lion
strong
think
very
We saw in the last chapter that many words contain a free morpheme which may occur on its own, with
nothing appended to it. The vast majority of root morphemes that are capable of appearing on their own are
CONTENT WORDS, i.e. nouns, verbs, adjectives and adverbs (see
section 2.2.1
).
We also saw that there are very many morphemes which are always kept in bondage. These are the bound
morphemes, which are totally barred from occurring independently. Many roots fall into this category.
Examples of BOUND ROOTS are provided below. In each case the root is italicised
and separated from the
rest of the word (which may contain one or more morphemes).
[4.2]
santc-
‘holy’
vir-
‘man’
tox-
‘poison’
loc-
‘place’
sanct-
ify
vir-
ile
tox-
in
loc-
al
sanct-
um
vir-
il-ity
tox-
ic
loc-
al-ity
sanct-
uary
vir-
ago
non-
tox-ic
dis-
loc-ate
sanct-
ity
trium-
vir-ate
in-
tox-ic-ate
loc-
um
The data in [4.2] illustrate another important point. Some of the forms to which affixes are attached are
words in their own right or bound roots which are not words. There is a need for a common term to refer to
forms, be they bound roots or self-standing words, to which affixes may be attached. The term BASE is
used to meet this need. Any form to which affixes are appended in wordformation is called a base. Bases to
which affixes are added can be bare roots like
loc- in
loc-al, loc-um, loc-us or they can be independent
words, e.g.
govern in
governor and
govern-ment. They can also be forms which already contain other
affixes, e.g.
[[loc-al]-ity], [[loc-al]-ism].
An issue raised by the segmentation of words into morphemes is what counts as a morpheme. I suspect
that
sanct-, vir- and
tox-do not jump at you as obvious examples of morphemes in the same way that
govern
does because their meanings are not immediately obvious. An etymological dictionary, or a knowledge of
Latin, might help to persuade you that these are indeed recognisable morphemes. But your intuitions as a
speaker of English would probably not enable you readily to reach that conclusion. So
the question arises of
how to treat such awkward cases. If a form was recognisable at some time in the past as a morpheme, does
that mean that for ever more it should be recognised as a morpheme? Where do we draw the line between
forms whose analysis is only of historical or etymological relevance and those whose analysis is well
motivated in the synchronic study of the language? We will just note this tricky problem for the moment
and defer tackling it until section (
6.6
).
Do'stlaringiz bilan baham: