people
would agree that the plural words listed in 2.4, 2.4, and 2.6 are indeed the same words as
their 'singular' counterparts. The analysis proposed in 2.4'-2.6' has the further advantage
of treating teeth as an allomorph of the morpheme tooth, an allomorph which is
constrained to occur preceding the plural morpheme -es.
Here is a summary of what we have learned so far about the plural morpheme:
2.7
Morpheme Label: -es
Meaning: 'plural'
Function (Position): Occurs at the end of nouns
Variant Forms (Allomorphs):
The s allomorph as in:
(1a) boys, bags, tails, trains, hams (pronounced 'z' in
speech)
(1b) tacks, bats, cuffs (pronounced 's' in speech)
The es allomorph as in:
(2) bushes, watches, judges
The en allomorph as in:
(3) oxen, children
12
WORDS: THE BUILDING BLOCKS OF GRAMMAR
The 'zero' allomorph as in:
(4) sheep, fish, series
(5) teeth, geese, feet
(6) data, criteria, alumni
It is especially common for word endings like the plural morpheme to have more
allomorphic variation than morphemes which can themselves be whole words. But some
whole-word morphemes can have significant allomorphic variation, especially in speech.
Here is a summary like the one for the plural morpheme in 2.7 for one such morpheme:
2.8
Morpheme Label: sign
Meaning: 'something that stands for or points to something else'
Function (Position): noun (not necessarily attached to any other morpheme)
Variant Forms (Allomorphs):
The sign allomorph (pronounced 'sayn') as in: sign, assign, consign,
cosign
The sign allomorph (pronounced 'zayn') as in: design, resign
The sign allomorph (pronounced 'sign' -- with 'hard g') as in: signify,
signature
The sign allomorph (pronounced 'zign' -- with 'hard g') as in:
designate, resignation
There is a strong tendency in English, despite the complaints that we often hear about
English spelling, to be consistent in spelling allomorphs of the same morpheme the same
way, even when they are pronounced differently. Thus even though the morpheme sign
has four different pronunciations, it is always spelled sign. This makes the English
spelling system much more consistent and effective in its primary purpose: to convey
meaning visually. Its primary task is not to represent the pronunciation of words (adult
speakers of English after all know how to pronounce words) but to communicate ideas.
Although morphemes are often defined as 'the basic units of meaning in language,'
form and function play an equal role in defining them. Here, as a summary of this
section, is an explicit definition of this important linguistic concept: A morpheme, such
as the -es morpheme in English, encompasses (a) one or more forms (sound or letter
sequences such as s, es, and en) that (b) share the same function, i.e., position in a word
(in the case of -es, the final position in a noun), and (c) convey the same meaning (in the
case of -es, the idea of 'plural').
MORPHEME TYPES
Roots and Affixes
The task of morphological analysis is not complete even when all the morphemes are
labeled, their meanings and functions specified, and their variant forms (allomorphs)
listed. The morphemes which compose a word may be of several types. Consider the
word subtract, which is composed of two morphemes, sub- and -tract. Even though their
meanings are abstract, they are fairly consistent; sub meaning "down" or "under" and
tract "to pull" or "to take." Now, let us see what we can learn about these two morphemes
13
WORDS: THE BUILDING BLOCKS OF GRAMMAR
by searching for other morphemes that can substitute for each of them in the word
subtract. We begin with tract. Here is just a sampling of words with another morpheme
substituted for tract after sub-:
2.9a
subjoin, sublease, sublet, submerge, subtend
2.9b
subject, submerse, submit, subsist, subsume, subvert
If we searched in a dictionary, we could surely find dozens, if not hundreds, of other
words to add to the above lists. However, even with access to a dictionary, we would
find relatively few morphemes that can substitute for sub- before tract. Here is a
virtually complete list:
2.10
attract, contract, detract, distract, extract, protract, retract
The lists in 2.9 and 2.10 point out an important difference between sub- and tract. If we
view the word subtract as composed of two positions (one occupied by sub- and the other
by tract), we immediately notice that for each position there are only certain morphemes
that can replace the one already there and still make up a genuine word. Very many
morphemes can occupy the position of tract but only relatively few can occupy the
position of sub-. Furthermore, none of the morphemes that can replace sub- can replace
tract and vice versa. This is not an isolated phenomenon true only of the word subtract.
Consider the word express, composed of the morpheme ex- and the morpheme press.
How many morphemes can you think of that can substitute for press after ex-? Here is a
partial list:
2.11a
exclaim, exchange, expose, extend
2.11b
exalt, excise, exclude, excuse, exempt, exhume, expand, expect, expel, explain, explode, export,
expunge, extol, extort, extract, exude
Now, how many morphemes can you think of that can substitute for ex- before press?
Here is a virtually complete list:
2.12
compress, depress, oppress, repress, suppress
Many other examples like these can be cited, examples which show that some
morphemes, like tract and press, occupy positions in words where there is a relatively
unlimited potential for substituting other morphemes (linguists call such morphemes
roots) and that other morphemes, like sub- and ex-, occupy positions where there is a
relatively limited potential for substitution (these are called affixes). In our two
examples, affixes came before roots. Affixes may also follow roots. Consider the word
14
WORDS: THE BUILDING BLOCKS OF GRAMMAR
traction, composed of the morpheme tract and the morpheme -ion. It would be easy to
show that many morphemes can replace tract in front of -ion (for example, tension,
fusion, lesion, version, portion, fission), and thus tract is a root, but that very few
morphemes (perhaps only -able and -or) can replace -ion after tract, and thus -ion is an
affix. But notice also that, even though we can call sub-, ex-, and -ion affixes, only the
first two (sub- and ex-) can precede roots, and only the last one (-ion) can follow roots,
which means that we have to sub-divide affixes into two types. Affixes preceding roots
are called prefixes, and affixes following roots are called suffixes. But prefixes may also
precede other prefixes, as in the word decompression where de- is a prefix that precedes
another prefix, com-, and both precede the root press. And suffixes may also follow
other suffixes as in the word tractability, where -ity is a suffix that follows another suffix,
-abil, and both follow the root tract. Furthermore, a word may have both prefixes and
suffixes (for example, distractible, retraction, expressible, compression). Linguists
generally cite prefixes with a hyphen after them (for example, sub- and ex-) and suffixes
with a hyphen in front of them (-ion and -ible).
Let us consider another phenomenon related to the classification of morpheme types.
You may have asked yourself earlier why the examples in 2.9 and 2.11 were divided into
two groups under a and b. Reexamine these lists of examples, trying to determine what
distinguishes 2.9a and 2.11a from 2.9b and 2.11b. They are reprinted here for your
convenience:
2.9a
subjoin, sublease, sublet, submerge, subtend
2.11a
exclaim, exchange, expose, extend
2.9b
subject, submerse, submit, subsist, subsume, subvert
2.11b
exalt, excise, exclude, excuse, exempt, exhume, expand, expect, expel, explain, explode, export,
expunge, extol, extort, extract, exude
First, note that the roots in 9a and 11a can function as words without the prefix attached:
join, lease, let, merge, tend, claim, change, and pose. Second, notice that all these words
are verbs, which suggests that the functional position of the root morpheme following
sub- and ex- is that of a verb root. Finally, notice that none of the roots of the words
listed in 9b and 11b can function as separate verbs without the prefixes attached: -ject, merse, -mit, -sist, -sume, -vert, -alt, -cise, -clude, etc.. For these reasons, the roots in (9a)
and (11a) are called free roots, and those in (9b) and (11b) are called bound roots. Here
is a summary of the terms and definitions introduced in this section:
2.13a
Roots: morphemes which occupy positions in words where the greatest potential for substitution
exists
(i) Free: capable of functioning as a word
(ii) Bound: incapable of functioning as a word
15
WORDS: THE BUILDING BLOCKS OF GRAMMAR
2.13b
Affixes (all are bound): morphemes which occupy positions in words where only a limited
potential for substitution exists
(i) Prefixes: affixes which precede roots and possibly other prefixes
(ii) Suffixes: affixes which follow roots and possibly other suffixes
Some English Prefixes
Examine the sampling of English prefixes in 2.14. The prefixes are grouped into six
categories based on their meaning. These meaning categories are not exhaustive, but
they do encompass a large number of English prefixes. Two examples of prefixes from
each meaning type are given. As you examine the example words listed beside each
prefix, notice that when a prefix is added to a root, the composite word is generally the
same part of speech as the root without the prefix. Notice also that, while prefixes do not
normally have as many allomorphs as did the -es morpheme we considered earlier, there
is enough variation to reinforce our definition of a morpheme as an abstract meaning
label which may group together a variety of spellings or pronunciations. In this regard,
note especially the different pronunciations of poly- in polyglot and polygamy and the
different pronunciations of hyper- in hyperactive and hyperbole.
2.14a
Number bipoly2.14b
Time
postpre2.14c
Place
intersub2.14d
Degree hyperultra2.14e
Privation un-1
dis2.14f
Negation nonun-2
bifocal, bilingual, biceps, bicycle
polysyllabic, polyglot, polygamy
postwar, postelection, postclassical, postpone
prewar, preschool, pre-19th century, premarital
international, intercontinental, interact, intermarry
subway, subsection, subconscious, sublet, subdivide, subcontract
hypercritical, hyperactive, hypersensitive, hyperbole
ultraviolet, ultramodern, ultraconservative, ultraliberal, ultramarine
undo, untie, unzip, unpack, unleash, unhorse
disconnect, disinfect, disown, displease
nonconformist, nonsmoker, nonpolitical, nondrip
unfair, unwise, unforgettable, unassuming, unexpected
PRACTICE 1 (WORKING WITH ENGLISH PREFIXES)
All the words listed below begin with a prefix. It is possible to sort them out and create a
display very much like the one immediately above in 2.14, containing the same six meaningbased sub-types. Four of the six will have two example prefixes; however, one will have three
examples, and one will have only one example. First, sort them out and make the six-part
display. Second, identify the root of every example word on your list and determine whether it is
bound or free. Here are the words:
16
WORDS: THE BUILDING BLOCKS OF GRAMMAR
decentralize, decode, semiconscious, discourteous, tricycle, archduke, transfer, de-escalate,
archenemy, semifinal, dislike, illogical, disloyal, defrost, amoral, transmit, disobey, semiofficial,
monoplane, transplant, protoplasm, tripod, improper, monorail, irrelevant, insane, asexual,
foreshadow, supersonic, foretell, monotheism, asymmetrical, superstructure, arch-traitor,
prototype, forewarn
FEEDBACK TO PRACTICE 1 (WORKING WITH ENGLISH PREFIXES)
Below is the display called for in Practice 1. Notice how the prefix in- (last line) dramatically
changes its spelling and pronunciation to match the initial sound in the base to which it is
attached.
To provide answers to the second task in Practice 1 (identifying roots as bound or free) I have
underlined the letters in each word that realize the root morpheme. Free roots are simply
underlined; bound roots are both underlined and in boldface type. It is often very difficult to
decide just which letters in a word realize the root morpheme and even more difficult to
determine whether that root is bound or free. Only through careful study of the history of a
word’s form (its etymology) can such determinations be made. For example, the letters “loy” in
the word disloyal in section f below historically descend from the French word loi, which means
“law.” One could thus argue that the letters “loy” in disloyal are an allomorph of law and
therefore realize a free root. I think that “loy” is not an allomorph of law, and thus the root
morpheme represented by those letters is a bound root. There are at least a dozen other similarly
problematic words listed below. If you are not sure why I have identified certain letters as the
root of a word or why I have indicated it as bound or free, you should examine its etymology in a
dictionary
a
Number monotrib
Time
foreprotoc
Place
supertransd
Degree archsemie
Privation def
aNegation disin-
monotheism, monoplane, monorail
tripod, tricycle
foretell, forewarn, foreshadow
protoplasm, prototype
superstructure, supersonic
transplant, transfer, transmit
archduke, arch-enemy, arch-traitor
semi-final, semi-official, semi-conscious
decode, defrost, decentralize, deescalate
amoral, asexual, asymmetrical
disloyal, discourteous, disobey, dislike
insane, illogical, improper, irrelevant
Derivational and Inflectional Suffixes
Although we focused on prefixes as examples in distinguishing between roots and
affixes, we did note that affixes may also be suffixes. In fact, the majority of English
affixes are suffixes. Let us now examine the behavior of some English suffixes.
17
WORDS: THE BUILDING BLOCKS OF GRAMMAR
Consider the verb root pave in the sentences of 2.15. It occurs in (15a) without a suffix
and in (15b) and (15c) with suffixes.
2.15a
Those workers pave streets for a living.
2.15b
The pavement is cracked.
2.15c
They paved it only a few years ago.
Note that the combination of the verb root pave and the suffix -ment in 2.15b results in a
word that is a noun. Other examples of a verb and -ment combining to form a noun are
arrangement and astonishment. We may conclude that when the suffix -ment is affixed
to a verb, the result is a noun. Now notice that when the verb root pave combines with
the suffix -ed in 2.15c the result is still a verb. Thus -ment changes the part of speech of a
root to which it is affixed, but -ed does not. Here is another example. The noun root
symbol occurs without a suffix in 2.16a but with suffixes in 2.16b and 2.16c:
2.16a
Purple is a symbol for royalty.
2.16b
White flags symbolize surrender.
2.16c
Both the circle and the triangle can be symbols.
Note that the combination of the noun root symbol with the suffix -ize in 2.16b results in
a verb. Other examples of a noun root and -ize combining to form a verb are hospitalize
and vaporize. We may conclude that when the suffix -ize is affixed to a noun, the result
is a verb. But notice that when the noun root symbol combines with the suffix -es in
2.16c the resultant word is still a noun. Thus -ize changes the part of speech of a root to
which it is affixed, but -es does not. Here is a list of the two roots and four suffixes we
have just discussed with summary comments in parentheses:
2.17a
pave
symbol
-ment
-ize
-ed
-es
(a verb root)
(a noun root)
(a suffix that changes verbs to nouns)
(a suffix that changes nouns to verbs)
(a verb suffix that means "past")
(a noun suffix that means "plural")
Suffixes like -ment and -ize, which change the part of speech of a word, are called
derivational suffixes. Suffixes like -ed and -es, which do not change the part of speech
of a word, are called inflectional suffixes. But let us look further into the behavior of
these two types of suffixes.
18
WORDS: THE BUILDING BLOCKS OF GRAMMAR
I noted earlier that a root may have two or more suffixes attached to it. Here are two
examples:
2.18a
pavements
2.18b
symbolized
Notice that in both 2.18a and 2.18b each word is composed of a root, a derivational
suffix, and an inflectional suffix. In 2.18a, -ment changes pave to a noun, and thus the -es
that indicates plural in nouns can be added. In 2.18b, -ize changes symbol to a verb, and
thus the -ed that indicates past in verbs can be added. Suppose we tried to reverse this
order of affixation, reasoning as follows: if pave is a verb root, then -ed can be added
giving paved, which is still a verb. Now if -ment changes verbs to nouns, why not add it
to the verb paved giving the noun *pavedment? But the combination of morphemes
indicated by the spelling *pavedment is in fact impossible in English. And so is the
combination *symbolsize, which might be formed by adding first an inflectional and then
a derivational suffix to the root symbol. Such forms are not unreasonable, nor would they
be without a certain usefulness. But they are not in fact possible in English. And thus we
have a second criterion for distinguishing between derivational and inflectional suffixes:
inflectional suffixes can follow derivational suffixes in words, but derivational suffixes
cannot follow inflectional suffixes. Notice that the inflectional suffixes -ed and -es occur
only last in the words listed in 2.19:
2.19
pavement
symbolize
paved
symbols
symbolizable
pavements
symbolized
symbolizability
The morpheme or combination of morphemes to which a derivational suffix is
attached we may call its base, and the morpheme or combination of morphemes to which
an inflectional suffix is attached we may call its stem. So far, we have seen that
derivational suffixes change the part of speech of bases to which they are attached, but
inflectional suffixes do not change the part of speech of stems to which they are attached.
Secondly, derivational suffixes cannot follow inflectional suffixes, but inflectional
suffixes must follow all derivational suffixes.
There is also a third difference between these two types of suffixes, a difference that
can help you distinguish the one type from the other: Words with derivational suffixes
(e.g., pavement and kingdom) are entered separately in the dictionary, but the inflected
forms of words, e.g., the plural of a noun or the past tense of a verb, do not have separate
dictionary entries. This is the case because derivational suffixes are lexical, i.e., they
exist in English precisely to increase the number of words in the language (and so each
19
WORDS: THE BUILDING BLOCKS OF GRAMMAR
new word that a derivational suffix creates deserves its separate entry in a dictionary).
But inflectional suffixes are grammatical, i.e., they exist, not to create new words (the
singular and plural forms of a given noun are forms of the same word) but to shape words
to fit appropriately into grammatical contexts. There is a sense in which a work does not
take on an inflectional suffix until it comes out of the dictionary and is placed in a
sentence. (A word like these or three preceding a noun determine that the plural of the
noun must appear.)
Here is a summary of the three differences between derivational and inflectional
suffixes:
2.21
Derivational Suffixes:
a Usually change part of speech.
b Cannot follow inflectional suffixes.
c Words with derivational suffixes
have separate dictionary entries.
Inflectional Suffixes:
a' Never change part of speech.
b' Cannot precede derivational suffixes.
c' Words with inflectional suffixes
do not have separate dictionary entries.
ENGLISH DERIVATIONAL AND INFLECTIONAL SUFFIXES
Some English Derivational Suffixes
Examine the eighteen derivational suffixes and example words in 2.22. They are in
six groups of three, based on the change they effect in parts of speech. (Notations like
N<V in 2.22c should be read "derives nouns from verbs.") As you examine the words in
2.22, notice two things: First, some derivational suffixes do not meet all three of our
criteria of definition; for instance, the suffixes in 2.22a are affixed to noun bases and the
resultant words remain nouns. Nonetheless the other two criteria classify these suffixes
as derivational. Second, notice that derivational suffixes, like prefixes, do not ordinarily
have many pronunciation variants (allomorphs), but they provide reinforcement for our
definition of a morpheme as an abstract meaning label because they frequently produce
pronunciation variants of morphemes in their bases. Notice how the morphemes sane,
vain, and chaste change vowel pronunciation when -ity is affixed: sanity, vanity, chastity.
And notice that the "k" sound at the end of public, elastic, and fanatic changes to an "s"
sound when derivational suffixes are added: publicity, elasticity, fanaticism.
2.22a
N<N
-hood neighborhood, sisterhood, bachelorhood, boyhood, maidenhood,
knighthood
-ship lordship, township, fellowship, championship, friendship, membership,
lectureship, kinship
-ster gangster, gamester, trickster, songster, punster, mobster, prankster,
speedster
20
WORDS: THE BUILDING BLOCKS OF GRAMMAR
2.22b
N < ADJ/N
-ism
-ity
-ness
2.22c
N<V
2.22d
V < ADJ/N
-al
-er
refusal, dismissal, upheaval, denial, survival, trial, approval, proposal
worker, writer, driver, employer, swimmer, preacher, traveler, teacher,
baker
-ment arrangement, amazement, puzzlement, judgment, astonishment,
pavement
-en
-ify
-ize
2.22e
ADJ < N/V
idealism, impressionism, fanaticism, dualism, realism, imperialism,
romanticism, patriotism
sanity, vanity, rapidity, banality, elasticity, ability, actuality, agility,
chastity
meanness, happiness, cleverness, usefulness, bitterness, brightness,
darkness, goodness
-ful
-ish
-able
ripen, widen, deafen, sadden, harden, lengthen, deepen, strengthen,
neaten
beautify, diversify, codify, amplify, simplify, glorify, nullify, Frenchify
symbolize, hospitalize, publicize, popularize, modernize, idealize
useful, delightful, pitiful, helpful, careful, awful, rightful, sinful, cheerful
foolish, selfish, snobbish, modish, hellish, sheepish, Swedish
acceptable, readable, drinkable, livable, commendable, comfortable,
changeable
2.22f
ADV < ADJ/N -ly
happily, strangely, oddly, athletically, basically
-wise clockwise, lengthwise, weatherwise, educationwise
-ward earthward, homeward, eastward
PRACTICE 2 (WORKING WITH ENGLISH DERIVATIONAL SUFFIXES)
All the words given below end with a derivational suffix. It is possible to sort them out on
the basis of the final suffix and make a display similar to the one in 2.22 (immediately above).
All of the functional groups in 2.22 except the last one (ADV < ADJ/N) are represented. Sort out
the words and make that display. Here are the words:
alarmist, authorization, Baptist, capacitate, cellist, certification, Chinese, Christendom,
civilization, contestant, creamy, deodorant, earldom, earthen, Elizabethan, exploration, facilitate,
famous, flowery, formalization, freighter, fruity, glorification, glorious, hyphenate, icy,
Indonesian, informant, inhabitant, Japanese, journalese, juicy, kingdom, leaden, Londoner, meaty,
modification, novelist, officialdom, orchestrate, Parisian, participant, poisonous, Portuguese,
rainy, Republican, riotous, sandy, silken, stardom, starvation, steamer, stylist, thirsty, traitorous,
typist, vaccinate, waxen, wooden
FEEDBACK TO PRACTICE 2
(WORKING WITH ENGLISH DERIVATIONAL SUFFIXES)
a
N<N
-dom
-er
Christendom, earldom, kingdom, stardom, officialdom
steamer, freighter, Londoner
21
WORDS: THE BUILDING BLOCKS OF GRAMMAR
b
N < ADJ/N
c
N<V
d
V<N
e
ADJ < N/V
-ist
-ese
-(i)an
stylist, cellist, Baptist, alarmist, novelist, typist
Chinese, Portuguese, Japanese, journalese
Indonesian, Parisian, Elizabethan, Republican
-ant
inhabitant, contestant, informant, participant, deodorant
-ation fixation, exploration, starvation, modification, certification, glorification,
authorization, formalization, civilization
-ate
facilitate, capacitate, hyphenate, orchestrate, vaccinate
-en
-ous
-y
wooden, leaden, silken, waxen, earthen
famous, glorious, riotous, traitorous, poisonous
meaty, sandy, creamy, icy, rainy, thirsty, flowery, juicy, fruity
English Inflectional Suffixes
The eight inflectional suffixes of English are listed in 2.23. The word or phrase given
first in parentheses serves both as a name and a rough designation of meaning. Next is
the morpheme label, then words which end in the inflectional suffix. Each line represents
a variant pronunciation (allomorph) of the suffix itself or an allomorph variant it requires
in the pronunciation of certain stems. Examine the lists of examples, preferably by
reading them aloud. Notice that 2.23a and 2.23b are noun inflectional suffixes, 2.23c to
2.23f are verb inflectional suffixes, and 2.23g and 2.23h are ordinarily adjective
inflectional suffixes (though a few adverbs can also have them affixed). Notice also that
many words given to exemplify the past tense inflectional suffix in 2.23d are spelled and
pronounced the same as words listed with the past participle inflectional suffix in 2.23e.
I shall comment below on this and other aspects of the display in 2.23.
2.23a
(plural)
-es
boys, bags, tails, trains, hams,
tacks, bats, cuffs
bushes, watches, judges
wives, knives, thieves
oxen, brethren, children
teeth, geese, feet
data, criteria, alumni
sheep, fish, series
2.23b
(possessive)
-poss
Jim's, Joe's, Mom's, Mr. Moore's
Pat's, my wife's, Mike's
the judge's, the witch's, the boss's
2.23c
(present tense)
-prs
go/goes, buy/buys, sell/sells
take/takes, bat/bats, buff/buffs
push/pushes, watch/watches
am/is/are, have/has, do/does
2.23d
22
WORDS: THE BUILDING BLOCKS OF GRAMMAR
(past tense)
-pst
stayed, tried, behaved, barred
walked, stopped, stuffed, watched
sighted, banded, reminded
chose, rose, spoke, stole, froze
threw, knew, grew, slew
took, stood, forsook
bought, taught, fought
cut, let, hit, bet
went, did, had, made
2.23e
(past participle)
-en
stayed, tried, behaved, barred
walked, stopped, stuffed, watched
sighted, banded, reminded
bought, let, made
taken, eaten, fallen, known
broken, chosen, spoken, stolen
written, given, driven, risen
gone, done, been, seen
2.23f
(present participle) -ing
coming, going, buying, selling
2.23g
(comparative)
-er
bigger, older, quicker
better, worse
2.23h
(superlative)
-est
biggest, oldest, quickest
best, worse
The plural inflectional suffix, which we have labeled -es, has "z," "s," "ez," and "en"
as allomorphs in spoken English, or, when its zero allomorph occurs, it can require a
vowel change in a stem (e.g., tooth > teeth), or a change of pronunciation at the end of a
stem (e.g., criterion > criteria) -- all of which was discussed earlier in this chapter. We
have also noted that the zero allomorph of plural can occur without requiring any change
in the stem to which it is attached: the stems of sheep, fish, and series in 2.23a do not
change pronunciation in any way when the -es inflectional suffix is added, as in John
bought three sheep. Note also that the -es inflectional suffix can cause a variant
allomorph of a stem to appear: whereas the words wife, knife, and thief end in an "f"
sound, the stems of their plurals end in a "v" sound.
Earlier, I defined a morpheme as an abstract meaning label, and I have reemphasized
and illustrated that definition throughout this chapter. One reason I have done so is that it
has important practical implications. As we move ever more deeply into the details of
English grammar later in this book, we will need to refer to word classes such as past
participles. When we do, it will be highly efficient to identify them by noting that they
end with the -en inflectional suffix, and not have to point out that this suffix may be
spelled with the letters en (as in eaten), ne (as in gone), ed (as in arrived), t (as in taught),
etc. The point is that grammatical descriptions focus on abstract form and in doing so
ignore spelling and pronunciation variants. Thus it is important for you to begin right
now to ignore allomorphic variations such as the examples in 2.23 manifest, and to think
23
WORDS: THE BUILDING BLOCKS OF GRAMMAR
of an inflectional suffix like the plural morpheme as if it were always -es no matter how it
might be pronounced or spelled and even if it is not pronounced or spelled at all.
The possessive inflectional suffix, labeled -poss, has three pronunciations: "z" as in
Jim's, "s" as in Pat's, and "ez" as in judge's. But, again, we can ignore such variations
when discussing the behavior of this inflectional suffix and represent the morphemes in
these three examples as Jim-poss, Pat-poss, and judge-poss.
The present tense inflectional suffix, labeled -prs, presents special problems. Many
linguists speak of a "third person singular agreement suffix," which is affixed to a verb in
the present tense when the verb is preceded by he, she, it, or a noun that these pronouns
can replace. This accounts for the "z," "s," and "ez" pronunciations at the end of some
example words in the first three lines of 2.23c; it would also account for is, does, and has
in the fourth line: all three "s-like" allomorphs occur after third person singular pronouns
and nouns. But these "s-like" sounds add no more meaning to a verb stem than it
communicates when preceded by I, we, you, they, or nouns these pronouns can replace.
For this reason, we will consider the "z," "s," or "ez" pronunciations to be allomorphs of
the -prs morpheme which alternate with a zero allomorph that occurs after I, we, and so
forth. Thus the verb in both We eat cheese and He eats cheese is analyzed
morphologically as eat-prs.
The past tense inflection, which we have labeled -pst, has a considerable number of
differently pronounced allomorphs, many of which are exemplified in 2.23d. Notice two
general tendencies: (1) a tendency to indicate 'past' by affixing a "d," "t," or "ed"
pronunciation to the verb stem, as in stayed, walked, and sighted respectively, and (2) a
tendency to indicate "past" by changing the pronunciation of a vowel in the verb stem, as
in chose, threw, and took. Some verbs, like bought, combine both tendencies. Some
verbs, like cut, call for a zero allomorph and make no change at all. Other verbs seem to
manifest the first tendency but also to call for a different pronunciation of the stem: in
went, did, had, and made a "d" or "t" pronunciation of the past tense inflectional suffix
appears, but the pronunciation of the stem in each case varies somewhat from the
ordinary non-past pronunciations (go, do, have, make).
The morphological
representations of went, did, had, and made are go-pst, do-pst, have-pst, and make-pst
respectively. The same kind of analysis is made of those verbs with past tense variants
that change a vowel in the stem: chose is analyzed as choose-pst, threw as throw-pst, and
took as take-pst.
The past participle inflectional suffix, which we have labeled -en, has a familiar
variant pronounced "en" or "n" as in taken, broken, written, gone. Notice that broken,
written, gone, and the words that accompany them in 2.23e change the pronunciation of a
vowel in their stem when the -en inflection is affixed. However, even though the most
frequently used verbs of English call for an "en" or "n" allomorph, most English verbs
pronounce -en exactly like the past tense inflection. Notice that the first three lines of
example words in 2.23e are the same as the first three lines in 2.23d. We may ask what
determines whether a given occurrence of stayed, or walked, or made is to be analyzed as
stay-pst or stay-en, walk-pst or walk-en, make-pst or make-en. In a later chapter, we shall
examine detailed criteria for distinguishing these two inflectional suffixes. Here, I wish
only to demonstrate that the distinction is a real one and to discuss a simple analytical
procedure that can help you determine which of the two inflectional suffixes occurs in a
given word. Notice that in He sighted a UFO the "ed" of sighted clearly means "past",
24
WORDS: THE BUILDING BLOCKS OF GRAMMAR
but in UFOs are sighted every week nothing about the sentence means "past". Similarly,
the "ed" of lived in He lived there last year is clearly "past", but the sentence He has lived
there for two years implies that he is still living there, and is thus as much a comment on
the present as on the past; we are not likely to say or hear a sentence like He has lived
there last year, and the reason is of course that the "ed" of lived in this sentence is not a
variant of the past tense inflection but of -en.
Here is a procedure that can help you decide whether a given verb stem ending in the
letters ed has the -pst or -en morpheme affixed. Find a verb like know, see, choose, take,
break, write, or go (whose pronunciation of -en is clearly distinct from -pst) and
substitute it for the word in question. If, for instance, go were substituted for live in He
has lived there, it would immediately be clear that He has gone there not *He has went
there is appropriate, and therefore that lived is to be analyzed as live-en.
The present participle inflectional suffix, which I have labeled -ing, presents few
difficulties. Aside from the fact that the final "g" sound is sometimes dropped in
conversation, it really has no more than one allomorph.
The comparative and superlative inflectional suffixes, which I have labeled -er and est respectively, have only a few allomorphic variations. When -er is affixed to the stem
good it requires an allomorph of the stem, spelled bett, to appear, and when -est is
affixed, yet another allomorph, spelled simply b, appears. But as with the other
inflectional suffixes, we will focus on the abstract and analyze the morphemes of better
and best as good-er and good-est respectively. Worse and worst would be analyzed
similarly as bad-er and bad-est.
PRACTICE 3 (WORKING WITH ENGLISH INFLECTIONAL SUFFIXES)
Printed below is a passage by Henry David Thoreau with forty words italicized. Each
italicized word ends in an inflectional suffix (sometimes with a zero allomorph). Copy the words
onto a sheet of paper, and beside each word write the label of the inflectional suffix that is part of
the word (-es, -poss, -prs ,-pst, -en, -ing, -er, -est).
At a certain season of our life we are accustomed to consider every spot as the possible site of a
house. I have thus surveyed the country on every side within a dozen miles of where I live. In
imagination I have bought all the farms in succession, for all were to be bought, and I knew their
price. I walked over each farmer’s premises, tasted his wild apples, discoursed on husbandry
with him, took his farm at his price, at any price, mortgaging it to him in my mind; even put a
higher price on it, -- took his word for his deed, for I dearly love to talk, -- cultivated it, and him
too to some extent, I trust, and withdrew when I had enjoyed it long enough, leaving him to carry
it on. This experience entitled me to be regarded as a sort of real-estate broker by my friends.
Wherever I sat, there I might live, and the landscape radiated from me accordingly. What is a
house but a sedes, a seat? -- better if a country seat. I discovered many a site for a house not
likely to be soon improved, which some might have thought too far from the village, but to my
eyes the village was too far from it. Well, there I might live, I said; and there I did live, for an
hour, a summer and a winter life; saw how I could let the years run off, buffet the winter
through, and see the spring come in. The future inhabitants of this region, wherever they may
place their houses, may be sure that they have been anticipated. An afternoon sufficed to lay out
the land into orchard, woodlot, and pasture, and to decide what fine oaks or pines should be left to
stand before the door, and whence each blasted tree could be seen to the best advantage; and then
25
WORDS: THE BUILDING BLOCKS OF GRAMMAR
I let it lie, fallow perchance, for a man is rich in proportion to the number of things which he can
afford to let alone.
FEEDBACK TO PRACTICE 3
(WORKING WITH ENGLISH INFLECTIONAL SUFFIXES)
Below is a reprint of the passage presented for analysis in Practice 4. Following each
italicized word (in parentheses and boldface type) is the label of the inflectional suffix that is
attached to the word.
At a certain season of our life we are (-prs) accustomed (-en) to consider every spot as the
possible site of a house. I have (-prs) thus surveyed the country on every side within a dozen
miles of where I live. In imagination I have bought (-en) all the farms (-es) in succession, for all
were to be bought (-en), and I knew (-pst) their price. I walked over each farmer’s (-poss)
premises, tasted (-pst) his wild apples (-es), discoursed on husbandry with him, took (-pst) his
farm at his price, at any price, mortgaging (-ing) it to him in my mind; even put a higher (-er)
price on it, -- took his word for his deed, for I dearly love (-prs) to talk, -- cultivated it, and him
too to some extent, I trust, and withdrew (-pst) when I had (-pst) enjoyed (-en) it long enough,
leaving (-ing) him to carry it on. This experience entitled me to be regarded (-en) as a sort of
real-estate broker by my friends (-es). Wherever I sat (-pst), there I might live, and the landscape
radiated (-pst) from me accordingly. What is a house but a sedes, a seat? -- better (-er) if a
country seat. I discovered (-pst) many a site for a house not likely to be soon improved (-en),
which some might have thought (-en) too far from the village, but to my eyes the village was too
far from it. Well, there I might live, I said (-pst); and there I did (-pst) live, for an hour, a
summer and a winter life; saw (-pst) how I could (-pst) let the years (-es) run off, buffet the
winter through, and see the spring come in. The future inhabitants (-es) of this region, wherever
they may (-prs) place their houses, may be sure that they have (-prs) been (-en) anticipated (-en).
An afternoon sufficed (-pst) to lay out the land into orchard, woodlot, and pasture, and to decide
what fine oaks (-es) or pines should be left to stand before the door, and whence each blasted tree
could be seen to the best advantage; and then I let (-pst) it lie, fallow perchance, for a man is (prs) rich in proportion to the number of things which he can afford to let alone.
COMPOUND WORDS
We have seen that many English words have more than one morpheme, but in all the
examples given so far, all words of more than one morpheme have had only one root,
even if combined with assortments of prefixes and suffixes. It is also possible in English
for one word to have more than one root, as in words like classroom or blackboard.
There are in fact hundreds, perhaps thousands, of such compound words in English.
Many of them are well established and entered in dictionaries. Many are recent creations
of productive compounding processes. There are really no patterns governing whether
compounds are written as one word, or as two words, or with a hyphen. The stress
pattern is the surest indicator of compounding: the main stressed syllable in the first
word of a compound will be louder than the main stressed syllable in any following
words. Here are some examples of compounds in which at least one of the compounded
words is a noun and where the whole compound functions like a noun, i.e., typically as
HEAD of a noun phrase:
26
WORDS: THE BUILDING BLOCKS OF GRAMMAR
2.24a
Type: noun + noun (sunrise < "The sun rises.")
bee-sting, catcall, daybreak, earthquake, frostbite, headache, heartbeat, landslide, nightfall,
rainfall, sound change, toothache
2.24b
Type: verb + noun (rattlesnake < "The snake rattles.")
crybaby, driftwood, flashlight, glow worm, hangman, playboy, popcorn, stink weed, tugboat,
turntable, watchdog
2.24c
Type: noun + noun (blood test < "X tests blood.")
birth-control, book review, crime report, dress-design, haircut, handshake, meat delivery, office
management, suicide attempt, self-control, self-destruction, tax cut, word formation
2.24d
Type: noun + noun-er (babysitter < "X sits with the baby")
city dweller, factory-worker, gate-crasher, housebreaker, playgoer, sun-bather, theatergoer,
daydreamer
2.24e
Type: noun + noun (doorknob < "The door has a knob")
arrowhead, bedpost, cartwheel, piano keys, shirt-sleeves, table leg, telephone receiver, television
screen, window-pane
2.24f
(These fit the previous 5 types, but the meanings are not literally related to the components of the
compound.)
birdbrain, blockhead, butterfingers, egghead, fathead, feather brain, hardhat, hardtop, hunchback,
loudmouth, pale face, paperback, pot-belly, scarecrow
Here are two types that produce compound past participles and one that produces a compound
adjective; all three can function as MODIFIERS in noun phrases:
2.25a
Type: noun + past participle (heartfelt < "X feels it in the heart.")
airborne, custom-built, handmade, home-made, suntanned,
typewritten, thunder-struck,
weather-beaten
2.25b
Type: adjective/adverb + past participle (quick frozen < "X was frozen quickly.")
dry-cleaned, far-fetched, fresh-baked, long-awaited, well-meant, widespread
2.25c
Type: noun + adjective (foot sore < "sore in respect of one's foot")
airsick, air-tight, camera-ready, carsick, dust proof, duty-free, fireproof, foolproof, homesick,
oven-ready, tax-free, war-weary
27
WORDS: THE BUILDING BLOCKS OF GRAMMAR
PRACTICE 4 (Labeling Morpheme Types)
The best way to acquire full and lasting understanding of linguistic categories is to apply
them to the analysis of texts. In principle, you should be able to take any text at random -- a
paragraph from a favorite book, a favorite poem, or even a page from this book -- and label all of
the morphemes in all of the words as examples of the various types of morphemes we have
discussed in this chapter. In fact, this is not so easy to do, because it is often very difficult to
decide about morpheme divisions in a word. So, instead of trying to label all of the morphemes
in the words of some random text, why don't you try to label just some morphemes that I have
underlined in a text that I have selected. The next paragraph provides the directions. The text,
with numbered lines for reference, follows the directions.
Determine which of the following morpheme types are represented by the underlined parts of
certain words in the following selection: BR (bound root), FR (free root), P (prefix), D
(derivational suffix), -es (plural), -poss (possessive), -prs (present tense), -pst (past tense), -en
(past participle), -ing (present participle), -er (comparative), -est (superlative). Then, on a
separate sheet of paper, list the line number of each word, copy the word beside it, and then write
the appropriate label from the list above immediately after the word.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
In the fall of 1924 Thomas Wolfe, fresh from his courses in play writing
at Harvard, joined the eight or ten of us who were teaching freshman
composition in New York University. I had never before seen a man so tall
as he, so beady-eyed and so ungraceful. I pitied him and went out of my
way to help him get adjusted to his work and to make him feel at home.
His students soon let me know that he had no need of my protectiveness.
They spoke of his ability to narrate a simple event in such a manner
as to have them roaring with laughter or struggling to keep back their
tears, of his readiness to quote at length from any poet they could name,
of his habit of writing three pages of commentary, on a one page theme,
and of his astonishing ease in expressing in words anything he had seen
or heard or tasted or felt.
Indeed, his students made so much of his powers of observation that
I decided to stage a little test and see for myself. My opportunity came
one morning when the students were slowly gathering for nine-o'clock
classes.
Upon arriving at the university that day, I found Wolfe alone in
the large room which served all the freshman composition teachers as an
office. He made no protest when I asked him to come with me out into
the hall, and he only smiled when we reached a classroom door and I bade
him enter alone and look around.
He stepped in, remained no more than thirty seconds, and came out.
"Tell me what I see," I said as I took his place in the room, leaving
him in the hall with his back to the door. Without the least hesitation
and without a single error, he gave the number of seats in the room,
designated those which were occupied by boys and those occupied by girls,
named the colors each student was wearing, pointed out the Latin verb
conjugated on the blackboard, spoke of the chalk marks which the charwoman had failed to wash from the floor, and pictured in detail the
view of Washington Square from the windows.
As I rejoined Wolfe, I was speechless with amazement. He, on the
contrary, was wholly calm as he said, "The worst thing about it is that
28
WORDS: THE BUILDING BLOCKS OF GRAMMAR
34
35
36
37
38
I'll remember it all."
I felt no surprise whatsoever when, five years after that revealing
experience I read Wolfe's first novel, LOOK HOMEWARD ANGEL, and
recognized it as perhaps the richest compendium of actually remembered sense
impressions which any author had ever committed to writing.
PRACTICE 5 (Labeling Morpheme Types II)
Here is another text, where the analytical task is the same as in Practice (1): Determine which
of the following morpheme types are represented by the underlined parts of certain words in the
following selection: BR (bound root), FR (free root), P (prefix), D (derivational suffix), -es
(plural), -poss (possessive), -prs (present tense), -pst (past tense), -en (past participle), -ing
(present participle), -er (comparative), -est (superlative). Then, on a separate sheet of paper, list
the line number of each word, copy the word beside it, and then write the appropriate label from
the list above immediately after the word.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Though the truth may not be felt or generally acknowledged for generations
to come, the only school of genuine moral sentiment is society between equals.
The moral education of mankind has hitherto emanated chiefly from the law of
force, and is adapted almost solely to the relations which force creates. In the
less advanced states of society, people hardly recognize any relation with their
equals. To be an equal is to be an enemy. Society, from its highest place to its
lowest, is one long chain, or rather ladder, where every individual is either
above or below his nearest neighbour, and wherever he does not command he
must obey. Existing moralities, accordingly, are mainly fitted to a relation of
command and obedience. Yet command and obedience are but unfortunate
necessities of human life: society in equality is its normal state. Already in
modern life, and more and more as it progressively improves, command and
obedience become exceptional facts in life, equal association its general rule.
The morality of the first ages rested on the obligation to submit to power; that
of the ages next following on the right of the weak to the forebearance and
protection of the strong. How much longer is one form of life to content itself
with the morality made for another? We have had the morality of submission,
and the morality of chivalry and generosity; the time is now come for the
morality of justice.
FEEDBACK TO PRACTICE 4 (Labeling Morpheme Types)
Here is the complete and correct analysis of the text in Practice (1) laid out as specified in the
directions for Practice (1). You should of course have worked out your own analyses before
consulting the answers given below.
(line)
(word)
(answer)
(line)
(word)
(answer)
1
2
3
3
4
courses
joined
composition
had
beady-eyed
-es
-pst
FR
-pst
D
1
2
3
3
4
writing
teaching
University
seen
ungraceful
FR
-ing
D
-en
P, D
29
WORDS: THE BUILDING BLOCKS OF GRAMMAR
4
6
6-7
9
9
10
11
12
12
13
13
15
17
20
21
24
27
29
30
32
32
33
36
38
pitied
students
protectiveness
roaring
struggling
readiness
writing
astonishing
words
heard
felt
decided
classes
protest
classroom
leaving
designated
conjugated
had
rejoiced
amazement
worst
experience
impressions
FR
-es
P, D
-ing
FR
FR
FR
D
-es
-en
FR
-pst
-es
FR
FR
FR
FR
BR
-pst
P
D
-est
P
P, D
5
6
8
9
10
10
11
12
12
13
14
16
18
20
23
25
28
29
30
32
33
35
37
38
adjusted
had
ability
laughter
tears
length
pages
expressing
seen
tasted
powers
slowly
alone
asked
stepped
hesitation
wearing
blackboard
failed
speechless
wholly
felt
richest
committed
P
FR
FR
D
-es
D
-es
FR
-en
-en
-es
D
P
-pst
-pst
D
-ing
FR
-en
D
FR
-pst
-est
BR
FEEDBACK TO PRACTICE 5 (Labeling Morpheme Types II)
Here is the complete and correct analysis of the text in Practice (2) laid out as specified in the
directions. You should of course have worked out your own analyses before consulting the
answers given below.
(line)
(word)
(answer)
(line)
(word)
(answer)
1
3
5
7
9
10
13
15
17
18
truth
emanated
recognize
lowest
mainly
unfortunate
exceptional
forbearance
had
morality
FR
-en
BR
-est
D
P
P
FR
-en
D
1
4
6
8
9
12
14
16
17
18
acknowledged
adapted
highest
does
fitted
progressively
rested
protection
submission
generosity
-en
-en
-est
-prs
-en
BR
-pst
BR
P
D
30