At various points in this book we have talked about how affixes are ordered with respect to each other. For example, we noted that inflectional affixes generally come outside of derivational affixes in the languages of the world. We also mentioned that among inflectional affixes, tense and aspect inflection tends to come closer to the stem than person and number inflection, the reason being that tense and aspect are more closely relevant to the meaning of the verb than person and number.

In this section, we will look more closely at the issue of how English derivational affixes are ordered with respect to each other, as this has been a matter of theoretical dispute for some time. The problem is this: if the only thing that constrained the ordering of affixes in English were the categorial restrictions on their attachment (for example, that -ness only attaches to adjectives, or that -ize attaches to both nouns and adjectives), we would expect to find many more combinations of affixes than we do find. In fact, we find very few of the potential combinations of affixes. The theoretical issue, then, is what restricts the combination of affixes, and how we explain those restrictions.

We have already touched upon this problem. In chapter 9 we noted that native derivational affixes in English typically come outside of non-native ones. One explanation that has been proposed for this generalization is that the two strata of English derivational morphology are ordered with respect to each other such that non-native affixation precedes native affixation. If the two strata are strictly ordered with respect to each other, then non-native affixes will never be able to attach outside of native ones. We might call this the Stratal Ordering Hypothesis.

There are two problems with this hypothesis, however. One is that it’s not quite accurate. There are non-native affixes that do not cause stress or phonological changes, like other non-native affixes, and that are perfectly happy attaching to native bases, for example -ee (standee), -ize (winterize), -able (singable). Occasionally it is even possible to attach a non-native affix to a word formed with a native suffix; the words softenable or whitenable, for example, have native -en followed by non-native -able. This has led theorists to lump -ee, -ize and -able in with native affixes, thus blurring the lines between the strata.

More seriously, if the only thing which constrained the ordering of derivational affixes in English were the ordering of the two strata, we would still expect to find many more combinations of affixes than we do. For example, the suffixes -age and -ize are both non-native. The suffix -age forms nouns, and -ize attaches to nouns, so -ize should attach to -age words. But we never find words with the combination -ageize, and words we might coin on the spot sound quite odd (orphanageize?, baggageize?). Similarly, -ify forms verbs, and non-native -ance forms nouns from verbs, but we never get nouns like purifiance. So another problem for the theory of stratal ordering is that the combinations of affixes within strata are more limited than the Stratal Ordering Hypothesis would lead us to expect.

How else might we constrain the ordering of affixes? One possibility that has been proposed (Plag 1999; Giegerich 1999) is that affixes cannot only select what they attach to (native or non-native bases of particular categories), but also what attaches to them. For example, according to this hypothesis, the reason that we don’t find words like purifiance is that the suffix -ify selects the suffix -ation as its nominalizer. So we find purification, and we predict that any new verb in -ify that we create (say, Bushify), will allow -ation to attach to form its nominalization (therefore Bushification). Similarly, any verb formed with the prefix en- in English (e.g. entomb), will form its nominalization with -ment, because en- selects -ment as its nominalizer (so entombment, rather than entombal or entombation). This sort of selection is called base-driven selection.

Another proposal is called Complexity Based Ordering. According to Hay and Plag (2004: 571), the gist of this proposal is that “the less phonologically segmentable, the less transparent, the less frequent, and the less productive an affix is, the more resistant it will be to attaching to already affixed words.” For example, the suffix -ness is extremely productive, its meaning is always transparent, and it’s easily segmentable from its bases. According to this hypothesis, it should not be resistant at all to attaching to other affixes, and this is indeed what we find, as the examples in (21a) show. By contrast, the verb-forming suffix -en (as in shorten, deepen), is not terribly productive or frequent, and because it is vowel-initial, is less easily segmentable from its base than is -ness, which is consonant-initial. As the hypothetical examples in (21b) show, it is difficult to find any suffixes that -en can attach to:

Indeed the only affix that -en can attach to is -th (lengthen, strengthen), which is completely unproductive in English, and even less segmentable from its bases than -en itself is.

What we can see is that there are a number of hypotheses that partially explain how derivational affixes are ordered with respect to each other in English, but that this question is by no means settled. Theorists will continue to work on this problem for some time to come.

مواضيع ذات صلة

Appendix 2: Hapax legomena from the Cobuild corpus

Appendix 1: 20th century neologisms from the OED

Morphological Productivity Structural Constraints in English Derivation Conclusion

Verbal derivation and the separation hypothesis

The separation hypothesis

Theoretical implications: Against the separation of form and meaning in morphology

Rival morphological processes 4: Where have all the rivals gone? Summary

Conversion vs. -ize, -ify and -ate

Rival morphological processes -ize vs. -ate

Rival morphological processes -ize vs. -ify

Rival morphological processes 4: Where have all the rivals gone?

The structural properties of other verb-deriving processes Summary