Andrew Radford - Linguistics An Introduction [Second Edition]

Incorporation is found in a large number of language groups; many languages

of the Americas, such as the Iroquoian languages, the Mayan languages, Nahuatl (the language of the Aztecs), large numbers of languages of the Pacific including Maori, Samoan and Tongan, a number of Australian languages, certain of the

languages of India and a host of others exhibit incorporation.

What are referred to as polysynthetic languages are those that make use of

incorporation in their morphology, though they may also have agglutinating or inflectional processes, or even show isolating tendencies. Chukchee, for instance, is typical in having a large number of very regular derivational processes, which are relatively agglutinating, just like Turkish. However, it also has a rich inflectional system showing cumulation, extended exponence, syncretism and so on

(exercise 4).

162

words

Incorporation processes like those described above strike us as ‘exotic’. It is noteworthy, then, that a similar phenomenon is found with a very common type of compound in English. This is illustrated in (128):

(128) a.

Tom drives taxis

b.

Tom is a taxi-driver

The compound in (128b) includes the object of the verb drive from which the deverbal noun driver is derived. Similar examples are taxi-driving, insect repellent and motorcycle maintenance. In these compounds, the head is derived from a verb (drive, repel, maintain). The non-head of the compound functions effectively as the object of the verb (see drive taxis, repel insects, maintain motorcycles). This is referred to as synthetic compounding. If it were possible to form a verb from these, as in (129), we would have proper noun incorporation in English: (129) a.

*Tom taxidrove yesterday

b.

*Agent Orange insectrepels very effectively

c.

*Bikers should motorcyclemaintain regularly

Even where it looks as though we have such a case, as in Dick babysat for Tom and Harriet, we generally find that there is no syntactic (analytic) equivalent in which the object and the verb are separated: *Dick sat the baby for Tom and Harriet. The verb babysit is just an idiosyncratic form, not a regular compound, and we are justified in concluding that English does not exhibit proper incorporation.

Types of morphological operations

We have already seen numerous examples of prefixation and suffixa-

tion, and the examples of vowel changes and suppletions, as in English past tense forms sang and brought, have indicated that there are additional ways in which the morphological structure of a word can be modified. The Chukchee example in

(125) provides another case, where the root teŋ in the word nə-teŋ-qin is simultaneously prefixed and suffixed to form the adjective. A similar phenomenon is seen in German. In regular verbs, the perfect/passive participle is formed by simultaneously adding a prefix ge- and a suffix -t to the verb stem. Thus, from the stem hab ‘have’ we get ge-hab-t ‘had’. Since the prefix and suffix are added together, we can think of nə-…-qin, or ge-…-t as a composite, discontinuous morpheme.

Such a morpheme is called a confix or circumfix.

The languages of the Philippines illustrate another type of affixation. Here are some verb forms in the major language of those islands, Tagalog:

(130)

verb stem

infinitive

meaning

a.

aral

umaral

‘teach’

b.

sulat

sumulat

‘write’

c.

basa

bumasa

‘read’

d.

gradwet

grumadwet

‘graduate’

Morphology across languages

163

The crucial thing about these examples is that aral, sulat, basa and gradwet are single, undecomposable morphemes. In (130a), we see the prefix um- added to a vowel-initial stem. However, (130b, c, d) do not have the infinitive forms *umsu-lat, *umbasa, *umgradwet. Rather, when the stem begins with a consonant, the affix goes inside the stem morpheme, after the onset of the first syllable. This is a regular and pervasive process in Tagalog and several hundred related languages, as can be seen from the fact that it applies to the recent English loan word from graduate (130d). An affix which is inserted strictly inside another affix or stem like this is known as an infix.

Prefixes and suffixes (and circumfixes) behave like things which are added

to stems. This is like compounding in that we simply concatenate two entities, and, indeed, such affixation often develops historically from compounding.

Morphology of this type is called concatenative, and it encourages the view, briefly discussed in section 10, that complex word forms consist simply of strings of morphemes. However, very often a morphological process seems to be realised by a phonological operation performed on the stem itself, as in the case of the vowel changes in sing ~ sung ~ sang. Indeed, infixation can be construed in this way as involving first affixation, then a phonological operation which moves the affix to a position inside the stem. It should also be clear that infixation represents another type of deviation from strict agglutination.

Tagalog illustrates a further way in which affixation looks more like a process than a straightforward concatenation of morphemes. Here are some more verb

forms in this language:

(131)

verb stem

future

meaning

a.

sulat

susulat

‘write’

b.

basa

babasa

‘read’

c.

trabaho

tatrabaho

‘work’

From (131) we can see that the future tense form of the verb involves taking the first syllable and copying the first consonant from its onset and its vowel to create a new syllable which appears as a prefix. This type of process is known as

reduplication, and it provides a rather vivid demonstration of the inappropriateness of suggesting that Tagalog has a morpheme FUTURE with various lexically conditioned allomorphs. Obviously, the list of such allomorphs would be rather long and such a list would fail to make explicit the fundamental fact about Tagalog future formation. This fact is acknowledged by suggesting that there is a morphological feature, say [+future], which can attach to verb lexemes. When this

happens, a phonological process is triggered which produces the correct future form of the verb by consulting the syllable structure of the stem form and

performing the appropriate operations (exercise 5).

On several occasions, in this and the previous section, we have invoked examples of vowel changes in English verb forms as another type of phonological operation which subserves a morphological purpose. Alongside sing ~ sang ~ sung, we find

164

words

ring ~ rang ~ rung, hang ~ hung, fling ~ flung, etc., and it is now time to introduce the technical term for this sort of process. It is known as ablaut (sometimes called apophony). A larger number of English verbs combine a vowel change

with suffixation, especially in the participle, so we find such sets of forms as the following: write ~ wrote ~ written, give ~ gave ~ given, take ~ took ~ taken, do ~

did ~ done. Each of these simply involves a vowel change in forming the past tense form (the second member of each set); for the participles (the third member of each set), however, there is suffixation of -en with or without a vowel change.

A specific kind of ablaut, which is particularly common in Germanic languages (and a number of other language groups), occurs when a back vowel is replaced by a front vowel. A number of German plurals are formed this way: /apfl ~ epfl/

‘apple’, /fogl ~ føgl/ ‘bird’, /brudr ~ brydr/ ‘brother’. This type of vowel fronting is known as umlaut, and there are vestiges of this in English irregular plurals such as men, teeth and geese.