Inflection and Derivation

So far we have distinguished 2 principal types of morphemes: bases, like [rat], and affixes, which are either prefixes, like [re-], or suffixes, like [-es]. Before we can proceed to the identification of words, which is the ultimate goal of morphemics, we must look a bit more closely at the various types of affixes and the ways in which they occur. [...]

[We can differentiate between] two types of suffixes, a distinction that will be of considerable importance in our discussion of words, as well as when we come to discuss grammar. ... These suffixes which must always come at the end of the morpheme groups to which they belong we will call inflectional suffixes. Those which may be followed by other suffixes we will call derivational suffixes. We can make a similar distinction between the types of paradigms in which these suffixes take part. Thus a paradigm like [әgri: – әgri:d], the second form of which consists of the stem plus the inflectional suffix [-ed], can be called an inflectional paradigm, and the form [әgri:d] can be called an inflected form of agree. On the other hand, the pair [әgri: – әgri:mәnt] illustrates a derivational paradigm, and the form [әgri:mәnt] is a derivative form or simply a derivative of agree.

The suffixes of present-day English can thus be divided into 2 groups, inflectional and derivational suffixes. No such distinction exists in the case of prefixes, however; they are all derivational. By means of prefix like [dis-], for instance, a whole new set of derivatives of agree can be made, corresponding to the derivatives already formed by adding suffixes. In turn, these new derivatives may add inflectional suffixes, so that we may get such forms as ‘disagreed’, ‘disagreements’, and ‘disagreeablenesses’. Since in adding suffixes all derivational ones must be added to the base before the final inflectional one, we assume the same of prefixes. That is, inflection takes place on a level of structure higher than that of derivation. What this comes to is that, in terms of our examples, we treat a form like ‘disagreements’ as consisting of [disәgri:mәnt] + [-es], rather than [dis-] + [әgri:mәnt]. Or, looking at it from the other direction, we may say that in analyzing linguistic forms into their constituent morphemes, we separate inflectional suffixes first, before we separate derivational prefixes or suffixes.

Bound Bases. If we study such combinations as ‘conclude’, ‘conceive’, and ‘consist’, we can observe that the stem of a derivative is not always a free form; it may be bound. Thus, by comparing ‘conclude’ with ‘occlude’, ‘preclude’, ‘include’, and ‘exclude’, we come to the conclusion that there is a morpheme [-klude], which serves as a stem for these various derivational forms. Yet we never find it as a free form; that is, we can find no environment into which [-klude] fits in ....

1. Bound morphemes are of 3 types: suffixes, prefixes, and bound bases.

2. Suffixes are either inflectional or derivational.

a) Inflectional suffixes are always final in the morpheme groups to which they belong. They are of wide occurrence, making large form-classes. Their distribution tends to be regular.

b) Derivational suffixes may be final in the morpheme groups to which they belong, or they may be followed by other derivational suffixes or by inflectional suffixes. They are of relatively limited occurrence, and their distribution tends to be arbitrary.

3. Prefixes are always derivational.

4. Bound bases are morphemes which serve as stems for deriva­tional forms but which never appear as free forms.

Ø What is W.N. Francis’s definition of a morph?

Ø In what does W.N. Francis see the difference between the two types of paradigms: inflectional and derivational?

Ø What proves that inflection is relevant for a level of structure higher than that of derivation?

Ø What types of bound morphemes does W.N. Francis identify?

 

Text 2

 

Nida E. Morphology

The Descriptive Analysis of Words. Introduction to Morphology

Morphology is the study of morphemes and their arrangements in forming words. Morphemes are the minimal meaningful units which may constitute words or parts of words, e.g. re-, de-, un-, -ish, -ly, -ceive, -mand, -tie, boy- and like- in the combinations receive, demand, untie, boyish, likely. The morpheme arrangements which are treated under the morphology of a language include all combinations that form words or parts of words. Combinations of words into phrases and sentences are treated under the syntax.

The Identification of Morphemes. Morphemes as Minimal Units

One of the first tasks which confront the linguist in examining a new language with a view to discovering and describing its structure is the identification of the minimal meaningful units of which the language is composed. These minimal units are called ‘morphemes’, and in many instances they are readily recognized. For example, in the English words boyish, maddening, condense, receive and up, we have little difficulty in identifying the various component units: boy-, -ish, mad-, -en, -ing, con-, -dense, re-, -ceive, and up. With practically no complications we have thus ‘broken down’ these longer expressions (i.e. words) into their constituent parts. The process by which this is accomplished appears at first to be almost instinctive, but if we consider closely what we have done, we recognize that we compared words, or at least drew upon our knowledge of such comparisons. In order to identify the morphemes we must have certain partially similar forms in which we can recognize recurring particles. What we need for comparison would be provided by the following series: boy, girlish, mad, fatten, fattening, constrain, density, return, deceive, start, up. These forms contain each of the morphemes in a different situation. By this means we compare and isolate, and it is only by such comparison with other forms that we can discover morphemes. ...

a) Determination of Allomorphs by Complementary Distribution

The plural forms of English nouns illustrate a number of points in allomorph identification. The predominant pattern of formation consists in the suffixation of [-әz - z - -s], but there are other ways of forming the plural. For example, the ox has in the plural the suffix ‘-en’. There is absolutely nothing in the phonological form of the stem ox to indicate that it does not take the regular plural suffixal set. A word such as box, which is phonologically similar, does take the suffix [-әz]. The only way in which we may know which words occur with which suffixes is to make a list, and the specific class for oxen contains just this one word. Since, however, the allomorphic set [-әz - z - -s] and the form [-әn] are in complementary distribution and have a common semantic distinctiveness (i.e. they are indicators of pluralization), we may combine all these forms as allomorphs. Some plural nouns do not differ in any overt way from the singular nouns, e.g. sheep, trout, elk, salmon, and grouse. For the sake of descriptive convenience we may say that these words occur with a zero suffix.

b) Basis of Complementation

The three types of plural formatives (1) [-әz - z - -s], (2) [-әn], and (3) [-0] (0 = zero) are all in complementary distribution. If they are combined as a single morpheme, then each of these forms constitutes an allomorph. Nevertheless, the relationships between these allomorphs are quite different because the basis of complementation is very different. The allomorphs [-әz - z - -s] are in complementation on the basis of phonological environment. This type of complementation we symbolize by ~. The complementation which exists between the three types of plural formatives is based upon the morphological environment. That is to say, we can describe the environment only by specifically identifying particular morphemes. This type of complementation we can symbolize by ∞. Accordingly, the series noted above may be written as [-әz ~ z ~ -s] ∞ [-әn] ∞ [-0].

d) Basic and Nonbasic Allomorphs

In treating phonologically defined allomorphs it is sometimes helpful to select a single form as phonologically basic, i.e. one from which the other allomorphs may be phonologically ‘derived’. For example, of the three allomorphs [-әz ~ z ~ - -s] we may set up [-әz] as phonologically basic. This is done in view of two types of data: (1) comparison with other similar series in English, e.g. [iz ~ z ~ s], atonic forms of ‘is’: Rose’s dead, Bill’s dead, Dick’s dead [-әz ~ z ~-s], atonic forms of ‘has’: Rose’s done it, Bill’s done it, Dick’s done it, and (2) congruence with general patterns of phonological change, by which we note that it is ‘phonologically simpler’ to explain or describe the loss of a phoneme than the addition of one. For the most part, however, we do not concern ourselves greatly about the rank of allomorphs as determined by their possible phonological relationships.

There is much greater value in determining the basic or nonbasic character of morphologically defined allomorphs. The basic allomorph is defined in terms of three characteristics: statistical predominance, productivity of new formations, and regularity of formation. An allomorph which occurs in more combinations than any other may generally be selected as being the basic form. A form which is statistically predominant is also likely to be productive of new combinations. For example, in English the so-called s-plural is productive of new plural formations, e.g. radios and videos. Whether a form is regular (i.e. consists of phonologically defined allomorphs) may also be a factor in determining its allomorphic rank.

The determining of the basic forms of a morpheme makes it possible to refer to the entire morpheme by a single allomorphic form. For example, in discussing the English plural formatives we may refer to the allomorphic series as a whole by using the symbols { } to enclose the basic allomorph, e.g. {-әz}. In many instances, there is no foundation for, or particular value in, attempting to set up a basic allomorph, but one may arbitrarily select a particular characteristic form of an allomorphic series and use it to refer to the entire series.

e) Types of Zero

When the structure of a series of related forms is such that there is a significant absence of a formal feature at some point or points in the series, we may describe such a significant absence as ‘zero’. For example, with the words sheep, trout, elk, salmon, and grouse, there is a significant (meaningful) absence of a plural suffix. We determine that there is an absence because the total structure is such as to make us ‘expect’ to find a suffix. This absence is meaningful, since the form with the absence (i.e. with zero) has a meaning which is different from the singular form, which has no such absence. A significant absence in an allomorphic series may be called an allomorphic zero.

Sometimes the general structure suggests a zero element. For example, in Totonac the subject pronouns are as follows:

k- first person singular -wi first person plural

-ti second person singular -tit second person plural

– third person singular -qu third person plural

The third person singular is never indicated overtly, i.e. it has no obvious form. The absence of some other form is what actually indicated the third singular. Structurally, this is a type of significant absence; it is not, however, an allomorph zero, but, rather, a morpheme zero. That is to say, this significant absence does not occur in a series of allomorphs, but in a series of morphemes. Both types of zeros are structurally and descriptively pertinent, but should be carefully distinguished (It is possible to say that in English the nouns have a zero morpheme for singular and {-әz} for plural. This would mean that sheep in the singular would have a morphemic zero and in the plural an allomorphic zero. One should, however, avoid the indiscriminate use of morphemic zeros. Otherwise the description of a language becomes unduly sprinkled with zeros merely for the sake of structural congruence and balance).