Volume 39, No.3 - Fall 1993
Editor of this issue: Antanas Klimas, University of Rochester 
ISSN 0024-5089
Copyright © 1993 LITUANUS Foundation, Inc.



Lithuanians will tell you that their language has a lot in common with Sanskrit. Baltists will tell you that Latvian is a sort of less well-preserved Lithuanian. Listen to the Lithuanians. Ignore the Baltists.

Latvians can now take pride in their language, too. For they can learn that it is the northern branch of West Baltic and stands with the southern branch, Prussian, against East Baltic Lithuanian.

Baltists, from the beginning, recognized Prussian as West Baltic. It seems so different from Lithuanian and Latvian. And wherever they noted agreements between Latvian and Prussian, they seemed to react to them as if they were isolated curiosities to be semi-forgotten.

But they remember well the agreements between Latvian and Lithuanian. They misname many of them. They call them East Balticisms. But they are more careful about the agreements between Prussian and Lithuanian. These they do not call East Balticisms. Nor do they call them West Balticisms.

Since agreements between Prussian and Lithuanian and Lithuanian and Latvian run from north to south together with transitional zones (e.g., the diphthong ei remaining constant in Prussian: deiws 'God', fluctuating in Lithuanian: dievas/deivė 'God, goddess', and uniformly metathesized in Latvian: dievs, dieve "Cod, goddess'), I suggest that we refer to them as North Baltic versus South Baltic. Here, Prussian-Lithuanian agreements are to be designated "South Baltic" while Latvian-Lithuanian ones are to be named "North Baltic". But agreements between Prussian and Latvian are West Baltic. Characteristically, these stand, without transitional zones, against Lithuanian East Baltic features.

There is clear phonological and morphological evidence to view Latvian, Prussian, and Slavic as a West Satem bloc v.s. the East Satem one with Indie and Iranian giving the clearest portrayal of some of its features. Lithuanian tends to show features parallel to these, specifically closer to Indie phonologically and closer to Iranian morphologically with respect, at least, to one feature shared by Old Lithuanian and Avestan, the definite adjective with a preceding pronominal element: OLith. paioprasta: Av. yaiš vahistaiš. This we do not find either in Prussian, or in Latvian, or in Slavic where the pronominal element always follows.

In Latvian, Prussian and Slavic, the reflexes of Indo-European palatals, k', g'(h) and of s before them are s, z. In Lithuanian they are š, ž. Thus, Latv. sirds, Prus. setr, Russ. serdce/ Lith. širdis 'heart'; Latv. zeme, Prus. zemė (written semme), Russ. zemlja;/ Lith. žemė 'earth'; Latv. vasks, Russ. vosk/ Lith. vaškas 'wax'.

And where Latvian, Prussian and Slavic do show š or ž from sj, aj, Lithuanian, at most, shows palatal(ized) s, z as in Latv. sut, OCS. šiti 'sew' Prus. schuwikis 'cobbler'/ Lith. siūti, siuvikis 'sew, cobbler' and Latv. atgriezu/atriežu 'turned back/turn back', Russ vozil/vožu 'conveyed/I convey', Latv. zvai(g)zne/ zvai(g)žnu 'star/of stars' (Lith. žvaigždė/žvaigždžių show no z/ž alternation).

As for ruki law reflexes where Slavic, with limitations, agrees with Lithuanian, Indic, and Iranian, we see Latvian and Prussian in agreement against them: Russ. verx, Lith. viršus, OInd. varšma/ Latv. virsus 'top'; Lith. veršis 'calf' OInd. vrša-'ox', Av. varšni- 'ram'/ Prus. werstian, Latv. versis 'calf'. Why? Because Baltic alone uniformly has joint ruki law and palatal reflexes whether they be East Baltic Lithuanian š, ž or West Baltic Prussian and Latvian s, z.

Otherwise, note West Baltic Prussian and Latvian morphological features with syncope: o-stem nom. sg. masc. Prus. deiws, Latv. dievs 'God'; definite adj. masc. nom. sg. Prus. pirmois, Latv. pirmais 'the first'. Against these stand East Baltic analytical forms dievas pirmasis with -a- and -s- intact.

Note the following items showing very early phonological parallels between Lithuanian and Sanskrit. These support statements by Lithuanians referring to special ancient ties between the two languages. One feature, that of s to š (Lith.), č (from š) (Sanskr.) via assimilation in Indo-European times to palatal k', indicates a truly ancient neighborship with sustained contact between them.

Lith. šimtas: Sanskr. satam '100' with š: s from I.E. k'

Lith. ieškoti 'to search': Sansr. icchati 'he searches' with šk: cch from I.E. xšk' from sk'

Note later, continuing parallelism in resisting sj to š.

Lith. siūti 'to sew': Sanskr. syūta- 'sewn'.

Lexically, most Lithuanian native items minus Latvian correspondences and Latvian native items minus Lithuanian correspondences are classifiable as "western". But most Lithuanian words of that kind match up with Slavic and/or Prussian ones only and are, therefore, relatively recent. Latvian words of that kind match up more with demonstrably much older items, ones matching Hittite and Tokharian elements as well as Latin, Greek, Gothic, etc. items, ones that tend to last in the west but die out in the east. This indicates a more (long lasting) "western" character also on the lexical level for Latvian than for Lithuanian. Here is a comparison of the oldest lexical items in both languages pointing to their western versus eastern positions. Since I count Hittite and Tokharian as "neutral,"1 Indic, Iranian and Armenian (together with Phrygian which I consider the ancestor of Armenian) as "eastern," but Italic, Celtic, Germanic, Albanian, Prussian, and Slavic2 as "western", I weight my statistics to compensate for the far greater number of "western" languages. This is important since "western" or "eastern" classifications often depend on the preponderance of evidence one way or the other.3

For Lithuanian: "western" - angu 'or': Hitt. -uk and items in Prus. (anga), Lat, Gk., Goth./OInd.; dosyti 'make mad': Lydian kandaulės and items in Gk., Lat., Illyr., Goth. (dauths 'death'), Russ. (davit')/Phryg.; piešti, paišas, pišas 'draw, soot': Tokh.A. pek-, pik-, pekant- and items in Prus. (peisaton), Russ. (pisat', pestryj), OHG., Goth., Lat., Gk./OInd., Av.,: petys 'shoulder': Hitt. pittar and items in Prus. (pette, petis), Gk. (unsure), Lat., Oscan, OEng., Eng. (fathom), Scot.Av., Arm .; sūnus 'son': Tokh.A. se, B. soya and items in Prus. (souns), Goth. (sunus), OCS. (synju)/Olnd., Av.; to which may be added auksas (with West Baltic ruki law s!) 'gold': Tokh. A. was, B. yasa with items in Prus. (ausis), Lat. (aurum), Sab. (ausom), the lone example without "eastern" language correspondence if we disregard this word's disqualifying tie to aušra 'dawn' which in turn is connected with Latvian austra, austrums 'dawn, east'.

For Lithuanian: "eastern" ant 'on': Tokh. A. ant with items in Gk., Lat., Goth./Oind., Arm.: duktė 'daughter': Tokh. A. ckacar, B. tkacer with items in Prus. (duckti, poducre), OBlg., Goth., OHG., Gk./OInd., Av., Arm.: našlys 'orphan': Tokh. AB. nak-, Hitt. henkan (Henk-, hink-) with items in Gk., Lat., Irish, MWelsh, Welsh/OInd., Av., OPers.: pas 'by': Tokh. B. ompostam with items in Arc. Cypr., OBlg. (pozde), Lat. (post)/ OInd., Av., Arm.

To determine the "old western" character of Lithuanian, we subtract the above four "eastern" from the six "western" items minus Latvian correspondences. This leaves us, at most, two words of "old western" character in Lithuanian. Against these two words, after a similar reduction of only two "old eastern" words, we will find six 'old western' words remaining in Latvian minus Lithuanian correspondences. These are as follows.

For Latvian: "eastern" — agri, agrums 'early': possibly Hitt. be-kur, be-gur with OInd., Av. items (an East Baltic substratic element?); asins 'blood': Tokh. AB. ysar, B. yasar, Hitt. ešhar/ ešnas, ešhanas with items in Lat., Gk./OInd., Arm.

For Latvian: "western" — izpaluot, palas, pelt, appelt, nuopelt libel': Tokh. A. pal-, pal- with items in Goth., Eng., (gospel), NHG. (Beispiel), Alb., Gk./Arm.: legens 'slack': Tokh. A. slakkar (unsure) with items in Russ., Gk., Lat. (laxus), OIrish, NHG., Dial., NHG (link), OIcel. (slakr)/OInd.; patape 'thaw': Tokh B. tsatsapau with items in Prus. (Tappelauken), Alb., Lat. (tepidus), Russ. (topit', teplyj), Irish/OInd., Av., NPers.: smaidit, smidinat, sminet, smiet, smaida 'smile': Tokh. A. smimam with items in Lat., Russ. (smejus'), Gk., OEng., Eng. (smile)/OInd.; the rest with "eastern" language correspondences: buozties 'get angry': Tokh B. pakware with items in OIcel., OHG., OIrish; dupeties 'thud': Tokh.A. tap- (*tup-), inf. tpassi, partic. pass. capcu with items in Gk. Hom., SC (dupiti); varsmis, varsms 'pile of grain': Hitt. waršiya-, warš-, Tokh.A. wsar, B. ysare with items in Lat., Russ.Ch.Sl., Russ. (vorox, verš'), Blg., OIcel. (vorr); veriba, vert, veru verties, veruos T'being alert, wary': Tokh.A. war-, war-sk-with items in Lat., Gk., OIrish, Welsh, NHG. (warten), OHG, Eng. (beware, warn, ward).

Statistically, if we disallow Lithuanian's auksas, Latvian's "old western" words outnumber Lithuanian's by six to one. If we allow it, then on this oldest of lexical levels, Latvian is three times more western than Lithuanian. This situation essentially continues in words like Latvian bariba, baruot 'food, feed' matched by Russ. borošno, Lat. far(ina), farris, Goth. barizeins, OIcel. barr, OEng. bere minus Tokharian or Hittite counterparts. These, though quite venerably old, may not belong to the oldest level of Latvian "western" vocabulary. And even here with its share of similar older "western" words like žvakė 'candle, torch' matched by Lat. Fax, OLat faces, Gk. fops Hesych.), Alb. duken, Lithuanian shows smaller percentages of these items minus ones with Slavic and/or Prussian correspondences only, that is, newer ones, Latvian shows 9.8 times as many "western" as "eastern" words (II:9) while Lithuanian shows only 5 times as many "western" as "eastern" words (120:24). So, in that comprehensive way, Latvian is twice as western as Lithuanian. These words are much more general, durable than the large number of more special, more recent lexical innovations which Lithuanian seems to have developed and shared in a much more restricted area with Prussian and Slavic, a fact which points to Lithuanian's relatively recent arrival in the west.

Other indications of this fact that Lithuanian is mostly an East Baltic intrusive element in what used to be West Baltic territory are the following: 1. river names with the root lat-from latvis 'Latvian' in the Vilnius area indicating a Latvian presence there before the arrival of Lithuanians (Fraenkel, 1962, 343), 2. the pronunciation krikstas for today's krikštas 'cross' in Western Old Lithuanian texts (Stang, 1966, 99), Western Lithuanian zuikis modified from Curonian *zoiekas cognate with Slavic zajeci 'hare' versus Eastern Lithuanian kiškis (Fraenkel, 1965,1281), 4. fluctuations of š, s, in ruki law positions after i, u with substratic West Baltic s (visas, ausis, glusnus 'all, ear, obedient') versus native East Baltic š (kriaušė, maišas, glušas 'pear, sack, stupid'), etc.

In sum, even with what seems like massive agreements between Lithuanian and Latvian (most of which are, in reality, more recent North Balticisms), it is time for Baltists to stop classifying Latvian as East Baltic, as a less well-preserved Lithuanian with little of interest to offer historical linguistics. It is time for them to view Latvian as most Lithuanians do, as something quite different. It is time for Baltists to re-evaluate Latvian from a West Baltic perspective. That way, they will gain a much more comprehensive understanding of that branch of Baltic than they have now with only relics from South-West Baltic, from Prussian, to consider. Now, they can add elements from North-West Baltic, from Latvian.

1 Before moving east, (Pre-)Tokharian was probably to the north of the rest of Indo-European while (Pre-) Hittite was to the south.
2 Classifying Slavic as "western" gains support from specialists who define Slavic as a one-time Prussian dialect.
3 I use the terms "eastern language," "western language" with prototypes in mind. This means, for example, that Gothic, Old English, Old High German, etc. are to be calculated together as, ultimately, one language, Germanic. With this understood, my statistical weightings are reflected as follows. I classify as "eastern", Latvian or Lithuanian words matched by two "eastern" language and one, two, or three "western" language words, and those matched by three "eastern" language and one, two, three, or four "western" language words. I classify as "neutral", Latvian or Lithuanian words matched by one word of each type, by two "eastern" language and four "western" language words, by three "eastern" language and five "western" language words, and by Hittite and/or Tokharian words only. I classify all other combinations, irrespective of items from Hittite and/ or Tokharian as "western".