Professional Documents
Culture Documents
Voyles-Cardinal Numerals in Pre - and Proto-Germanic 1987
Voyles-Cardinal Numerals in Pre - and Proto-Germanic 1987
Voyles-Cardinal Numerals in Pre - and Proto-Germanic 1987
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms
University of Illinois Press is collaborating with JSTOR to digitize, preserve and extend access
to The Journal of English and Germanic Philology
In the following article I shall reconstruct what I shall argue are the
most probable forms of the late Indo-European or immediately pre
Germanic cardinal numerals. My reconstructions will be based on the
forms of these numerals as they occur in the earliest attested Ger
manic languages, namely Gothic, Old Icelandic, Old English, Old Fri
sian, Old Saxon, and Old High German.1 As my methodological
framework for reconstructing the pre-Germanic forms of these nu
merals I make the following three basic assumptions.
First, I assume that cardinals can influence other cardinals in the
series by a kind of frequently observed paradigmatic pressure. Ex
amples of this type of change are found in many Indo-European lan
guages, such as Russian devjat' '9' with its initial [d] replacing earlier
[n] (*nevjat') on the model of desjaf '10'. Another and perhaps more
familiar example of such paradigmatic change?but here in the or
dinal series?is the pronunciation by some American speakers of sec
ond as [sek^nt] with word-final [t] carried over from first.
Next, I assume that cardinals usually?and apparently invariably as
far as the Germanic languages are concerned?determine the form
of the ordinal numerals and not the other way around. This seems to
have been always the case in the attested instances found among the
Germanic languages: e.g., ModEng. tenth instead of the phonologi
cally regular and earlier tithe (whence tithe) from the cardinal ten +
the ordinal suffix -th; and ModHG zweit '2nd' instead of earlier ander
from the cardinal zwei + the ordinal suffix -t. There exists to my
knowledge no clear instance in the history of any of the Germanic
languages of an ordinal numeral determining the form of a cardinal.2
1 Any of the standard handbooks suffice for the attestations. I have relied for Gothic
on Wolfgang Krause, Handbuch des Gotischen (Munich: C. H. Beck, 1953); for Old Ice
landic on Adolf Noreen, Altnordische Grammatik (University: Univ. of Alabama Press,
1970); for Old English on Alistair Campbell, Old English Grammar (London: Oxford
Univ. Press, 1964); for Old Frisian on Walther Steller, Abriss der Altfriesischen Grammatik
(Halle [Saale]: Max Niemeyer, 1928); for Old Saxon on Ferdinand Holthausen, Alt
s?chsisches Elementarbuch (Heidelberg: Carl Winter, 1900); and for Old High German on
Wilhelm Braune, Althochdeutsche Grammatik, ed. Hans Eggers, 13th ed. (T?bingen: Max
Niemeyer, 1975).
2 This includes the case of OI fern '5' instead of the phonologically regular but non
occurring fif. The form/<?m is often considered derived from the earlier ordinal *ftmft
487
with subsequent reduction of the consonant cluster to *fimt-. But this is not necessarily
the correct account since the consonant cluster could also and just as easily have been
simplified from the cardinal *fimftehan '15', later fimt?n.
3 Varying versions are found in Joseph Voyles, "Simplicity, Ordered Rules, and the
First Sound Shift," Language, 43 (1967), 636-60; and in Elmar Seebold, "Die Ver
tretung idg. gvh im Germanischen," Zeitschrift f?r vergleichende Sprachforschung, 81
(1967), 104-33.
regularly into Gmc. *sebun and * newun by the ist ss, verner, str
shift, syl res, and finally cor c dele and nas dele?the latter two
changes in the order given.
The third difference of my reconstructions from the traditional
ones constitutes only a minor innovation in that I pattern the '10' after
some suggestions of Szemer?nyi: "The fact that '10' appears, in some
areas at any rate, to derive from * dekm is easily explained on the as
sumption that * dekm is the preconsonantal sandhi-variant of * dekmt
[the original form]. . . ,"16 I modify Szemer?nyi's reconstruction to
* d?knt with * n instead of * m by dint of the IE nas assim rule noted ear
lier. This form then developed to Gmc. * tehun regularly and by much
the same route as '7' and '9' described in the preceding paragraph.
The fourth difference is the reconstruction of the formations for
'70' through '90' as they occurred in early Germanic and perhaps in
pre-Germanic as well. I have reconstructed them as consisting of an
abstract noun formed with the suffix *-t and in the gen. pi. This noun
was followed in turn by a noun meaning 'a basic 10-ness', e.g., Gmc.
* sebunt? * hunda = 'of 7s a 10-ness' = 'a 10-ness of 7s' = '70'. This
diverges totally from Szemer?nyi's account, whereby forms like IE
*penkw?kont- '50' and * okt?kont- '80' developed into Gmc. * femf?xanp
and *axt?xanp-.17 Then, according to Szemer?nyi, "The divergence
between E Gmc. [as in Go. sihuntehund '70'] and W Gmc. [as in OHG
sihunzo '70'] is merely due to the fact that the inherited system, with
-?- in '50' but -?- in '80', and no other juncture vowel between '40' and
'90', gave two possible ways of realization: either the extension of -?
(Gothic) or that of -0- (W Gmc.)"18?which difference, it should be
noted, is only coincidentally paralleled by the vowels of the gen. pi.:
Go. dage 'of days' versus OHG tago 'of days'. This striking coincidence
alone renders Szemer?nyi's account suspicious.
Szemer?nyi himself realizes first, that there are severe problems
with his theory and second, that the most likely competitor is one like
mine wherein the first constituent of the construction is a gen. pi.
noun. Accordingly, Szemer?nyi defends his own account while at the
same time attempting to discredit any possible "gen. pi." theory.
It should at this point be noted that a version of a gen. pi. theory
different from mine has already been proposed by Brugmann.19 His
theory differs from mine, however, in that he posits as the Indo
16 Szemer?nyi, Studies in the Indo-European System of Numbers, p. 68.
17 As outlined in Szemer?nyi, Studies in the Indo-European System of Numbers, esp.
PP- 33-44
18 Szemer?nyi, Studies in the Indo-European System of Numbers, p. 36.
19 Karl Brugmann, "Die Bildung der zehner und der hunderter in den indo
germanischen sprachen," Morphologische Untersuchungen, 5 (1890), 1?61 (Nachtrag
138-44), Leipzig.