Acta Linguistica Lithuanica

LVII (2007), 1 – 15

Old Prussian riste, Lithuanian rykštė, Latvian rīkste and cognates¹

BERND GLIWA

Riga

OPr. riste ‘rod, switch’, Lith. rykštė, Latv. rīkste have hitherto been connected with Lith. rišti ‘bind, tie, bundle’. Smoczyński has recently connected it with either Lith. rėžti ‘cut, scratch’ or rėkšti ‘gather’. Formerly terms like Lith. vytis, Old Latv. wīhte rather than rykštė etc. were used for binding rods, which suggests a connection with Lith. rėžti ‘cut, scratch’ < PIE *ureh₁ǵ- ‘break, snap’. A reflex of the inherited zero grade might be Lith. vìržis ‘heather’, formally divergent from Latv. vìrsis, Russ. бе́ре̄ск ‘id.’. The renewed zero grade underlying Lith. rykštė appears also in rýžti ‘disappear; appear, bud etc.’, rýkšti ‘fray out, decay’. The change ž > š could have been induced by following consonants, e. g. -sta- in the present *rýž-sta or by k-insertion. Lith. ruõšti, rṹošti ‘prepare, make household work’ could also belong here if it continues ruõžti ‘cleanse, sweep’. The new zero grade underlies new ablaut series, cf. Lith. rãžas ‘stubble, old broom’, raižýti ‘cut, scratch’ etc. Lith. ryškùs, raiškùs ‘clear, distinct’, ryškėti ‘become distinct’, raiškà ‘meaning, expression’, rėikšti ‘mean’ could also be based on rýkšti ‘appear, become visible’, similarly Lith. rãštas, Latv. raksts ‘writing, pattern’ from ‘scratched sign’.

Whereas in the older literature Old Prussian riste – rute (E639) and Latvian rīkste, rīksts, rīksts, riste (ME₃₃₃ 538f., LEV₂ 122), Lithuanian rýkštė ‘rod, switch, pole’, in Zietela also rýštė, rýštė (LKŽₓₗ 707), are generally connected with the Lithuanian verb rišti ‘to (tie) bind’, as a switch that served as a binding material (LEW 734; Mažiulis 1997: 29; LEV₂ 122), Smoczyński proposes a connection with the Lithuanian verbs rėžti ‘to scratch, (cut) off; strike’, parýšta, parýžta ‘disappear, diminish’ (2007: 507), or rėkšti ‘to pick’, *rīkšti, as its source, to the

¹The research was conducted within the framework of the research fellowship awarded by the German Research Foundation for “Etymological Studies of Latvian Plant Names”.

1

Bernd Gliwa

letzteren mit dem Ablaut wie trýkšti: trėkšti (2007: 512). Vom Autor wurde (2008) bei der Untersuchung des litauischen Gottesnamens Dviriks für den Anschluss an lit. rėžti plädiert.

Lithuanian rėkšti ‘to pick’ has no further possibilities for comparison (Smoczyński 2007: 507), so it seems possible to assume that it developed within Baltic from the same source as Lithuanian rėžti, namely Proto-Indo-European *ureh₁ǵ- ‘to break, tear’ (LIV: 698)—even though the circumflex in rėkšti argues against this. In a marginal note (Gliwa 2006: 65), Lithuanian rakštis ‘(embedded) splinter (in the skin)’ and rãkštas, rakštis, rãkštis ‘coffin; grave, gravestone’ were also connected with Proto-Indo-European *ureh₁ǵ-.

The oldest Baltic attestation, namely Old Prussian riste in the Elbing Vocabulary, paired with = Middle High German/ Middle Low German rute ‘switch, pole, staff, rod’, does not support the meaning of riste as a rod for binding. The same is true of the other two languages. Latvian rihste is translated in the Manuale Lettico-Germanicum as Ruthe, Scepter, Fiddelbogen (Fennell 2002: 70); for early Lithuanian, the entry rozga / virga, ferula. rikšte. _{(SD₁: 157)} may be cited.

Withies—that is, those branches, especially of willows, bird cherry, hazel, etc., that are used for binding—are usually called Lithuanian vytis, vytėlė and their equivalents. Such an entry appears both in the Dictionarium trium linguarum: Wić/ restis. witis. and witka/ dimin. witele. _{(SD₁: 192)}, as well as in early Latvian, where Füpecker has wihte in the context Schautrs. Ein Knüttel. ein Prügel oder Stock, damit man die Kluge oder wihte umbdrehet, as well as No kasas wihtoleem darra Kluhgas, tee ir Šihksti. Von bach=willows one makes withies, which are pliant (Fennell 2000: 304). Although Fennell formulates this very cautiously as “sense unclear,” he is of course right to point to Kluhga, Sem wizze, Curl. Eine band weide, wede. (2000: 278). In Old Prussian, the corresponding designation for the willow, or a species of willow, is witwan (E603). Thus, the situation is that rykštė, etc., did denote rods and switches, but only ones that were not used for binding. A separate term was used for these binding rods, the withies—and this at a time when rods and withies were indeed in everyday use. Since the attestations in all three languages are highly comparable, a late semantic shift cannot simply be assumed—it is unlikely to have proceeded analogously in all three languages. This makes the assumed connection between rišti and rykštė diachronically highly improbable—

2

Apr riste, lit rykštė, lett. rikste und Verwandtschaft

probable and synchronically limited to folk-etymological influences and conflations.

On the other hand, there are no formal arguments against connecting it with rišti and thus with PIE *ureik- ‘to twist, wrap up’ (LIV 699), especially since forms with k also occur in this connection: riškėlė ‘neckerchief, tie’, raikštis ‘rope; connection between the handle and the club of a threshing flail’, reikštis ‘same’, Latvian rieksts ‘hazelnut’ alongside Lithuanian riešutas ‘same’, rieškutis ‘open hand, handful’, related to riešas ‘wrist, palm’, formed from *rie(k)š(C)- → rieškV-.

For purely semantic reasons, therefore, I propose connecting rykštė with Lithuanian rėžti ‘to scratch, cut (off); strike’ < PIE *ureh₁ǵ- ‘to break, tear (intrans.)’ (LIV 698), although rykštė, Latvian rikste, and Old Prussian riste can only be Baltic formations, which cannot be explained from the PIE inventory of forms. The meaning of rykštė would thus not be a branch for binding, but a switch for driving or whipping livestock, or indeed a rod for scoring or drawing. The insertion of k before consonant clusters is a widespread phenomenon in the Baltic languages (Stang 1966: 3), and is occasionally found before a single consonant as well, specifically before s: Lithuanian áuksas ‘gold’: Old Prussian ausis ‘same’; Lithuanian úoksas ‘tree hollow’: úostas ‘harbor (< river mouth)’ (Smoczyński 2006: 141). After this inserted k, Lithuanian ž becomes š (Latvian z becomes s), e.g. Lithuanian gražùs ‘beautiful’: grakštùs ‘elegant’; brėkšti ‘to dawn’ from PIE *bhreH₁ǵ- ‘to shine, gleam’ (LIV 92); tūžti ‘to flood; swell’ (LKŽ XVII 203); tūkšti ‘same’, the latter of which might also belong to PIE *teuh₂- ‘to become strong’ (LIV 639), or may have been influenced by derivatives originating from this root (Gliwa, in press). Assimilation can also proceed in the opposite direction, k → g, or k may be replaced directly by g: Lithuanian driegžlys alongside driežlys, driežas ‘lizard’, and druožglė alongside druožlė, drožlė ‘shaving’, or where g/k is part of the root, augti: aukštas.

Old Prussian riste ‘switch’ does not contradict the proposed derivation from PIE *ureh₁ǵ-, since Old Prussian s corresponds to both Lithuanian s and ž, e.g. Old Prussian soalis (– Middle Dutch krewtecht): Lithuanian žolė ‘grass, herb’. A direct derivation from this PIE root is not possible, however; rather, this can only be a later formation, which may be termed Baltic here, based on other reflexes of the root. Let us therefore first consider those formations that can be derived directly from PIE. First there is Lithuanian rėžti ‘to scratch, strike’ (-é- < *-eh₁-), with no exact correspondences in Latvian

3

Bernd Gliwa

or Old Prussian, but with Greek ἔρρηξα ‘I broke’, and rúožas ‘strip; layer; plot of land; old broom, handle’; rúoštas ‘piece of land’ (Lithuanian -uo- < Baltic -ō- < PIE *-oh₁-) (Smoczyński 2006: 118, 140; 2007: 512). Here, too, there is no Latvian correspondence, although it would have merged with roza [ruôza] ‘hill, ridge’, related to Latvian rozīt ‘to stretch, extend; bend’ and Lithuanian rąžinti ‘to tug, stretch’, and could thus be identified as a homophone at most on the basis of nuances of meaning—interesting in this connection is dialectal roze ‘strip of deciduous trees in a coniferous forest’ (LEV_II 131). As a zero grade *urh̥ḱ-, with *r > ur, Lithuanian uržti ‘to scratch, gnaw, scrape’ and perhaps urkštis ‘narrow vestibule in a cowshed’, if derived from ‘strip’ or the like, may be proposed, as may—through the more common *r > ir—Lithuanian viržis ‘heather, Calluna vulgaris’. A possible formal Latvian correspondence could be urkša ‘a stick used to dig potatoes out of the furrow’ (ME_IV 306), unless this arose by way of urkšt ‘to grunt, growl’ → urkšnāt ‘to root about with the snout while grunting’ (ME_IV 306) → *‘a stick that roots about like a wild boar’.

Lithuanian viržis already does not match Latvian here, where forms with s predominate²: virsis, virsi, vērsnis, viršņi, among others, although forms with z are also occasionally found: virzāji, virži, virzenes, whose attestations come from Roja, Gārsene, and Salaca and thus cannot all be explained by Lithuanian influence (EO 440; ME_IV 619). The more suitable form virzis, virgzdu zāle, by contrast, is used for the dissimilar and botanically unrelated plants Stellaria spp., particularly S. media, and Anagallis arvensis. I see no reason for confusing the two plants (groups), or for their sharing a name on account of a particular use. At most, might this be due to large, dense, ground-covering stands in suitable locations? That, however, is far too general. Different origins are therefore likely³ and possibly also a

² The mismatch between the two Baltic names is simply passed over, while emphasizing the much rarer Latvian forms with z (LEW 1264), or an attempt is made to resolve it through root etymology by positing root extensions that have no more specific justification (LEV_II 539). Brückner (1957: 634) assumes “ž zamiast s” for Lithuanian viržis.

³ For chickweed, Latvian virzis, a connection with Lithuanian viržė ‘rope, cord’, virželė ‘thread’, etc., may then be assumed, after its tangled, soft, low-lying, threadlike stems. Mühlenbach offers a similar explanation: “Very branchy and, when growing densely, forms a mass that is difficult to disentangle; hence probably [...] related to the root of varza.” (ME_IV 619).

4

Apr. riste, lit. rykštė, lett. rīkste und Verwandtschaft

The differentiation in Latvian into heather versus twig serves to avoid confusion, for example by using regional forms. The plant name’s association with scrape still requires semantic justification, namely that the plant is used to make coarse brooms⁴ and sweeping with these can certainly be described as scraping; cf. broom-scraper, broom-scratcher or the broom wore out—only the stubble remained (LKŽ_XV 37ff.). On the other hand, prickly heather is also responsible for scratches and abrasions, which are mentioned particularly often in Latvian folk songs, e.g. the heather pricks my legs (ME_IV 620) or the heather tore my shoes (ME_IV 612).

Lith. vīrzis and Latv. virsis are usually compared with further near-equivalents in other Indo-European languages. These include first of all the Slavic names for heather, Polish wrzos, Russian береск, Czech vřes, and others, which sometimes also designate other heather plants such as Erica tetralix, Empetrum nigrum, and Juniperus spp. (Boryś 2005: 712). Furthermore, Old Irish froech and Welsh grug derive from PIE *uroiko-, as does Greek ἐρείκη (GEW_I 551), which presupposes PIE *uereik-. In addition, *uerǵ- is posited for Baltic and *uerk- for Slavic (Boryś 2005: 712)⁵. Smoczyński (2007: 761) explains this somewhat differently, deriving the Baltic and Slavic forms from the zero-grade or e-grade of PIE *uŕk-/*uerk-, with Greek ἐρείκη as a comparator. On this account, Lith. -rš- became -rž- through assimilation or under the influence of veržti ‘to bind tightly (together)’ and viržėti ‘same (iterative)’. It should be objected, however, that -rš- is retained in viršus ‘above’, and this form could likewise have analogically influenced its retention. Moreover, the acute accent remains unexplained. Still, the ablaut alternation appearing in the variants *uŕk-/*uerk- finds a parallel in Lith. pušis ‘pine’

⁴ Concrete data for this are lacking from Lithuania and Latvia, but they are attested, for example, in Bulgaria even in the present day (Nedelcheva et al. 2007), and the scientific genus name Calluna is derived from Greek καλλύνει ‘to clean, sweep out’ (Genaust 1996: 118). The German name Besenheide (‘broom heather’) speaks for itself.

⁵ A more correct reconstruction would be *ur(H)ǵ- for Lithuanian and *ur(H)k- for Latvian. If *uerk- is posited for the Slavic forms, one wonders why Russian †берех is not obtained; compare Lith. viršus: Russian верх. Vasmer seeks to avoid this (REW, 296) by comparing Lith. virkščia ‘stem, herb of various plants, especially legumes and potatoes’, which is supposed to contain a -sk- suffix that also underlies the Slavic forms, given the isolated Russian береск. Thus, *uer-sk- is implicitly posited. The comparison with virkščia cannot be maintained, since this belongs, with an intrusive k, to the already-mentioned vīržė, virželė, Latv. virzis, etc.—once again, these are soft stems that usually lie on the ground and wind around.

5

Bernd Gliwa

in contrast to Old Prussian peuse ‘pine tree’ and Greek πεύκη, even if Old Prussian peuse may go back to a zero grade (Smoczyński 2007: 493). Most likely, different forms of the declension were generalized here, and perhaps the accentual pattern of this declension could also explain the Lith. acute accent?

Benseler (BGD 305) notes that Greek ἐρείκη is not the “heather” familiar to us, Calluna vulgaris, but rather an unspecified shrubby plant, identified by Genaust (1996: 233) as Erica arborea. Of course, other or broader meanings cannot be ruled out, but the emphasis on heather honey in Pliny (GEW₁ 551) does not preclude identifying it as Calluna vulgaris, nor as Erica arborea. Heather is native to Greece and parts of Turkey (Zander 2000: 246), so its distribution is no obstacle either. If we abandon the idea of exact formal correspondence in plant names—which is hardly ever encountered anyway, particularly for a plant that does not grow in most of the supposed Indo-European homelands—and instead favor semasiological correspondences, the first thing that stands out is the correspondence between Greek ἐρείκη ‘heather’ and ἐρείκω ‘to break, crush, tear apart, burst’. Regardless of whether it is identified as Erica arborea or Calluna vulgaris, either could have been used as a rough broom in earlier times as well (Nedelcheva et al. 2007). At present, I can offer no plausible proposal for the Slavic forms.

In any case, deriving Lith. vìržis from *urh́ǵ- has the advantage that the accent is explained as expected, in contrast to vìržis ‘rope on horse harness’ from *urǵh- (cf. LIV 688). Here the Latvian and Slavic forms also differ from Lithuanian in stress: Latv. vìrsis, Serbo-Croatian vrês, Russian ве́рек (Rudzīte 1993: 103).

With rēkšti ‘to pluck’, the interesting question arises whether the insertion of -k- is the cause of metatony. Such cases should be brought forward when considering whether rēkšti belongs with rēžti. The evidence mostly argues against metatony arising for this reason: brēkšti ‘to dawn’ < PIE *bʰreh́ǵ-, úokṣas < *tiosas (Smoczyński 2006: 141), plúokštas alongside plúoštas ‘fiber bundle, tuft’. Isolated counterexamples are bróžti ‘to scrape off-

6

Apr. riste, lit. rykšte, lett. rikste und Verwandtschaft

‘to strip, slide, drag one’s feet’ and brõkšti ‘to strip off, slide’, whereby influence from homophonous, onomatopoeically influenced brõkšti ‘to churn butter’: braškėti ‘to creak, crack’ may be invoked. Moreover, precisely with rẽžti the rare variant rẽžti is also attested, so that one cannot entirely rule out a formation deriving directly from rẽžti / rẽžti. On the other hand, rẽkšti has different present-tense forms, namely rẽškia and rẽškia (LKŽₓᵢ 415). This indicates that in the infinitive, generalized from rẽškė, this is a Baltic lengthened grade unrelated to laryngeal influence. This presupposes a new e-grade, formed only in Baltic or in an individual language, which, for example, was newly created on the basis of rýkšti ‘to decay, come apart, split open, fray’. It thus seems likely that k-insertion did not take place in the old -ē- und -úo- grades, but later, especially in the new zero grade. One indication of this is also the comparison of Old Prussian riste: Latvian rīkste, rīksts, rīksts, riste: Lithuanian rýkšte. The cause of this new i-grade was initially likely the perceived absence of a zero grade, after inherited forms with virž- or urž- were no longer regarded as belonging to the same family. Three scenarios. First: pure analogy; since a long vowel with acute is found in the e- and o-grades, this was adopted for the zero grade. The result is identical to what one would expect from *(u)rehₕ́-: *(u)rohₕ́-: *(u)rihₕ́-, so this process, too, cannot be ruled out as involving an early neo-zero grade. Second: initially there was a zero grade with short i *rīž⁶, from which y was formed as a lengthened grade, for example in past-tense forms. In that case, however, ȳ would initially be expected, and metatony must be assumed for ý, accompanied by a change in word class, since when the preterite root is generalized to the infinitive, the accent is retained (see the examples in Kazlauskas 1968: 351). Third, as above, except that the lengthened grade ȳ arose through contraction from -in- with an n-infix in the present, as for example in rýja, gýja, šlýja alongside the infinitives rýti, gýti, šlýti (ý from iH before C) and the preterite rìjo, gìjo, šlìjo (ì with laryngeal hiatus from iH before V) (Smoczyński 2003: 19, 40; Hyllested, Gliwa, in press). This, too, would require

⁶ Possibly Latvian riksna ‘the strip’ alongside riksna ‘same’ (ME₃ 525, 538). Short i is clearer, though its chronology is uncertain, in formations from the structurally identical Indo-European root *bʰrehₕ́-, namely brìžes ‘harrow’, brìžinti ‘to brush, rub, saw’, brìžterti ‘to draw, write’, alongside brýžis ‘lime, strip’, brẽžti ‘to score’, bruõžas ‘line, strip; trait, property’.

7

Bernd Gliwa

A metatony would be needed to account for the acute in rýkšte, rýšte. Metatony would be conceivable if the acute occurred only in nouns. But it is also found in rýkšti ‘to decay, come apart, split open, fray’ and in rýžti ‘to diminish, disappear; bud, swell; determine a dowry’. The variants with š instead of ž are likely to be explained by the sta-suffix (Smoczyński 2007: 507), while the present rýžta alongside rýža points to a -sta-infix—in this case, however, with generalization of the ž from the infinitive and preterite to the entire lemma.

The purely verbal formation rýžti ‘to diminish, fade (of a celestial body); disappear, hide’ from réžti ‘to cut off, destroy’, as well as the derivative rýkšti ‘to decay, come apart, split open, fray’, follows here the pattern láužti ‘to break (trans.)’: lūžti ‘to break (intr.)’, as well as plýšti ‘to tear, split, fray (intr.)’: plėšti ‘to tear, pluck (trans.)’. Here, too, the intransitive verbs have present forms with the -sta-suffix: lūžta and plýšta; no additional n-infix is discernible.

Following the same pattern as réžti ‘to incise, mark (make something clear)’, rýžti ‘to bud, swell (become apparent)’ and rýkšti ‘to become apparent; turn up, appear’ arose. By contrast, there seems to have been no semantic change in the pair (at)rýžti ‘to determine the dowry; provide with a dowry’: atrėžti (žemę, dalį) ‘to measure off, separate a piece of land; determine the dowry’; semantically compare also skirti dalį, krauti ‘to determine the dowry’ (literally: ‘to separate off’). Also belonging to this ablaut grade are Lithuanian ryžė, ryžmė, ryžtas ‘strip’. On this basis, a new o-grade rãžas ‘dry stalk, stubble; old broom’ arose on the one hand, along with further derivatives following the usual ablaut pattern e: a: i/y, for example šerkšnas, šarmas, širmas (Gliwa 2005). One should also add rakštis ‘(embedded) splinter (in the skin)’ and rãkštas, rakštis, rãkštis ‘coffin; grave, tombstone’⁸, originally perhaps a designation for a coffin carved out and hollowed from solid wood? On the other hand, there is a new o-grade of the kind expected in an i-ablaut series: Lithuanian raižyti ‘to incise, cut off’, formally an iterative to réžti ‘to incise, cut off’, structurally identical in every respect to braižyti ‘to draw’ from brýžis

⁷ The old zero grade *pḷHk- would yield Lithuanian *pilš- or *pùlš-; forms of this kind are absent.

⁸ Otherwise usually grouped with Lithuanian rakti ‘to poke’

8

Apr. riste, lit. rykštė, lett. rīkste und Verwandtschaft

‘line, strip’, brėžti ‘to score’, brūožas ‘line, strip; trait, property’ (Indo-European *bʰreh₁gʰ-) or plėšyti ‘to tear, split, fray (intr.)’ as an iterative to plýšti ‘to tear apart, split, fray (intr.)’, plėšti ‘to tear, pluck (trans.)’, as well as plėštas ‘wedge’, (at)plaiša ‘splinter, bast’, plaišinti ‘to split, blast’ and others, or braidýti ‘to wade’ from bristi ‘same’, brastà ‘ford’ (Indo-European *bʰredʰ-).

Above, both k-insertion and the use of the -sta-suffix were held responsible for the appearance of forms with -š- or -kš- instead of -ž-, -gž-. Moreover, ž can already become š before consonants⁹. For example, težė, težis, težlà, težlis, tešlė, tešlis ‘birch bolete’, from tėžti ‘to become slippery; soften; swell’, tižti ‘same’, tažėti ‘to drip, patter’, težėti ‘to drip, seep’, tāžinti ‘to sprinkle with drops’, težlynė ‘bog, mire’ and tẽšlas ‘dough’, as well as, through k-insertion, tekšoti ‘to stand (of a body of water, puddle)’, tãškas ‘splash, spot’ (Gliwa 2006: 64ff.). Or lūžnà, lūšnà ‘hut, cottage’ from lūžti ‘to break, disintegrate (intr.)’, laušiuoti ‘to break, smash, beat’: láužti ‘to break (trans.)’; gùštà: gùžtà ‘nest’, gùštáuti: gùžtáuti ‘to seek a place for a nest’; gošti: gõžti ‘to thrive, spread (of plants), overgrow’.

Given this, the question arises whether Lithuanian to complete, prepare and Latvian to get things done ‘complete, prepare; be active, act’ can also be assigned to the family under discussion—especially since there are only uncertain possibilities for comparison outside Latvian and Lithuanian. The circumflex is no reason for exclusion, but, as also in the case of to incise, scratch ‘incise, scratch; sweep out’, could point to a denominal formation. Initially, both Baltic languages show only one ablaut grade—if we set aside the comparison with Lithuanian tumult, revolt, coup (LEW 750). Further comparison with Polish movement etc. (LEW 750, LEW^II 129) presupposes a PIE root *reus-, with the operation of the ruki rule in Slavic and Baltic¹⁰. This root can then be identified with *reus- ‘to rummage’, in which case Lithuanian to scratch, rummage appears as the Baltic form for comparison (LIV 511; Boryś 2005: 528). Tumult as ‘commotion’ can also be accommodated here quite well semantically, albeit figuratively,

⁹ The precise conditions for this, and any regional tendencies, would be a subject for more detailed investigation.

¹⁰ After au in Lithuanian aušra, as against auksas.

9

BERND GLIWA

but what about the other Baltic forms? Especially since Baltic *rō(u)š- would then also need to be accounted for¹¹, or rather the transition from (East) Baltic *rauš- to ruoš-.

Semantically, two possibilities suggest themselves. On the one hand, starting from ruõžti ‘sweep out’, generalized to ruõšti ‘carry out work around the house and yard’, or ruõšti ‘sweep out → prepare for the reception of guests, celebrations → prepare in general’. On the other hand, the semantic correspondence ryžtis: ryžtingas, pasirýžęs¹² alongside ruošis: pasiruošęs is striking.

The possible semantic developments of rýžti have already been discussed above. In this case, rėžti ‘incise, mark (make something clear)’ → rýžti ‘become clear’ seems the most fitting. Thus, originally, the meaning was not preparing something in a qualitative sense, but rather the becoming clear of an intention or attitude, as is evident, for example, in the sentence žąsinas ryžtai atsigręžė į berniuką ‘the gander turned energetically toward the boy’ or in the combination ryžtas balsas ‘clear voice’ (LKŽ_XI 774). Since such meanings, however, are essentially restricted to the reflexive forms, at least in the case of ruošti one should assume the first-mentioned semantic development—provided the connection with ruõžti is justified.

Another large word family in Baltic that can be interpreted as developing from *ureh̑ǵ- comprises the terms around Lithuanian rãštas, Latvian raksts ‘writing, script’. According to Karulis (LEV_II 103), Karaliūnas had already proposed this, although this proposal does not appear in the bibliography.

Karulis views raksts as a formation from the verb rakstīt ‘write’, which was formed as an iterative of rakt, Lithuanian ràkti ‘dig, scratch, poke’, following the pattern lēkt: lēkāt: lakstīt (LEV_II 102). One would therefore actually have to separate the Lithuanian terms from the Latvian ones, since in that case one would expect not rašyti but †rakstyti, following precisely

¹¹ Die ausführliche Untersuchung von Karaliūnas (1987: 174–215) zur Problematik der Apophonie uo: au verzichtet zwar auf Laryngale und ist daher in mancher Hinsicht erneuerungsbedürftig. Allerdings brächte dies kaum neue Erkenntnisse bezüglich der erst (ost)baltischen Apophonie uo: au.

¹² With the typical metatony in the transition from adjective to noun: rýžtas ‘brave, energetic’ (adjective) → rýžtas ‘courage, will to assert oneself, responsibility’, without requiring a change of stem (Hyllested, Gliwa, in press)

10

Apr. riste, lit. rykštė, lett. rikste und Verwandtschaft

that pattern in Lithuanian: lėkti: lakštyti. To avoid this, Fraenkel invokes the influence of piešti (LEW 702). As Fraenkel (LEW 702) also acknowledges, the pair Lithuanian raštas, Latvian raksts actually invites the supposition of an infixed k, as is especially common in Latvian, cf. Lithuanian pìrstas: Latvian pirksts ‘finger, toe’. “Yet it is advisable [...] to connect the Latvian words for ‘write’ and ‘writing’ with Lithuanian rãkti ‘poke with a pointed object, dig out’, Latvian rakt ‘dig’, etc.” (LEW 702). The reason for this recommendation is not stated, but apparently it is because no suitable comparisons can be made for a root *raš-. Smoczyński (2007: 500f.) sums up: “Bez dobrej etymologii”.

As discussed above, under suitable conditions š can develop from ž. Semantically too, there is no significant difference from the proposal to connect raštas with rãkti ‘poke...’. The base rėžti ‘scratch’ is equally suitable, indeed more so. A standard semasiological example is Latin scribere ‘write’ ← ‘scratch, scrape’; further examples are given by Karulis (LEV_II 102f.). The transition in the ablaut series to forms with a alongside older ē, uo is also indisputably attested here by rãžas ‘dry stalk...’. However, I consider the starting point for the nouns raštas, rašmuo ‘incised mark, sign’ to be probably deverbal or derived from a participle of *rašti, *ražti, cf. rėžis ‘incised mark, cut; line...’, whereas the verbs Lithuanian rašyti, Latvian rakstīt ‘write’ are denominal, cf. baslỹs: bãšlyti, burnà: bùrnyti, ginčas: ginčyti, among others (Skardžius 1943: 535). Finally, the word family around rãkti contains no further trace pointing toward a development in the direction of signs, symbols, or writing, in stark contrast to rėžti.

If one connects rykšti ‘become clear; be found, appear’ with rėžti, as done above, then one cannot help connecting the entire family with ryškùs, rá iškus, raiškùs ‘clear, noticeable, conspicuous’, ryškėti ‘become clear’, raiškà ‘expression’, rėikšti ‘mean’, especially since these have no Latvian counterparts.

To return to the starting point of the investigation—rykšte—derived from this is Lithuanian rykštėnė ‘Solidago sp.’. The only question is whether this is a folk name or a name from books. The equivalents are Lithuanian tvira, auksarykštė (LBŽ), Latvian zeltgalvītes, zeltslotiņa, zeltrīste, zelta rīkstes, among others (EO 457), German Goldrute, and many others. Since the only native species of Solidago

11

Bernd Gliwa

virgaurea is¹³ and, moreover, the modern epithet Virga aurea was also used as a botanical (genus) name (Genaust 1996: 685); Latin virga ‘rod, switch’, aureus ‘golden’. In many cases it is difficult to decide whether this is a calque, for example auksarykštė, or merely names motivated in the same way (cf. Marzell 2000: IV 390f.). After all, the term rykštė is used in plant names, e.g. Šv. Jurgio rykštė, Juozapa rykštė, Arono rykštė, Erodo rykštė ‘Gladiolus sp.’ (Genelytė 2004: 53f.; LKŽ_XI 611), which are obviously not terminologically formed. tvira, tvyra, on the other hand, can certainly be regarded as a book name¹⁴, as a genus name and equivalent of Solidago: solidus ‘solid’. The same applies to rykštinis in binomial botanical names; it is simply a translation of Latin virgata.

Finally, it should also be noted that connecting rykštė with rėžti rather than with rišti finally permits a reasonable interpretation of the theonym Diviriks. The few attestations come from 13th-century Russian chronicles and, as most researchers agree, despite the absence of incontrovertible evidence, refer to the thunder god Perkūnas. Catholic theology is then no longer needed to explain the pagan theonym as *Dievo rykštė ‘God’s scourge’; instead, the thunder god can be understood as what he immediately appears to be: *Dievarykštis ‘God-striker’, who flashes, thunders, and strikes. For unlike the family of words around rišti ‘to bind’, rėžti ‘to scratch, strike’ is also explicitly used in descriptions of thunderstorms: Griausmas kaip rėžė, ir užmušė kumelę ‘The thunderstorm struck and killed the mare’ (LKŽ_XI 514; Gliwa 2008).

¹³ A fairly common and handsome plant, up to 1 m tall, which is also said to have had folk names because of its use as a wound-healing and medicinal plant. Unfortunately, no entries can be found in old dictionaries.

¹⁴ The primary sources, namely botanists, did not provide stress markings. The stress information in LKŽ_XVII 305, following Niedermann’s dictionary of the Lithuanian literary language, may be plausible, but it suggests a deceptive degree of precision.

12

Apr. riste, lit. rykštė, lett. rikste und Verwandtschaft

LITERATURE

BGD = [G. E.] BENSELER, Benseler’s Greek-German Dictionary, edited by A. KAEGI. 17th edition – unchanged reprint of the 15th edition, 1931. Leipzig: Verlag Enzyklopädie, 1990.

BORYŚ, W. 2005: Słownik etymologiczny języka polskiego. Kraków: Wydawnictwo Literackie.

BRÜCKNER, A. 1957: Słownik etymologiczny języka polskiego. Warszawa: Wiedza Powszechna.

EO = I. ĒDELMANE, A. OZOLA, Latvian Plant Names. Riga: Augsburgas institūts, 2003.

FENNELL, T. G. 2000: Fürecker’s Dictionary: A Concordance II N–Ž. Rīga: Latvijas akadēmiskā bibliotēka.

FENNELL, T. G. 2002: A Concordance to the Manuale Lettico-Germanicum II P–Ž. Rīga: Latvijas akadēmiskā bibliotēka.

GENAUST, H. 1996: Etymologisches Wörterbuch der botanischen Pflanzennamen. Basel–Boston–Berlin: Birkhäuser.

GENELYTĖ, A. 2004: Liaudiškų fitonimų motyvacija (šiaurės panevėžiškių ir kitų lietuvių kalbos tarmių lyginamasis aspektas). Daktaro disertacija. Vilnius: Lietuvių kalbos institutas.

GEW = H. FRISK, Griechisches etymologisches Wörterbuch I. Heidelberg: Winter, 1960.

GLIWA, B. 2005: Baltische Bestattungsrituale im Spiegel ihrer Bezeichnungen: litauisch šermenys und šarvojimas, altpreussisch sirmen. Acta Linguistica Lithuanica 53, 9–21.

GLIWA, B. 2006: Studien zu litauischen Pflanzennamen im indogermanischen Kontext. Rozprawa doktorska. Kraków: Uniwersytet Jagielloński.

GLIWA, B. 2008: Der litauische Gottesname Diviriks. In memoriam V. Toporov. Beiträge zur Namenforschung 43 (1), 25–35.

GLIWA, B., in press: Apr. tubo – filcz (E 448). Res Balticae.

13

Bernd Gliwa

Hyllested, A., B. Gliwa, in press: Metatony in Lithuanian suffixless derivatives. In: Th. Olander, ed., Stressing the past. Copenhagen.

Karaliūnas, S. 1987: Baltų kalbų struktūrų bendrybės ir jų kilmė. Vilnius: Mokslas.

Kazlauskas, J. 1968: Lietuvių kalbos istorinė gramatika. Vilnius: Mintis.

LBŽ = J. Dagys, red., Lietuviškas botanikos žodynas. Kaunas: L. Vailionis, 1938.

LEV = K. Karulis, Latviešu etimoloģijas vārdnīca I–II. Rīga: Avots, 1992.

LEW = E. Fraenkel, Lithuanian Etymological Dictionary II. Heidelberg–Göttingen: Winter, Vandenhoeck & Ruprecht, 1965.

LIV = H. Rix et al., Lexicon of Indo-European Verbs. The Roots and Their Primary Stem Formations. 2nd ed. Wiesbaden: Reichert, 2001.

LKŽ = Dictionary of the Lithuanian Language I–XX. Vilnius: Mintis, Institute of Science and Encyclopedias Publishing, Institute of the Lithuanian Language, 1956–2002.

Marzell, H. 2000: Wörterbuch der deutschen Pflanzennamen I–V. Köln: Parkland.

Mažiulis, V. 1997: Prūsų kalbos etimologijos žodynas IV. R–Z. Vilnius: Mokslo ir enciklopedijų leidybos institutas.

ME = K. Muhlenbachs, Latvian-German Dictionary I–IV. Edited and continued by J. Endzelīns. Riga, 1923–1932.

Nedelcheva, A.M., Y. Dogan, P.M. Guarrara, 2007: Plants traditionally used to make brooms in several European countries. Journal of Ethnobiology and Ethnomedicine 3: 20 (http://www.ethnobiomed.com/content/3/1/20).

REW = M. Vasmer, Etymological Dictionary of the Russian Language I. Translated from German and supplemented by O. N. Trubachev. Moscow: Progress, 1964.

Rudzīte, M. 1993: Latviešu valodas vēsturiskā fonētika. Rīga: Zvaigzne.

SD₁ = Szyrwid, C. 1621. Dictionarium trium linguarum. Vilnius: Acad. Jesu. Faksimile in: K. Pakalka, Hrsg., Senasis Konstantino Sirvydo žodynas. Vilnius: Mokslo ir enciklopedijos leidybos institutas, 1997.

Skardžius, P. 1943: Lietuvių kalbos žodžių daryba. Vilnius: Lietuvių kalbos institutas.

14

Apr. riste, lit. rykšte, lett. rikste und Verwandtschaft

Smoczyński, W. 2003: Hiat laryngalny w językach bałto-słowiańskich. Kraków: Wydawnictwo Uniwersytetu Jagiellońskiego.

Smoczyński, W. 2006: Laringalų teorija ir lietuvių kalba. Vilnius: Lietuvių kalbos institutas. [Bibliotheca Salensis, 2.]

Smoczyński, W. 2007: Słownik etymologiczny języka litewskiego. Wilno: Uniwersytet Wileński.

Stang, C. 1966: Vergleichende Grammatik der baltischen Sprachen. Oslo, Bergen, Tromsø: Universitetsforlaget.

Zander, R. 2000: Handwörterbuch der deutschen Pflanzennamen. 16. Auflage in Bearbeitung von W. Erhard, E. Gotz, N. Bödecker, S. Seybold. Stuttgart: Ulmer.

Bernd Gliwa

Latvijas universitāte

Baltu valodu katedra

Visvalža iela 4a, LV 1050 Rīga

[email protected]

15