This set consists of the various disjoint subsets, most of which are not relevant here.
The subsets include
for root property combinations,
for inflectional property combinations, and
for the properties of simplex words.
The set of ID categories for both lexical and inflectional morphemes
will be referred to as follows:
.
Discussion: What are the permissible properties of English lemmata and inflections, e.g. for the words bread, bake? The following definitions can be referred to.
For example, applying this to inflections for a specific noun entry bus:
,
where:
e.g.
[ Number=plural, Case=nongenitive, Paradigm=regular ]
,
where:
Abacus, ... , Zygote
abacus, ... , zygote
e.g.
[ Lemma = Bus]
For example, applying the function
to
and
, and applying
to the result, will give
:
= [ Lemma = Buses, Number=plural, Case=nongenitive ])
When applied to feature structures, the construction function
is defined as the unification operation, which requires that the resulting feature structure must contain all the feature specifications of the operands and be a permissible feature structure.
A permissible feature structure requires that each of the features it contains may only be specified once. This excludes contraditions, such as [ Number = singular, Number = plural ]. The notion of feature structure requires further definition, but this will not be given here.