Linguistics 1A Morphology

2 Complex words
In the previous lecture we noted that words can be classified into different categories,
such as verbs, nouns, adjectives, prepositions, determiners, and so on. We can make
another distinction between word types as well, a distinction that cuts across these
categories. Consider the verbs, nouns and adjectives in (1)-(3), respectively. It will
probably be intuitively clear that the words in the (b) examples are complex in a way
that the words in the (a) examples are not, and not just because the words in the (b)
examples are, on the whole, longer.

(1) a. to walk, to dance, to laugh, to kiss

b. to purify, to enlarge, to industrialize, to head-hunt

(2) a. house, corner, zebra

b. collection, builder, sea horse

(3) c. green, old, sick

d. regional, washable, honey-sweet

The words in the (a) examples in (1)-(3) do not have any internal structure. It does not
seem to make much sense to say that walk, for example, consists of the smaller parts
wa and lk. But for the words in the (b) examples this is different. These are built up
from smaller parts that each contribute their own distinct bit of meaning to the whole.
For example, builder consists of the verbal part build with its associated meaning, and
the part –er that contributes a ‘doer’ reading, just as it does in kill-er, sell-er, doubt-er,
and so on. Similarly, washable consists of wash and a part –able that contributes a
meaning aspect that might be described loosely as ‘can be done’, as it does in
refundable, testable, verifiable etc.

The smallest parts of words that add their own distinct meaning component to the
word are called morphemes. Thus, the (a) examples in (1) to (3) show that some
words consist of just one morpheme. Such words are called simple or simplex. Other
words consist of two or more morphemes. These are called complex words.

The definition of morphemes as the smallest meaning-bearing units of words is

actually not entirely unproblematic. In some cases, when we divide a word into
smaller parts because we detect parts that add their own distinct meaning component,
we are left with another bit that appears not to have any independent meaning.
Consider the following examples:

(4) blueberry
blackberry
gooseberry
cranberry

It seems clear that we want to say that these words all share the same morpheme berry
on their right-hand side, with the same meaning ‘a small round fruit on particular
plants and trees’. In the first cases, that leaves us with morphemes on the left-hand
side that also have an independent meaning (although in the case of goose it is unclear
what the relation is of this meaning to the meaning of the whole). In the case of
cranberry, however, this leaves us with the morpheme cran as left-hand part, which is
not a meaningful morpheme in (current) English. Such meaningless morphemes are
called cranberry morphemes, after this particular example (note that it is cran, not
cranberry, that is the cranberry morpheme in this case).

For the large majority of cases, however, defining morphemes as the smallest
meaningful units into which we can divide words works well. Not all morphemes are
of the same type. We have already seen that words can consist of just one morpheme.
Put differently, it is possible for a morpheme to constitute a word in its own right.
Examples are the words in (1a), (2a) and (3a), all consisting of just a single
morpheme. But not all morphemes can do this. Consider the following examples:

(5) a. Q: Is that green?

A: *No, at best it’s –ish.
b. Q: Is she any good at football?
A: *Yes, she’s a great –er.
c. Q: Does he play the piano often?
A: *Yes, he –s it all the time.

In contrast to morphemes like horse or green, morphemes such as -ish, -er or –s

cannot stand on their own as a complete word. They have to attach to another
morpheme and form a complex word with that.

When a morpheme can occur as a word on its own we speak of a free morpheme. In
contrast, morphemes that obligatorily attach to another morpheme are bound
morphemes, also called affixes. There are affixes that must attach to the right of their
host, the so-called suffixes:

(6) green-ish, build-er, wash-able, solid-ify, industry-al-ize

And there are affixes that must attach to the left of their host, the so-called prefixes:

(7) en-large, re-en-act, de-throne, ex-minister

Though much more exceptional, an affix can also consist of two parts, with one part
attaching to the left of the word the affix combines with and one part attaching to the
right. An example of such a circumfix is the affix that forms the past participle form
of verbs in Dutch. Note that in (8) the ge- and –d parts really do form a single affix –
you would not get the past participle of the verb if you left one of them out, while at
the same time neither of them does anything else but make the past participle form of
the verb.

(8) speel ge-speel-d leef ge-leef-d huur ge-huur-d

play played live lived hire hired

A very curious type of affix is the infix, which can go right into the word with which
it combines, rather than attaching to an edge of it. An example of an infix is the verb-
forming affix um in the Philippine language Tagalog:
(9) base: labas verb: l-um-abas ‘come out’
base: pasok verb: p-um-asok ‘enter’
base: bili verb: b-um-ili ‘buy’

Let us now take a closer look at how a complex word can be built up out of two or
more morphemes. There are two basic possibilities. The first is to combine two free
morphemes. Although free morphemes can form a simplex word on their own, it is
possible to combine two (or more) of them to create a complex word. This process is
termed compounding; the resultant words are compounds. Here are some examples:

(10) sea + horse → sea horse

head + strong → headstrong
hand + made → handmade

(Don’t be misled by the space that English orthography puts between sea and horse in
sea horse: a sea horse behaves just like any other noun, so it is a compound noun).

Some types of compounding are very productive indeed in English, others less so. We
will come back to this in the next lecture. For now, note that it is possible to take any
two simplex nouns in English and form a compound out of them. Some of these will
be existing words in the English lexicon, such as sea horse. Others, such as the ones
in (11), are not existing words, but these are nevertheless all possible words of
English. Their absence from the current English lexicon reflects that there is no
current need to use them. However, they are allowed by the principles of English
morphology.

(11) word + kitchen → word kitchen (e.g. a kitchen for morphologists)

tree + laptop → tree laptop (e.g. a laptop that has a tree painted on it)
gorilla + window → gorilla window (e.g. a window through which you
can see a gorilla)
window + gorilla → window gorilla (e.g. a gorilla that always sits
behind a window)

The second type of complex word, next to compounds, involves affixation. That is, it
involves the combination of a bound morpheme with a free morpheme, or with a
complex word. This type comes in two subtypes.

In one type of affixation process, the affix makes a whole new word out of the word it
attaches to, with a different meaning and often also of a different lexical category:

(12) buildVERB → build-erNOUN

largeADJECTIVE → en-largeVERB
industryNOUN → industry-alADJECTIVE → industry-al-izeVERB

This process of affixation is called derivation; the affixes involved hence are
derivational affixes.
Another type of affix does not make a completely new word out of the word it
attaches to. Rather, it expresses some grammatical information about that word, such
as for example the person and number of a finite verb, or whether a noun is plural or
not. Such affixes consequently never change the lexical category of the word they
attach to. This process of affixation is called inflection; the affixes involved are
inflectional affixes. Some examples of inflection are the –s, –ed and –s suffixes in the
following examples:

(13) workVERB-s (expresses that the verb has a third person singular subject)
workVERB-ed (expresses that the event expressed by the verb took place
in the past)
houseNOUN-s (expresses that we are dealing with more than one house)

English has very little inflectional morphology. There are languages in which it is
ubiquitous (as we will see in a later lecture).

The distinction between derivational affixes and inflectional affixes is not always
clear-cut. What about the English prefix re- for example? This does not change the
category of the word it attaches to: rebuild is a verb just like build is, and rematch is a
noun just like match. It does add a distinct meaning component to the word, however,
that of repetition. The question then is whether we want to say this is grammatical
information about the word it attaches to (its ‘repetitive aspect’ for instance) or
whether we want to say it is really a new word we are dealing with. Presumably we
want to say that re- is derivational, because of another aspect that distinguishes
inflectional morphology from derivational morphology. This is that in some syntactic
contexts it is obligatory to put inflectional affixes on certain words (to inflect those
words). There are no syntactic contexts in which only words containing derivational
morphology can show up. Since there are no syntactic contexts in which it is
obligatory to put the prefix re- on a verb we may conclude that this is a derivational
affix.

Let us consider one case in which inflectional affixation is required in English. It was
stated above that free morphemes can stand on their own. In that context, verbs such
as dance and walk were mentioned. In certain types of sentences, however, namely
those with a third person singular subject, these forms, despite being free morphemes,
cannot show up on their own. The verb must be inflected with the –s that expresses
the person and number of the subject:

(14) a. *Sylvia usually walk to work.

b. Sylvia usually walks to work.

The form of the verb stripped of its inflection, that is, the seemingly free morpheme
that nevertheless requires inflection in particular syntactic contexts, is called its stem.
Nouns and adjectives sometimes must carry inflectional morphology in particular
syntactic contexts as well, certainly in some languages other than English. We will
see later what kind of inflection that is.

Let us return to derivation now. Recall that it is possible to derive a new word, usually
of a different lexical category, by adding a derivational affix to a word. Interestingly,
it seems that sometimes it is possible to do exactly the same without adding any affix.
Consider the following examples:

(15) to build – a build-er

to dance – a danc-er
to run – a run-er
to kill – a kill-er
to cook – a cook

These examples show that it is possible to take a verb that implies some ‘doer’ and
derive a noun from it that refers to this ‘doer’ by adding the suffix –er. In the last
example, however, there is precisely the same relationship between the verb to cook
and the noun cook as in the other examples, but in this case no suffix –er shows up.
Cases in which one word appears to be derived from another without there being any
derivational morphology are known as zero derivation or conversion. The term ‘zero
derivation’ indicates what some people think is going on in such cases: there really is
a derivational affix present, but it happens not have any phonological content. It is a
null affix. Conversion can be very common in some languages. In English, there are
quite a few noun-verb pairs that are related by conversion, and conversion between
some other lexical categories is possible too:

(16) to runVERB – a runNOUN

yellowADJECTIVE – to yellowVERB
downPREPOSITION – to downVERB
greenADJECTIVE – a greenNOUN

Null affixes occur in the inflectional domain as well (but here the phenomenon is not
referred to as ‘conversion’, since inflectional affixes do not derive a new word). It was
noted above that in English a verb needs to carry a third person singular suffix if there
is a third person singular subject present. So, inflecting a verb in sentences with a
subject appears to be obligatory. However, if there is a first person or second person
subject, it seems as if this requirement does not hold any longer:

(17) I/You usually go- to the market on Saturdays.

If inflection is required in sentences with a third person subject, it is rather unexpected

that this requirement is voided when the subject is first or second person. At least, we
would have to think of a reason for why that should be so. Perhaps it is more likely
that (17) indicates instead that the inflectional affixes for first and second person
verbs in English happen to be null affixes.

It is certainly not always easy to decide whether a particular affix is just absent from a
language or whether it is null affix. However, when there are overt affixes for a
particular inflectional category in a language, such as for ‘person’ or ‘tense’ in
English, it seems to make sense to assume that, for those values of this category that
are not expressed by a visible affix, we are dealing with a null affix. When there are
no affixes at all for a particular category, however, it may well be that that inflectional
category is absent from the language altogether. For example, many languages have
verbal inflectional affixes that express the so-called ‘mood’ of a verb, which reflects
the speaker’s attitude towards the factuality of the event expressed by the verb -
whether it is to be taken as a possibility, something that is not the case, something that
may be the case in the future, a wish, an order, etc. In English, there is no inflection to
express these categories. Does this mean that English has a battery of null affixes
expressing all these categories? It seems more economical to assume that English
simply does not have inflection for mood. This is not just a more economical
assumption, but it may also account for the fact that English uses different means to
express verbal mood, namely by certain auxiliary verbs such as can and will. In
general, auxiliary verbs in one language fulfil the same function as morphological
inflection in another language.

We have seen the various ways in which a complex word can be made now: by
compounding, by derivation or by inflection. These processes can be combined, or
reiterated, so that a complex word consisting of three or more morphemes results. For
example, we can derive the verb teach with the suffix –er, resulting in the noun
teacher, which can be compounded with school to get school teacher, which can be
pluralized with the inflectional suffix –s: school teachers. The internal structure of
such complex words can be transparently represented by a so-called tree diagram.
The tree diagram for school teachers looks as follows:

(18) N

N plural

N N s

school V N
 
teach er

A tree diagram should be read as follows. The top node in the tree (the uppermost N
in (18)) branches into two other nodes (another N and ‘plural’). This means that the
whole word (represented by the upper node) is a noun that consists of two smaller
parts, another noun and a plural morpheme. The ‘plural’ node does not branch any
further, which means it has no further internal structure. The left-hand node, however,
branches further, namely into two other N nodes. This means that the part of the word
that that node represents (the part dominated by that node, so the school teacher part
in (18)) consists in turn of two smaller parts, both of which are a noun. The left-hand
part does not branch any further, so has no further internal structure (indeed, school is
not morphologically complex). But the right-hand part branches still further, into a V
node and an N node. This expresses that the N node dominating the teacher bit of the
structure is internally structured, consisting of a verb (teach) as its left-hand part and
an affix that makes a noun (-er) as its right-hand part. Thus, a tree diagram is a visual
means to indicate what the internal structure is of complex words. (In the syntax
module you will see that tree diagrams can be used in the same way to visualise the
internal structure of complex phrases and sentences. Note that the ‘tree’ metaphor for
such diagrams is used in such a way that the tree is actually upside down, the top node
being the root of the tree, and the tree branching out downwards).
Some complex words can be structurally ambiguous. This means that they have two
different meanings, which correspond to a different internal structuring of the word.
An example is the word ex-president-office-manager. This can mean ‘manager of the
office of an ex-president’ or ‘former manager of a president’s office ’:

(19) a. This ex-president-office-manager works for Bill Clinton.

b. This ex-president-office-manager is now a street artist.

These different meanings correspond to the two structures expressed by the tree
diagrams in (20a) and (20b), respectively. The difference between the two concerns
the hierarchical position that the prefix ex- occupies.

(20) a. N

N N

N N manager

ex N office

president

b. N

ex N

N N

N N manager
 
president office