104
also Anderson 1992: 61-62). If it is assumed that present tense stems are basic
as compared to derived past tense stems, then a process of past stem formation
may be identified that derives ritt from reif, the past tense marker is, then, a
change from ei to i (more exactly, /ai/ to III), and indeed, whenever a strong
verb in German has a stem in /ai/ and another stem in III, the former is a pre-
sent stem and the latter is a past (and past participle) stem. It would appear,
then, that an inspection of relations between stems might open up a viable
avenue towards uncovering the system of ablaut.
But in general, inspecting relations between a base form and a derived
form does not suffice to establish the interpretation of the derived form. The
reason is that different derived forms may compete, as it were, for the same
range of application; and in this case, the latitude a derived form has may be
cut by other more specific derived forms. An example that will be discussed in
more detail below is this: in the case of the verb
FECHTEN
w, the alternation
/e/—>/o/ serves as a general past marker: both the past stem and the past parti-
ciple stem show h ! (the principal parts being fechten-focht-gefochten). The
same alternation applies to
SPRECHENW.
But in addition,
SPRECHENW
(spre-
chen-sprach-gesprochen) shows another vowel change, /e/—>-/a:/, which
marks the stem of finite past forms; in this way the stem in lol is restricted to
the non-finite past, viz. the past participle. To cope with this situation item-
and-process-models have to integrate some paradigm-like mechanism.*
Eventually, the interplay among various forms of a paradigm must not be
disregarded if it is to be elucidated how the forms’ interpretations come about.
Thus an approach is needed that grants paradigms their due status in language
systems. At the same time, simply reappointing the traditional word-and-
paradigm-model would not do. The ‘structuralist’ insistence on uncovering in
detail how forms are made up in order to clarify how form signals function
must not be abandoned. An adequate analysis of ablaut must not be limited to
dealing with words and word forms but must also regard sub-word items,
stems in particular. What is needed, then, is an item-and-paradigm-model or
stem-and-paradigm model (cf. Blevins 2003). An approach to morphology
(and syntax) that meets these requirements is part and parcel of the larger
framework of Integrational Linguistics (Lieb 1983, Eisenberg 1998). Some
relevant concepts that are used in the following will now be introduced. 8
8 Cf. complex rules or disjunctively ordered sets of rules as in Bierwisch (1967) and in
Anderson’s ‘Extended Word and Paradigm Morphology’ (Anderson 1982). As Bierwisch
(1967: 267, cf. 270) says, “We may conceive of a disjunctively ordered set of inflectional
rules as displaying an inflectional paradigm.”
105
1.4 Stems and stem forms
An analysis of ablaut must set out to clarify the relations that hold between
items that are often referred to as ‘tense stems’ of strong verbs (‘present
stems’, ‘past stems’, ‘participle stems’); usually, ‘primary present stems’ are
taken to be basic while stems of other types are assumed to be derived. At-
tempts at grasping the alleged derivational relations in syntagmatic terms (by
reference to combinations of morphemes) have turned out to be inadequate. A
paradigmatic treatment would appear to be more promising. From this van-
tage point, a past stem is not construed as a ‘combination’ of a basic form and
an item that signals tense, but as an ‘alternative’ to the basic form (that occurs
in lieu of the latter under certain conditions).
This change of viewpoint brings with it a change in terminology. So far I
have followed the older tradition in using terms like ‘past stem’ and ‘participle
stem’ when referring to morphological ‘items’ such as sprach, sproch. An-
other practice that would seem more suitable for a paradigmatic approach
treats such items as ‘alternants’ or ‘alternating variants’; from this angle,
strong verbs “have more than one form of the stem” (Halle 1953: 46).9 As
words (such as the verb
SPRECHEN
w) are to be distinguished from their forms
{sprechen, spricht, sprach, ...), so, by this approach, stems — that is,
stem
lexemes — are to be distinguished from stem forms. Consequently, Lieb
(1976: 21) introduced the notion of morphological paradigm: as word forms
may be put into word-paradigms, stem forms may be put into stem-paradigms
(cf. also Lieb 1983: 169, Eisenberg 1998: 29). The relationship between stems
and stem forms may be construed analogously to the relationship between
words and word forms, as illustrated by the following comparison (see table
on next page).
As the example indicates, word-paradigms may be construed as sets of,
traditionally speaking, ‘grammatical words’, a ‘grammatical word’ being an
ordered pair comprising (i) a syntactic unit and (ii) a set of categories to which
this unit belongs (this set is called a categorisation)', thus spreche is the form
of the 1st person singular indicative of the present of the active of
SPRECHENW.
Likewise stem-paradigms may be construed as sets of pairs, each of which
combines (i) a morphological unit and (ii) a set of categories to which this unit
belongs; e.g., sprech (a stem form) is a form of
SPRECHL
(a stem lexeme).
9 Halle decides to adopt the term stem alternant and uses the term ‘stem form’ (1953: 47)
only informally.