stratum 2 ones. So, we have stratum 1
-ian right after the root and stratum 2
-ness on the outside as in
Canad
(a)ian-ness
. Further, as you will recall from section (
4.4.1
), derivational affixes usually go on the inside.
They are nearer to the root than derivational suffixes, which are attached later and appear on the periphery of
the word.
The second point concerns the way morphological rules interact in the lexicon. The organisation of strata
in a hierarchical manner can be used to explain the phenomenon of BLOCKING whereby one rule makes a
pre-emptive strike, robbing another rule of its potential input. I will illustrate blocking with two sets of
examples, one from inflectional morphology and the other from derivational morphology.
In derivational morphology it is not uncommon to find two or more affixes with virtually identical, or at
any rate closely related, meanings. In such situations, usually only one of those affixes can be attached to a
particular base. You can observe this if you consider the ways of deriving from verbs AGENTIVE NOUNS
(with the meaning ‘someone who does X’) by suffixing
-er or
-ant:
[6.8]
a.
Agentive noun
Agentive noun
-er
-ant
sell-er
contest-ant
send-er
defend-ant
gaol-er
serv-ant
There is no doubt that
-er is the more productive of the two suffixes. Most verbs can be turned into
agentive nouns meaning ‘the doer of X’ by suffixing
-er. More existing and potential nouns are formed by
suffixing
-er than are formed by suffixing
-ant.
The
-ant suffix is at stratum 1. It requires stress to fall two syllables before it, if it does not do so already:
[6.9]
a.
Stress-shifting
b.
Stress already 2 syllables before
`applicant (
`accountant (<`account)
`Protestant (pro`test)
`occupant (`occupy)
But
-er is neutral and belongs to stratum 2 since it has no phonological special effects (cf.
`manage ~
`manager; pro
`
duce; ~ pro`ducer etc.).
The existence of a comparable stratum 1 affix attaching to the same class of bases and having a meaning
that is very closely related to that of a stratum 2 affix usually deprives an otherwise general stratum 2 affix
of its input. That is why, for example, the existence of
accountant, formed at stratum 1, blocks the
formation of
*accounter, and
applicant blocks
*applier.
Observe, however, that blocking is only a strong tendency rather than an inviolable constraint.
Sometimes closely related stratum 1 and stratum 2 affixes may be used with the same base on different
occasions. Interestingly, this seldom results in synonymous words. Rather, what we tend to get is one word
which has a broad meaning and another which has a narrower, specialised meaning. For instance, a
Do'stlaringiz bilan baham: