132
including past participles. Even in the case of verbs like
STOSSENw
(stoßen-
stieji-gestoßen), Segeral and Scheer maintain that vowels in past participles
are derived from finite past vowels by ablaut. But, as a matter of fact, the
verbs of the fifth, sixth and seventh class do not employ ablaut in past partici-
ple formation at all, participles showing the base forms of the stems. It must be
recognised that the primary forms of a German strong verb stem mostly in-
clude just one ablaut form; only stems of the third and fourth class have two of
them.
The form-function-relation found with ablaut in German is straightfor-
ward and rigid: this is the pay-off of a strictly ‘surfacisf approach to morphol-
ogy. Its surface orientation sets the present analysis apart from those ap-
proaches that try to grasp ablaut in terms of rules that apply to ‘underlying’
forms. Taking recourse to underlying forms only clouds the actual regulari-
44
ties.
A further study of the foundations on which ablaut is built would surely
be worthwhile. Here I confine myself to pointing to the role of iconicity as a
basis for quantitative ablaut. Change of quantity type that produces forms of
normalised length is used as a component of simple ablaut, which has the
rather unspecific functional value of general past marking. On the other hand,
change of quantity type that produces forms of marked length is used as a
component of full ablaut, which has the more specific functional value of finite
past marking. The diagrammatic correspondence between form and function is
patent. The preference for bases in front vowels as opposed to ablaut forms in
back vowels is another iconic aspect of ablaut, which, however, I cannot go
into here.44 45
4
Inflectional classes
4.1 Ablaut classes
The traditional division of strong verb classes, each distinguished by a particu-
lar overall ablaut pattern, is not an antiquated descriptive tool of nineteenth
century linguistics; nor is its relevance restricted to diachronic matters or to
44 Segeral & Scheer freely use non-surface ‘lexical’ forms (cf. note 12, supra) in order to
defend their claim that German gradations are instantiations of a certain, allegedly universal
pattern (the ‘apophonic path’). But even so, by their own count, only 59 percent of German
strong verbs fit into the scheme; the remaining ones are, in their terms, only ‘half-
apophonic’ or even ‘non-apophonic’.
45 For some hints see already Adelung (1782: 785 et passim); cf. Tanz (1971: 269) and
Plank (1979: 144—145) with reference to English, and see Stedje (1987) on ‘mentalistic
theories’ of ablaut. On iconicity of quantitative ablaut see also Mayerthaler (1981: Section
5.2.3) with reference to Gothic.
133
older forms of German or Germanic. Rather, it is particularly appropriate to
the system of strong verbs of Contemporary Standard German — in spite of
widely accepted claims to the contrary. This, too, may be read off Table 5. A
division that is very close to the traditional one arises from a strictly syn-
chronic analysis of the ablaut system when stem lexemes are classified accord-
ing to the types of alternations exhibited by their forms (cf. also Fourquet
1970: 61). Only, the division is simplified in Contemporary German since the
historical third class and the historical fourth class have become one as regards
qualitative ablaut.
NO-FULL-ABL
fl-ABL
l-ABL
W-ABL
NO-O-ABL
1
V
VII
VI
C SCHLAF AI iase TRAG )
( KOMM
U-buse
RUF )
f
REIT
/
A
LES A
O-ABL
/-
a
V BIEG
use
1
SPRECHJ
11
II1/IV
Table 6.
Stem lexemes: ablaut classes
As the above analysis has revealed, change of quality type is a fundamental of
ablaut in German; two types have been distinguished: full ablaut (comprising
/-ablaut/a-ablaut/n-ablaut) and o-ablaut (a variety of simple ablaut). Accord-
ingly, stems may be classified into those that show both o-ablaut and full
ablaut (cf.
SPRECHL),
only the one or the other (cf.
LES
l
and
BIEGL)
or neither
(cf. REITl ,
which has /-monophthongisation, a second variety of simple ablaut).
(In addition, there are a few stems with diphthongal bases that show o-ablaut
or /-ablaut but no change of quality type, cf. Section 2.6.)
I assume, then, a cross-classification on the set of strong verb stem lex-
emes based on two criteria, viz. ‘o-ablaut: yes/no’ and ‘full ablaut: yes/no’ (in
the affirmative case: /'-,
a- or w-ablaut). As full ablaut has three subtypes, the
system allows eight types of stems, as illustrated in Table 6. The rows repre-
sent the classification according to o-ablaut; the columns represent the classi-
fication according to full ablaut. The heads of rows and columns introduce
names for the classes that are distinguished; for instance,
o-ABL
is the set of
verb stems that exhibit o-ablaut — an alternation pattern identified in Sec-