OF NOMINATIVES AND DATIVES 473Figure 3tiality (i.e., successivity) with respect to the moment of speaking as prior
(plus vector). A next more complicated system takes the "past" as a
moment of reference, and has a special marking of vectorial relationships
with respect to this moment in the past (note: this is distinct from merely an
aspectual relationship of two events narrated in the past, a matter of clause-
clause relative aspect and narrated sequentiality relationships), and then
next more complicated takes the "future" as a moment of reference for a
minus vectoral, and as I recall Bull located no more complicated systems.
In other words, what we have is a progression of gradually more and
more differentiated formal systems, which can be ordered as to complexity
of marking and compared one with another by giving criteria of intervals
and their sequentiality, defining relative "moment" or point-like intervals as
moments of reference, and defining "vectors" by sequentiality relation
ships; every tense-aspect inflectional system can be encompassed within this
regular array. The array constrains the possible tense-aspect systems of lan
guage, as well as provides the unifying hypothesis of what these formal
grammatical systems are doing, with some language-independent metalan
guage. There is, again, no analytic apparatus available to interpret and con
strain the formal devices of any language without such an hypothesis, just
as there is no understanding of the function of certain systematic categorial
interactions without such an hypothesis that says that the categories should
be constrained as those of tense-aspect. The minimal system defines a par
ticular functional domain which every formal system worthy of the name
"human language" must mark or signal in a grammatical way, and the max
imal system defines the limits of systematic effability in that domain, as far
as we know the domain to be structured (as opposed to lexicalized on the
basis of pragmatic considerations).