Volume 23, No.2 - Summer 1977
Editor of this issue: Antanas Klimas
ISSN 0024-5089
Copyright © 1977 LITUANUS Foundation, Inc.


The Pennsylvania State University

In standard Lithuanian the *w-stem nom. pl. ending is (sūn-)ūs 'sons', and the *i'-stem nom. pl. ending is (ak-)ys 'eyes.' In earlier papers I have proposed that these plurals had their origin in old dual endings *i( «-iN) and *-ū («-uN) to which the plural marker *-s was added. (See Schmalstieg, 1973a and 1973b). My original assumption was, however, that such endings as *-i, *-ū and *-ô (for the *o-stems) denoted a plural and only came to denote the dual when the new endings *-is, *-ūs and *-ôs took over the function of a generalized nom. -acc. plural. It is interesting to note that in Lithuanian dialects a somewhat similar phenomenon is being repeated where we find new forms abidvis rankis 'both hands' (abidvi + s and ranki + s). (See Girdenis and Rosinas.) The only difference here is that these new forms still denote the dual although the method of formation is exactly what I have proposed for Indo-European plurals.

In previous articles I have assumed that in Lithuanian the function of plural endings -ūs and -ys was limited to the nominative plural with the appearance of the new accusative plural endings *-uNs (<acc. sg.'-uN + plural *-s) and * iNs «acc. sg.*-iN + plural *-s). In Lithuanian the function of the *o-stem plural ending *-ôs was limited to the accusative with the appearance of the nom. pl. ending *-oi.

I should now like to call the reader's attention to some other i- and u- stem nominative plural forms found in Lithuanian dialects, viz., (sun-)aus 'sons' and (ak-lies 'eyes' (<*-eis). Kazlauskas, 1969,11,17 considered these forms archaisms and derived the u-stem ending from *-aues and the i-stem ending from *eies. Similar to Kazlauskas I would consider these forms archaisms but I would propose to derive the endings from the simple *e/o-grade form of the stem, *e/ous and *e/ois respectively.

I believe that in the past such proto-forms were not considered acceptable a) because reconstructions of the oldest stages of Proto-Indo-European were usually made on the basis on Sanskrit and Greek (and to a somewhat lesser extent Latin) and b) because it was customary to assume for Proto-Indo-European the largest attested morpheme and then derive the morphemes of the daughter languages by means of assuming reductions of this monster morpheme. I propose therefore two methodological deviations from the established norm. First, I propose that the evidence of Baltic and other Indo-European languages may be just as important as Sanskrit and Greek for the reconstruction of Proto-Indo-European. Second, I propose that it is just as reasonable to assume that the forms of the attested Indo-European languages reached their present state of development by the accretion of minimorphemes. It is, of course, easier to derive the attested forms by presupposing some morphophonemic monster from which deletions are made rather than to assume accretions of minimorphemes. On the other hand we have no contract with God to the effect that reconstructions are to be made on the simplest principle available to us. 1

It is customary to assume that the u-stem nominative plural ending derives from *-eues on the basis of such Sanskrit forms as (sūn-)avah, Slavic (syn-) ove 'sons' and Greek (pekh-)eos 'arms.' Now it is well known that the Sanskrit *u-stem genitive singular offers both a closed inflection (sun-)oh '(of the) son' (<*e/ous) and an open inflection (madh-)vah '(of the) honey' (<*-ue+os). It may be pointed out here that there also exists a Sanskrit nom. pl. madhvah (presumably derived from *-ue/os)' attested four times according to Macdonell, 1968, 298.

I propose that in Proto-Indo-European there existed two endings vying for the position of *u-stem nom. -acc. plural, viz., *-ue/os (attested by Sanskrit madhvah) and *-e/ous (attested by Lith. sūnaus). (This is parallel to the competition of the genitive singular forms represented by Sanskrit sunoh and madhvah.)

I should note here, however, that it is not only the Lithuanian nominative plural sūnaus which attests to *-e/ous in plural function. The Gothic nominative plural (sun-)jus 'sons' could just as easily be derived from *-eus. Likewise one might see the ending *-ous in the Hittite u-stem acc. pl. da-aš-ša-uš 'strong.' Kronasser, 1956, 115, would read this as tassawus, but I would see nothing wrong in reading it as tassaus.

Now the attested forms Sanskrit (sun-)avah, Slavic (syn-)ove 'sons', Greek (pekh-)eos 'arms' do indeed, in my opinion, derive from an ending *-eues. This ending *-mes, however, is nothing but a contamination of the two endings *e/ous and *-ue/os. One should find nothing surprising about such contaminations. In fact, the contemporary Russian nom. pl. synov'ja 'sons' is apparently the result of the contamination of the nom. pl. endings *-ove (or *-ovi, itself a contamination) and -ja.

I shall herewith follow a similar line of reasoning for the etymological *i-stem nouns.

Again it is customary to establish a Proto-Indo-European *i-stem nom. pl. *-eies on the basis of such forms as Sanskrit (agn-)ayah 'fires/ Slavic kostbje 'bones,' etc. But as with the *u-stem nouns Sanskrit shows both a closed and an open inflection in the *i-stem genitive singular, where we find both the forms (agn-)eh '(of the) fire' ( *-e/ois) and (av-)yab '(of the) sheep' (*-ie+os). In Sanskrit there is also one stem, viz., ari- 'greedy, eager, hostile' which has both the nom. and acc. masc. and fem, plural form aryah presumably reflecting *-ie/os. (See Macdonell, 1968, 285.) And as we have seen above, the Lithuanian nom. pl. (ak-)ies 'eyes' attests to an earlier *-eis. Therefore, I propose that the reconstructed *i-stem nominative plural ending *-eies reflects a contamination of the endings *-ie/os (cf. Sanskrit aryah) and *-e/ois (cf. Lith. akies).

To summarize then: Sanskrit shows *u- and *i-stem genitives singular which attest to both an open inflection ( *-ue/os, *-ie/os) and a closed inflection (*-e/ous, *-e/ois). Sanskrit shows a few marginal forms with an open inflection in the *u- and *i-stem nominative plural (*-ue/os, *-ie/os). Lithuanian shows some dialect forms (apparently archaisms) which attest a closed inflection in the u- and i-stem nominative plural (*-ue/os, *-ie/os). In those languages where it is necessary to reconstruct a *u-stem nominative plural ending as *-eues, this derives from a contamination of *-eus and *-ues. In those languages where it is necessary to reconstruct an *i- stem nominative plural ending *-eies, this derives from a contamination of *-eis and /*~ies.


Girdenis, A. and A. Rosinas, forthcoming. "Some remarks on the vocalism and morphology of Old Prussian." General Linguistics 17.
Kazlauskas, Jonas. 1969. "O nekotoryx arxaizmax v fleksii im. pad. mn. č. na -u i na -i baltijskix jazykov." Lingua Posnaniensis 14. 7-17.
Klimas, Antanas. 1970. "Baltic, Germanic, and Slavic." 263-269. In Donum Balticum, Ruke-Dravina, Velta, ed. Stockholm, Almqvist and Wiksell.
Kronasser, Heinz. 1956. Vergleichende Laut- und Formenlehre des Hethitischen. Heidelberg, Carl Winter.
Macdonell, A. A. 1968. Vedic grammar. Varanasi and Delhi, Indological Book House.
Schmalstieg, William R. 1973a. "Lithuanian nominative plurals in -ys and -ūs." 197-201. Baltic literature and linguistics. Eds. Ziedonis, Arvids, Jr., Jaan Puhvel, Rimvydas Šilbajoris and Mardi Valgemae. Columbus, Ohio, Association for the Advancement of Baltic Studies.
Schmalstieg, William R. 1973b. "New thoughts on Indo-European phonology." Zeitschrift für vergleichende Sprachforschung 87. 99-157.


1 One should note the comment of A. Klimas, 1970,263-264: "First of all, let us examine the non-chronological hierarchy of linguistic data: the very old and still very strong notion that some Indo-European languages are more important for positing and visualizing ancient linguistic changes and relationships. In other words, let us no longer measure almost everything in all areas of Indo-European linguistics according to the later interrelationships of the South-Eastern Indo-European languages: Greek, Italic, Indo-Iranian, etc. It is quite clear by now that, in spite of the relatively late written documentation of the Indo-European languages of the Northern region, they often show more ancient features. . . In spite of the general acceptance of this fact, to the present day many publications in linguistics still reflect the dominance of the Greek-Latin-Sanskrit triumvirate."