In the lexicon derivations start off with the UNDERIVED ROOT (i.e. the naked root without affixes) and
work their way through relevant rules starting at stratum 1 and then progressing through stratum 2. The theory
attempts to reflect the fact that at each stratum morphological rules are bonded together with phonological
rules. The application of a morphological rule brings in its wake the application of a phonological rule. The
reason for this is easy to see. When in the process of word-formation we put morphemes together using
morphological rules, we need to know how the forms that we create are pronounced.
Look back to the tables of suffixes in [6.3] and [6.4]. See how different types of lexical information play
a role in word-formation. As you can see, in order to attach an affix to a base, you need to know which kind
of bases it can go with. For instance,
-ian is attached to noun bases to yield adjectives. Given the noun
Canada
you can form the adjective
Canadian. If your intention is to form a verb from a noun, then
-ian is
not a suitable suffix. You must try instead another suffix, e.g.
-ify (as in
solidify). Of course, in addition to
knowing the grammatical characteristics of the word resulting from affixation, you need to know what it
will mean. This point is obvious. We will not dwell on it.
Knowing the meanings of the words formed and their syntactic characteristics is not enough. We need to
know how our words are pronounced. By associating morphological rules directly with phonological rules
that account for the pronunciation, we are able to address this problem effectively. As we saw in [6.3] and
[6.4] batches of affixes can be shown to have broadly similar phonological effects. Rather than state these
effects separately for each affix, it is preferable to state them in more general terms. That is what we are
doing here when we say, for example, that suffixes like those in [6.4] are non-neutral. They cause changes
in the location of stress or in the pronunciation of vowels or consonants, or both. If a suffix is placed at
stratum 1, it can be expected to have disruptive phonological effects. Conversely, if a suffix is at stratum 2,
we know that it will not cause phonological disruption in the base to which it is attached.
Sample derivations of representative words listed in [6.3] and [6.4] are provided below. In [6.5] only
stratum 1 rules apply while in [6.6] only stratum 2 rules apply. But in [6.7] both stratum 1 and stratum 2
rules apply.
The claim that affixes are arranged on hierarchical strata has additional advantages. We will consider
them briefly in turn.
First, as we saw in section (
4.4.1
), the order in which morphemes are arranged in words is normally
rigidly fixed. Thus,
Canad-ianness is the only permissible MORPHEME SEQUENCING in the word
Do'stlaringiz bilan baham: