HARALD BICHLMEIER

Saxon Academy of Sciences in Leipzig

Martin Luther University Halle-Wittenberg

Research fields: onomastics (toponymy), historical phonology of the Germanic languages, Indo-European linguistics, Aljamiado-texts.

DOI: doi.org/10.35321/all83-07

NOTES ON THE RIVER NAMES LITH. LIĖKĖ, LIEKĄ AND THEIR FAMILY

Some remarks on Liėkė, Lieką and other river names with the same root

ANNOTATION

The river names Lith. Liėkė, Lieką (and their family) occasionally occur in studies of ‘Old European’ hydronymy. Traditionally, an etymology based on the root PIE *(h₁)leik- ‘to bend, curve’ has been attributed to them. However, this root probably never existed; at least, it is not attested in the appellative vocabulary of any Indo-European language. Consequently, if the PIE root *(h₁)leik existed at all, no meaning can really be attributed to it. It therefore seems more reasonable to derive the two Lithuanian river names from the PIE root *uleikʷ- ‘to be fluid, to moisten’. Connecting them with Baltic lexical material is difficult, but possible.

At the same time, regarding the name of the river Lech, for whose etymologization the two roots mentioned have likewise been invoked, it must be noted that it cannot be derived from the PIE root *(h₁)leik because that root probably did not exist at all, but that the PIE root *uleikʷ is also ruled out because phonological reasons militate against it. Rather, this name is a Celtic formation meaning ‘the one with stone slabs, the stony one’: PIE *pḷ(h₂)k-(m)neh₂- > Early Celtic *φlikkā- > Common Celtic *likkā.

The conclusion is that Lith. Liėkė, Lieką etc. and German Lech are etymologically unrelated.

Straipsniai / Articles

151

HARALD BICHLMEIER

KEYWORDS: Old European hydronymy, Celtic hydronyms, Lithuanian hydronyms, ‘false friends’.

ANNOTATION

The river-names Lith. Liėkė, Liekà (and their siblings) are sometimes mentioned in papers dealing with the so-called ‘Old-European hydronymy’. Traditionally they are etymologized on the basis of a root PIE *(_h₁)leik- ‘bend’. This root, however, probably never existed, at least there seem to be no descendents from it in the appellatival lexicon of any IE language. Thus the root – if it ever existed – cannot be ascribed a meaning. So it seems more appropriate to derive the Lithuanian river-names from the root PIE *uleik^ᵘ- ‘moisten, be liquid’.

Parallelly the river-name Lech (a river in Austria and Bavaria) was also analyzed based on the two roots mentioned, PIE *(_h₁)leik- ‘bend’ and PIE *uleikᵘ- ‘moisten, be liquid’. As just mentioned, the first one, however, never existed – and the second one cannot be the starting point because of phonological reasons. The river-name Lech is based on a Celtic formation ‘the one with stone(-plate)s’: PIE *pl̥(h₂)k-(m)neh₂- > PCelt. *(q)likkā-.

Thus it is clear that the river-names Lith. Liėkė, Liekà etc. and Germ. Lech do not have a common etymology.

KEYWORDS: Old European hydronymy, Celtic hydronyms, Lithuanian hydronyms, ‘false friends’.

1. INTRODUCTION

The two river names Lith. Liėkė, Liekà and their family occur in various works on the older layers of water names in Europe, that is, on ‘Old European hydronymy’. Under certain conditions, a connection between these names and the name of the river Lech, which flows from Austria to Bavaria and there empties into the Danube, could also be established, as was indeed done in older literature. The aim of this contribution is to examine whether such a connection is justified.^¹

^¹ I thank the two anonymous reviewers of the contribution for their comments and suggestions, which helped to dispel misunderstandings and generally improve the contribution.

152

Acta Linguistica Lithuanica LXXXIII

Notes on the Lithuanian river names Liĕkė, Liekā and their family

2. THE LITHUANIAN NAME FAMILY

2.1. As regards etymologization and the other names to be connected, one must ultimately start from Lith. lieknas (also lieknas) m. ‘marsh, swampy, boggy meadowland’, liėkna (also lieknā) f. ‘low-lying swampy meadow’. Lith. lāikšės (1) f. ‘large swamp overgrown with brush’ probably also belongs here (for further details on the etymology, see below § 3.).

2.2. As for the names belonging to this family, several lists must be compared:

In addition to the ‘classical’ Lithuanian literature (especially Savukynas et al. 1963, Vanagas 1970, 1981, 1981a), a list should be mentioned that is given in SEJL² (2018–2020) s.v. lāikšės (in LED 656 s.v. lāikšės, however, the water names are not listed; instead, they appear in LED 696 s.v. lieknas), although comparison with the Lietuvos vietovardžių geoinformacinė duomenų bazė (http://lkiis.lki.lt/lietuvos-vietovardziu-geoinformacine-duomenu-baze; hereafter LVGDB) makes the attribution of the names to geographical objects carried out there occasionally appear questionable.

2.3. According to Smoczyński, the following names are based on the Lithuanian root leik–/liek–:

2.3.1. River names based on leik–/liek–:

a) Derivatives:

Liėke (2),^² Liėkė (4),^³ Liekā (4);^⁴

b) Compounds:

Lēikupis (1).^⁵

2.3.2. River names based on leikn–/liekn–:

a) Derivatives:

^² Cf. Savukynas et al. 1963: 91; not in Vanagas 1970; cf. Vanagas 1981: 189; Vanagas 1981a: 59; not in LVGDB.

^³ Cf. Savukynas et al. 1963: 91; not in Vanagas 1970; cf. Vanagas 1981: 189; Vanagas 1981a: 59; not in LVGDB.

^⁴ Cf. Savukynas et al. 1963: 89; Vanagas 1970: 46; Vanagas 1981: 185, 189; Vanagas 1981a: 59; not in LVGDB.

^⁵ Cf. Savukynas et al. 1963: 89; Vanagas 1970: 241; Vanagas 1981: 189; Vanagas 1981a: 59; not in LVGDB.

Straipsniai / Articles

153

HARALD BICHLMEIER

Liekšna (1);^⁶ 7x in LVGDB, including no water name: 116597^⁷ (meadow), 116664 (meadow), 116866 (meadow), 117072 (meadow), 117766 (meadow), 117807 (meadow), 118043 (forest).

Liekšnas (1; 4x)^⁸.

Liekšnas (2; 3x);^⁹ 5x in LVGDB, including no water name: 116808 (meadow), 117138 (meadow), 117244 (meadow), 117700 (meadow), 117767 (meadow).

b) Compounds: Liekšnupalis,^¹⁰ Liekšnupis,^¹¹ Pusiáulieknis^.¹²

2.3.2. Place names based on liekn-/liekn-:

a) Derivatives: Liek̨niai^¹³ (in LVGDB as a farm-part name [118387]); Liek̨nėlis,^¹⁴ Ūžliek̨nė (apparently a forest name),^¹⁵ Ūžliek̨nis,^¹⁶ Ūžliek̨niai (1)^¹⁷ (11748);

b) Compounds: Liek̨nag̨ala^,¹⁸ Jūodliek̨nis^,¹⁹ Šilaliek̨nis^.²⁰

2.4. In Smoczyński’s list cited above, the Liek̨a regularly mentioned in the hydronymic literature is not included, nor is it recorded in LVGDB or the ‘classical’ Lithuanian literature on the subject. According to the current state of research, much suggests that the water name Lith. Liek̨a

^⁶ Cf. Savukynas et al. 1963: 91; Vanagas 1970: 36; Vanagas 1981: 190.

^⁷ The numbers cited here and below are the geocodes assigned to the individual geographical objects in LVGDB.

^⁸ Cf. Savukynas et al. 1963: 91; Vanagas 1970: 36, 115; Vanagas 1981: 190; name form with acute not in LVGDB.

^⁹ Cf. Savukynas et al. 1963: 91; Vanagas 1970: 36; Vanagas 1981: 190.

^¹⁰ Cf. Savukynas et al. 1963: 91; Vanagas 1970: 89; Vanagas 1981: 190; not in LVGDB.

^¹¹ Cf. Savukynas et al. 1963: 91; Vanagas 1970: 235; Vanagas 1981: 190; not in LVGDB.

^¹² Cf. Savukynas et al. 1963: 130; Vanagas 1970: 254, 256; Vanagas 1981: 269; not in LVGDB.

^¹³ Not in Savukynas et al. 1963; not in Vanagas 1970; not in Vanagas 1981.

^¹⁴ Cf. Savukynas et al. 1963: 91; Vanagas 1970: 115; Vanagas 1981: 190; not in LVGDB.

^¹⁵ Cf. Savukynas et al. 1963: 181 (Ūžliek̨nės ūpis); Vanagas 1970: 272 (Ūžliek̨nės ūpis); Vanagas 1981: 356 (Ūžliek̨nės ūpis): concerning the forest name Ūžliek̨nė; not in LVGDB.

^¹⁶ Not in Savukynas et al. 1963; not in Vanagas 1970; not in Vanagas 1981; not in LVGDB.

^¹⁷ Not in Savukynas et al. 1963; not in Vanagas 1970; not in Vanagas 1981.

^¹⁸ Not in Savukynas et al. 1963; not in Vanagas 1970; not in Vanagas 1981; not in LVGDB.

^¹⁹ Not in Savukynas et al. 1963; cf. Vanagas 1970: 251, 256; not in Vanagas 1981; not in LVGDB.

^²⁰ Cf. Savukynas et al. 1963: 185; Vanagas 1970: 239, 247; Vanagas 1981: 331; not in LVGDB.

154

Acta Linguistica Lithuanica LXXXIII

Notes on the Lithuanian river names Liëke, Liekà and their family

that it is a Ghostname, perhaps based on a scribal error for Leikà. It is not possible for the time being to determine who first introduced it into scholarly discussion.

In addition to the list cited above, LVGDB records the following derivatives or onymizations of the Lithuanian appellatives liėknas, liėkna:

2x Liek̇nalis (2):²¹ 116716 (meadow), 116824 (meadow);

Lieknāle (2):²² 116997 (field);

Liek̇nai (2):²³ 116697 (pasture);

Liek̇iai (2):²⁴ 76426 (village);

4x Liek̇nos (1):²⁵ 117693 (meadow), 117701 (meadow), 117506 (swamp), 117507 (meadow);

Liek̇niniai pašėdžiai:²⁶ 116394 (meadow);

3x Liek̇nė (2):²⁷ 116370 (meadow), 116395 (meadow), 116613 (meadow);

Liek̇nės (2):²⁸ 116711 (meadow).

2.5. Furthermore, the literature on Lithuanian water names cites several water names offered neither by Smoczyński nor in LVGDB, namely Liekupis twice,²⁹ as well as Liekaitis, Liekaitė, Liėkno upalis and Lieknėlių rãvas once each.

2.6. For the time being, it is unclear how the discrepancy between the information in the database and the information in the name books (insofar as they concern water names) arose.

2.7. In addition, Udolph (1990: 116), among others, also associates with this group two water names with a zero-grade root form, Lith. Li̇kupis, Likupỹs. In this connection (and regarding names with Liek-, Leik-), Udolph also cites, among others, Vanagas

²¹ Not in Savukynas et al. 1963; not in Vanagas 1970; not in Vanagas 1981.

²² Not in Savukynas et al. 1963; not in Vanagas 1970; not in Vanagas 1981.

²³ Not in Savukynas et al. 1963; not in Vanagas 1970; not in Vanagas 1981.

²⁴ Not in Savukynas et al. 1963; not in Vanagas 1970; not in Vanagas 1981.

²⁵ Not in Savukynas et al. 1963; not in Vanagas 1970; not in Vanagas 1981.

²⁶ Not in Savukynas et al. 1963; not in Vanagas 1970; not in Vanagas 1981.

²⁷ Not in Savukynas et al. 1963; not in Vanagas 1970; not in Vanagas 1981.

²⁸ Not in Savukynas et al. 1963; not in Vanagas 1970; not in Vanagas 1981.

²⁹ Cf. Savukynas et al. 1963: 91; Vanagas 1981: 189.

³⁰ Cf. Savukynas et al. 1963: 91; Vanagas 1970: 76; Vanagas 1981: 189.

³¹ Cf. Savukynas et al. 1963: 91; Vanagas 1970: 76; Vanagas 1981: 189.

³² Cf. Savukynas et al. 1963: 91; Vanagas 1970: 270; Vanagas 1981: 190.

³³ Cf. Savukynas et al. 1963: 91; Vanagas 1970: 272; Vanagas 1981: 190.

³⁴ Cf. Savukynas et al. 1963: 92; Vanagas 1970: 241, 244; not in LVGDB.

Straipsniai / Articles

155

HARALD BICHLMEIER

1981: “185, 189f. – the forms with Lik are found, however, on p. 191 and are in no way connected there with the names containing Liek-, Leik; nor is a connection with the Lithuanian personal name Likas considered probable; by contrast, a connection with the Lithuanian verb likti appears possible in the same place. Yet a connection between the water names and Lith. likti ‘to remain, remain over, remain alive, leave, leave behind’ is hardly a particularly obvious proposal. Such a proposal would appear more plausible in an area where rivers disappear during the dry season—for there ‘the river that remains (despite the drought) (i.e. continues to carry water)’ would be a meaningful naming motive. For the Baltic region, this is provisionally rather unlikely.”

The connection can therefore at most be classified as a remote possibility.

2.8. Summarized in a table, the following picture emerges:

Name ^SEJL² LVGDB Savukynas et al. 1963 Vanagas 1970 Vanagas 1981 Vanagas 1981a Juódliek­nis + (ON) – – – – – Leikà + (FIN) – 89 46 185, 189 59 Liekáitė – – 91 76 189 – Liekáitis – – 91 76 189 – Leikupis + (FIN) – 89 241 189 59 Liẽkė + (FIN) – 91 – 189 59 Liekė̃ +(FIN) – 91 – 189 59 Liẽkiai – 76426 (village) – – – – Liẽkna + (FIN) 116597 (meadow), 116664 (meadow), 116866 (meadow), 117072 (meadow), 117766 (meadow), 117807 (meadow), 118043 (forest) 91 36, 115 190 – Liẽknagala + (ON) – – – – – Liẽknai – 116697 (pasture) – – – – Liek­nãlė – 116997 (field) – – – – Liek­nãlis – 116716 (meadow), 116824 (meadow) – – – – Liẽknas + (FIN) 116808 (meadow), 117138 (meadow), 117244 (meadow), 117700 (meadow), 117767 (meadow) 91 36 190 –

156

Acta Linguistica Lithuanica LXXXIII

Notes on the Lithuanian river names Liėkė, Liėką and their family

Name | SEJL^² | LVGDB | Savukynas et al. 1963 | Vanagas 1970 | Vanagas 1981 | Vanagas 1981a Liėknė | – | 116370 (meadow), 116395 (meadow), 116613 (meadow) | – | – | – | – Liėknėlių rāvas | – | – | 91 | 272 | 190 | – Liėknėlis | + (ON) | – | 91 | 115 | 190 | – Liėknės | – | 116711 (meadow) | – | – | – | – Liėkniaĩ | + (ON) | 118387 (farm part) | – | – | – | – Liėkniniai pasėdžiai | – | 116394 (meadow) | – | – | – | – Liėkno upālis | – | – | 91 | 270 | 190 | – Liėknos | – | 117693 (meadow), 117701 (meadow), 117506 (swamp), 117507 (meadow) | – | – | – | – Liėknupalis | + (FlN) | – | 91 | 89 | 190 | – Liėknupis | + (FlN) | – | 91 | 235 | 190 | – Liėkupis | – | – | 91 | – | 189 | – Pusiālieknis | + (FlN) | – | 130 | 254, 256 | 269 | – Šilalieknis | + (ON) | – | 185 | 239, 247 | 331 | – Užlieknė | + (ON), more precisely: Wa | – | 181 (Užlikenės ūpis) | 272 (Užlikenės ūpis) | 356 (Užlikenės ūpis) | – Užliekniai | + (ON) | 117748 | – | – | – | – Užlieknis | + (ON) | – | – | – | – | –

Abbreviations: A – field, FlN – river name, Ht – farm part, ON – place name, Su – swamp, Vw – pasture, Wa – forest, Wi – meadow

2.9. The tabular overview reveals further discrepancies: it is surprising how many toponyms (in the broader sense) are not recorded in LVGDB. In some cases, different classifications are made; for example, Liėknėlis is regarded by SEJL^² as a place name, but in the Lithuanian specialist literature as a water name. Here, quite a few things still need to be thoroughly rechecked; one or another name form may also yet prove to be a Ghostname.

Straipsniai / Articles

157

HARALD BICHLMEIER

3. ON THE ETYMOLOGY OF LITH. LIEKĖ, LIEKĄ ETC.

3.1. In the context of ‘Old European hydronymy’, only the two shortest forms, Lith. Liekė, Lieką, are usually mentioned. Two different etymologies have so far been attributed to these names, which will be presented and weighed against each other below: a traditional one and a more recent one.

On the one hand, the question of the etymology of the name of the river Lech, which flows from Austria to Bavaria, is closely intertwined with the etymology of these two river names; on the other hand, so is the question of the etymology of a series of Slavic river and area names.

3.2. In all cases, two Proto-Indo-European roots compete here as the basis of derivation, *uleikᵘ- ‘to be fluid, to moisten’ (the more recent proposal) and *(h₂)leik- ‘to bend, curve’ (the traditional view). From a semantic perspective, derivations from both roots appear quite suitable for naming flowing waters: certainly named after a characteristic bend or their abundance of curves are the Chamb (OHG Kamb [1058]; < OHG *kamp- ← Gaulish/Common Celtic *kambo- ‘bent’³⁵),³⁶ probably the Bogenbach (also place name Bogen; Lower Bavaria; < OHG Pogana, Bogana [PN: ca. 790 (copy 1254)], water name OHG Bogana [882])³⁷, perhaps the Memel (Lith. Nẽmunas), if the traditional etymology is correct,³⁸ and perhaps the Elbe, if it is nevertheless to be etymologized as PIE *h₂elh₂-bʰ(h₂)-o- ‘the meandering one’ rather than, in the traditional manner, as ‘the white one’.³⁹ And names based on roots meaning ‘moist, wet, fluid’ are famously very numerous; cf. the various German water names Neetze, Nette, Netze, which are to be assigned to the Proto-Germanic root *nat- ‘wet’,⁴⁰ in Austria the Mattig (probably from [Pre-]Celtic *madu-ko to the root

35 Cf. EDPC 186.

36 Cf. DGNB 81f.

37 Cf. Prinz 2007: 158–164; DGNB 67; Wiesinger, Greule 2019: 101f.

38 Cf. Bichlmeier 2019: 49f.; 2020: 130f., each with older literature. Similarly, already citing the two etymological proposals mentioned by Bichlmeier, Vanagas 1988: 28f.

39 Cf. Bichlmeier 2012, 2013; Bichlmeier, Blažek 2014, 2015; Scheungraber, Grünzweig 2014: 43–46 (each with further literature).

40 Cf. DGNB 372, 374f.

158

Acta Linguistica Lithuanica LXXXIII

Notes on the Lithuanian river names Liẽkė, Lieką and their family

PIE *mad- ‘to be/become wet’ [LIV² 421]⁴¹ and probably also the Sava / Save, which is most likely formed from – *sh₂eu̯ ‘to pour, rain’ (cf. IEW 912; LIV² 545),⁴² etc.

3.3. Before considering the potentially related forms outside the Baltic area, the internal Baltic relationship of the names should first be explored.

3.3.1. In general, the two Lithuanian river names and their family are connected with Lith. lieknas (also liėknas) m. ‘marsh, swampy, boggy meadowland’, Latvian liekns m. ‘lowland, low-lying damp meadow’, Lith. liėkna (also lieknā) f. ‘low-lying swampy meadow’, Latvian liėkna f. ‘swamp, lowland between two elevations, damp swampy place in the forest, swampy forest, large damp meadow’, Latvian liėkne f. ‘endless meadow along the river mouth’. Lith. laíkšės (1) f. ‘large swamp overgrown with brush’ should probably also be included.⁴³

3.3.2. The following preforms are to be reconstructed from the attested Lithuanian and Latvian word forms: Common Baltic *leiknā–/*laiknā–, *leikna–/*laikna–, *leikniā/*laikniā– and finally *laik-s-iā.

3.3.3. Semantically, this connection presents no problem. Initially, however, the intonation of the root syllable is problematic, since the circumflex in the water names contrasts with the acute of the appellative (which is the form most often encountered). The difference is easily explained if the acute is original and one starts from Common Baltic *leik- (/*laik-) > Proto-East Baltic *lēk. In this case, the circumflex in the water names could be explained as having arisen through metatony (métatonie douce) in the water name Lith. Liẽkė < Common Baltic *Leikiā and then having been transferred to the water name Lith. Lieką, where metatony could not actually take place. The question now is: Where did the root in the appellatives get its primary acute? In fact, there are only two alternatives here: either the root contained in

⁴¹ Cf. DGNB 340f., Wiesinger, Greule 2019: 126.

⁴² Cf. Bichlmeier 2011: 65–74; Repanšek 2016: 121f., Bichlmeier 2017, each with older literature.

⁴³ Cf. LitEW 1: 332, 364f.; Karulis 1992/2001: 523f. (with an outdated etymology). In SEJL 350, Smoczyński additionally writes: “Bez etymologii.” – For the general issue, see now SEJL² (2018–2020) s.v. liėknas, laíkšės, where the etymology based on PIE *uleik̯u is now actually preferred. This etymological connection is ultimately already about a century old (which Bichlmeier 2010 also failed to recognize); cf. LitEW 1: 332, where Fraenkel cites the older literature but rejects this proposal. Smoczyński likewise has no explanation for the acute; in SEJL² (2018–2020) s.v. laíkšės, we read: “Acute without historical justification.” In LED 656, 696 (where PIE *uleik̯u is now also regarded as the basis), the problem is not raised at all.

Straipsniai / Articles

159

HARALD BICHLMEIER

Proto-Indo-European contained a laryngeal, so it should be posited as PIE *(u)leiHᵏ⁽ᵘ⁾-,⁴⁴ or the formation has a (possibly Common Baltic) developed lengthened grade. In this case, (Pre-)Common Baltic *(u)lei(H)ᵏ⁽ᵘ⁾- must be postulated, with the assumption of a laryngeal in the root being unnecessary here. According to SEJL² (2018–2020) s.v. laĩkės, the “acute without historical justification” is found. One may therefore assume that the author likewise regards it as somehow analogical.

3.3.4. Since the root underlying the etymologization of the appellatives and consequently also of the place and water names allows the emergence of an acute only to a limited extent, it appears that the names preserve the older state (with circumflex), whereas the appellatives have been renewed. If this analysis is correct, this would at least be an unusual situation.

3.4. What, then, is the evidence for the two roots assumed as the basis?

3.4.1. The PIE root *uleikᵘ- is well attested, for example in Lat. liquēre ‘to be fluid’, liquidus ‘fluid’, etc. Regarding the putative PIE root *h₁leik- ‘to bend, curve’, Bichlmeier (2010: 44f.) probably said what was necessary:

“The second position [as opposed to the view that it is precisely a continuation of the PIE root *uleikᵘ-; H.B.] is represented, for example, by Anreiter (2005: 45; 2005a: 16; 2008: 150) and Anreiter/Haslinger/Roider (2000: 131), who assume a derivation from pre-Roman *lik- ‘to bend, curve’ < PIE *(h₁)lik- to the root *(h₁)leik-, cf. water name Lith. Liėkė, Liekà (< *[h₁]loikʰēn₂, *[h₁]loikéh₂), whereby the Lech would be ‘the bent one’. Ultimately, this is the position already held by Pokorny. In principle, this could of course be so; however, this derivation has a significant flaw: this Indo-European root does not appear to be attested in the appellative vocabulary, which ultimately means that its meaning cannot really be determined.”

And further, ibid. 46:

“On the other hand, this root *uleikᵘ- offers a starting point for the etymologization of the two Lithuanian rivers mentioned above, Liėkė, Liekà, which, given the fact that the Baltic languages have undergone precisely the sound change typical of satem languages, *-kᵘ- > -k-, and, on the other hand, no inherited word in the

⁴⁴ One should also point out a theoretical possibility that would explain the stress accent, but is semantically unconvincing: in principle, an opaque compound *leiH-h₃kᵘ- could underlie it, composed of PIE *leiH- ‘to pour, pour out’ (LIV² 405f.) and *h₃ekᵘ- ‘to see, behold’ (LIV² 297f.)—but the meaning remains unclear (‘looking like a pour’?). A derivation *loiH-ko from the aforementioned root would likewise contain a laryngeal, a formation parallel to PIE *h₃roiH-ko/h₂- > Common *ŕ̥ēka ‘river’ (from PIE *h₃reiH- ‘to surge, whirl’ (LIV² 305f.), which is continued, for example, in PIE *h₃reiH-no- > Common Celtic *rēno- (→ Lat. Rhēnus), Proto-Germanic *rīna- > German Rhein.

160

Acta Linguistica Lithuanica LXXXIII

Anmerkungen zu den Flussnamen lit. Lieḱe, Lieḱā und ihrer Sippe

Baltic language shows an initial cluster vl-, it can therefore be reconstructed as Indo-European *#ul- > Common Baltic (East Slavic). *#lgilt, as Common Baltic *laik(i̯)ā or *leik(i̯)ā- < Indo-European *u̯loiku̯ieh₂, *u̯loiku̯eh₂ or *uleiku̯ieh₂, *uleiku̯eh₂. The meaning, based on the root meaning given above, would be something like ‘moistening’, that is, ‘the moistening (body of water)’ or similar.”

3.4.2. In view of what was said last, it also becomes clear that the compilation of Slavic and Baltic hydronyms by Udolph (1990: 113–117), accepted, for example, by Wiesinger (2014: 682), proves nothing (there, alongside the Lech he mentions, for example, Slovak. Likava [1321 Lyka], the Croatian river and landscape name Lika⁴⁵, etc.), since the numerous names cited from the Baltic and Slavic spheres may or probably do represent continuants of precisely this Proto-Indo-European root *uleikᵘ- ‘to moisten’; the root proposed by Udolph (1990: 116), “*leikᵘ- ‘to flow’,” following Schmid 1975, is mistaken or outdated. The only Proto-Indo-European root *uleikᵘ- that is still generally continued in and in Iranian means ‘to (back, away), leave, withdraw from’ (EDIV 307f.; LIV² 406f.; this is precisely the root on which the Lithuanian *l̥kti ‘is based’ mentioned above is also based).

One blemish remains: the Proto-Indo-European root *uleikᵘ- ‘to moisten’ does not seem to have been continued appellatively in Slavic.

This Proto-Indo-European root *uleikᵘ should therefore, with probability bordering on certainty, also underlie the only pair of appellatives ever mentioned as an argument in the discussion, Lithuanian liekna f., Latvian liekņa f. ‘marshy meadow’.

^⁴⁵ This name (or the entire name family) harbors a problem that seems to have been overlooked until now: The oldest attestation (apart from Por[tus] Epi[icus] in the Tabula Peutingeriana) is found in Constantine Porphyrogenitus (De Adm. Imp. 30, 93) Λίκα, which, according to Schramm (1981: 286, with reference to older literature), is supposed to be based on the locative Common Slavic *vь̄ Licě. The next-oldest reads 1185 Lica. The question here is why the Croatian name form did not undergo the so-called third (progressive) palatalization, which should have applied here, since it affected Common Slavic *k, *g, *x after Common Slavic *i, *ь, *ę (or, in a more modern interpretation, Proto-Slavic *k, *g, *x after Proto-Slavic *i, *i, *j; also across a nasal [Holzer 2007: 51]). A direct parallel to our case is probably the suffix conglomerate Proto-Slavic *-in-i̯ā > Common Slavic *-bnica. As the more recent description shows, the accentual conditions do not seem to have played a role. Normally, loanwords too should be subject to these sound laws—and the river or region name must have been adopted soon after the Slavic settlement of the region in the late 6th century. This raises the question of why, on the one hand, all Slavic name forms fail to show progressive palatalization—and, on the other, whether Λίτζα might not ultimately be the regular sound-law form of the nominative after all. But this requires a separate investigation.

Straipsniai / Articles

161

HARALD BICHLMEIER

It takes, however, a number of semantic shifts to designate a marshy meadow as ‘the Curved One’.⁴⁶ Why this derivation from the root Proto-Indo-European *(h₁)leik- ‘to bend’, unattested appellatively, now appears as the sole one in Scheungraber, Grünzweig (2014: 210), remains unclear.

4. ON THE ETYMOLOGY OF THE RIVER NAME LECH

4.1. The river name Lech and its etymology now enter the discussion. Three etymologies have been proposed and discussed for it, namely first the ones already mentioned above:

a) a derivation from a Proto-Indo-European root *(h₁)leik- ‘to bend’, in which case either a pre-Celtic or ultimately Celtic formation based on it is assumed;

b) a derivation based on the Proto-Indo-European root *uleiku- ‘to be moist, to moisten’, etc., although the presentations do not always make clear whether the formation is regarded as Celtic, pre-Celtic, or pre-Celtic that has passed through a Celtic filter.

And finally: c) Proto-Celtic *liko/ā- ‘rock, stone’, together with an (expressive?) geminated variant *likko/ā- (required by a few attestations of the river name, some attestations of personal names [if one accepts them as belonging here], and the Celtic appellative comparanda). This designation fits well with the parts of the upper course where the river initially works its way through narrow rocks and later occupies a broad riverbed in which, at low water, extensive areas covered with pebbles become visible.

4.2. Despite the above-mentioned observation, made ten years ago, that precisely a derivation of the Lech name from the semantically naturally appealing Proto-Indo-European root *uleiku is entirely out of the question for phonological reasons, Wiesinger, Greule (2019: 107) present this etymology as the sole one. This represents a clear step backward from what had already been achieved:

⁴⁶ Walde, Pokorny 1927–1932, 2: 397 (s.v. *le̯iqᵘ) had already stated the following in connection with “Lith. lė̃knas ‘morass’, Latv. lėkns, lėkne ‘id., low-lying field’: ‘no more probably from *le̯iqᵘ- ‘to bend’ than ‘indentation, hence damp depression.’”

162

Acta Linguistica Lithuanica LXXXIII

Notes on the Lithuanian river names Liėkē, Liekā and their name family

Only this etymological solution is offered, and in a greatly abridged form:

“Etymologically, Rom. *Licus/*Lica is traced through Celt. *Likos/ā < *wlikos back to PIE *wlikwō-s ‘moistener’, and ultimately to PIE *wleikw- ‘to moisten’. L: Greule (2014), 304; Wiesinger (2014), 681–683; Greule (2015), 343.”

4.3. Let us begin the analysis from the end: The two authors cite only their own works, with Greule (2014 = DGNB: 304) standing out here, since, unlike Wiesinger, Greule (2019), it does still take the history of research into account: all three etymologies mentioned above are presented and briefly discussed, with the ‘stone’ variant being the only one rejected, while the other two initially appear equivalent, but it is pointed out that “*Likos cannot be securely established as Celtic”—and this “[d]espite the two Celtic appellatives Old Irish fluch ‘moist’ and Welsh gwlith ‘dew’ (< *ulik-)”—yet that this reconstruction, although one would have to assume “dissimilatory loss of the IE labiovelar /kʷ/ > /k/ and loss of initial /u-/,” is “to be preferred as the name of a river because of the meaning ‘moistener (of the land), water dispenser or the like’ that can be inferred from it.”

4.4. Compared with this correct presentation, furnished with the appropriate qualifications, the current one in Wiesinger, Greule (2019) is ultimately erroneous in several respects: The explanation that the name goes back as “Rom. *Licus/*Lica through Celt. *Likos/ā < *wlikos to IE *wlikwō-s ‘moistener’, and ultimately to IE *wleikw- ‘to moisten’,” contains the following (typographical) errors:

• The root, if one remains with the notation more usual in the Anglo-Saxon sphere and applied here, should be written *wleikw, not, as the spelling “*wleikw-” suggests, with six phonemes; the root has only five phonemes. Presumably this is merely a typographical error, but it distorts the meaning.

• “IE *wlikwō-s ‘moistener’” would, apart from the faulty notation, first have to be justified morphologically: agent nouns, as one is apparently posited here, do exhibit suffix stress, but normally have the o-full grade in the root; consequently, one would expect PIE *uloik-u̯-ō- ‘moistener’ as opposed to the noun

⁴⁷ Similarly, Greule 2015: 343 = Greule n.d.: 7: “< Roman. Licus (*Likius, Licca) through Celt. *wlikos < (IE) *wlikwōs ‘moistener’ to IE *wleikw- ‘to moisten’,” although here all signs are correctly set.

Straipsniai / Articles

163

HARALD BICHLMEIER

actionis PIE *ulōiku̯-o- ‘moistening, moisture’, or, where appropriate, a nomen rei actae ‘something moistened’.

• The dissimilation *u̯ – kʷ > *u̯ – k, not justified in the presentation cited above, is in principle possible—but ultimately entirely ad hoc and, so far as can be seen, has no parallels whatsoever in Celtic; in a possible example with the same sound sequence, at any rate, nothing of the sort occurs: PIE *uekʷu̯- ‘to say’ (LIV² 673f.; IEW 1135f.) can in any case underlie continuants of a Proto-Celtic stem *uekʷu̯-o- ‘face’ (probably < ‘language’), which show no trace of dissimilation: perhaps belonging here is Ogam VEQ-REQ, certainly Middle Welsh gwep ‘face, appearance, grimace’, Middle Breton goap ‘mockery’ (unless a loanword from Norman French), Gaulish PN Uepo-litanus, Uepos.^⁴⁸

• And even more ad hoc is the assumption that follows thereafter of a loss of the initial Proto-Indo-European/Proto-Celtic *u̯-: There is no evidence whatsoever for such a development in the Celtic languages: Proto-Indo-European *u̯- becomes Proto-Celtic *u̯- and this in turn becomes Old Irish f-, Welsh gw-, Breton g-, Cornish g-, Gaulish /u-/ <u-/ v->. And the same applies to all initial clusters Proto-Celtic *ul- < Proto-Indo-European *(H)ul- and Proto-Celtic *uliK- < Proto-Indo-European *(H)u̯lK-.⁴⁹

• This Celtic etymology becomes finally improbable because, of the word to which the river name Lech is traced here, Proto-Celtic *uliku̯-o-, there is likewise not even an approximate continuant in any Celtic language in which it survives that would exhibit one of the non-regular sound changes assumed above; one finds only the regular reflexes of *u- and *-kʷ-: Old Irish fliuch, Old Welsh gulip, Middle Welsh gulip, gwlyp, gʷlyb, Old Breton gulip, Middle Breton gloeb, glueb, Modern Breton gwleb, Cornish glibor (Latin gloss humor).

4.5. It can thus be stated in summary that the tracing of the name Lech to “IE *wlikwō-s (read: *wlikwō or *ulikʷu̯-o-; H.B.) ‘moistener’,” offered by Wiesinger, Greule 2019 as the sole etymology, would fit well semantically, but does not withstand historical-phonological scrutiny: it remains possible only if one accepts for (pre-Alpine) Celtic (the Bavarian Celtic [as an analogical formation to Bavarian Slavic] or Celtica bavarica) phonological developments otherwise unattested in Celtic, which, moreover, are nowhere attested in the Celtic languages for this very lexeme.

⁴⁸ Vgl. EDPC 409f.; OPEL 4: 155; 2007: 236; Raybould, Sims-Williams 2009: 14, 25.

⁴⁹ For examples, cf. EDPC 425f.

164

Acta Linguistica Lithuanica LXXXIII

Notes on the Lithuanian river names Liėke, Liėką and their name family

4.6. The work by Wiesinger (2014: 681–683) cited in this connection by Wiesinger, Greule (2019) is likewise problematic with regard to the problem just discussed: None of the difficulties accompanying this derivation is seriously addressed. Thus, it says there on p. 682:

“Since the Likates who settled on the Lech around the beginning of the Common Era bear a Celtic name, a Celtic designation of the river is also plausible, for which Celtic *Likā is to be posited.... However, the river name is probably not of Celtic origin, but already of IE pre-linguistic origin. Thus Jürgen Udolph demonstrates related formations, on the one hand, in eastern Croatia via Slovakia and Poland to Belarus and Lithuania and, on the other hand, in the west in Belgium, France, and England. One must then start from IE *lei- (*leiH-) ‘to pour, flow’ (Pokorny, 664) and an extension of the form to *leikᵘ⁻⁵⁰ (*uleikᵘ⁻) ‘to flow’ in Latin liquēō ‘to be liquid’ and liquidus ‘liquid, flowing.’”

In view of the derivation—which clearly differs from the etymology advocated in Wiesinger, Greule (2019), since here a pre-Celtic, in Wiesinger’s terminology ‘Indo-European pre-individual-language’ formation is assumed, whereas there an explicitly Celtic one is assumed—a few clarifying words would have been appropriate in the more recent text, if this text appears as one of only three bibliographic references.⁵¹

If, on this premise, one wanted to salvage the derivation of the name Lech from a Celtic hydronym *Likā, one would have to assume that this is merely a Celticized form, or the form of a pre-Celtic name integrated into Celtic. One would then have to assume further—if one continued to posit an underlying Proto-Indo-European root *uleikᵘ⁻ ‘to flow’ or

⁵⁰ The source for this outdated approach is probably the Udolph 1990: 116 cited in the same place (“*leikᵘ⁻, ‘to flow’ as a root extension to *lei- ‘to pour, flow’”), where he adopts this theory of root extension from Schmid 1975. From today’s perspective, however, the theory of such a root extension must be rejected (the root is to be posited as PIE *leiH- ‘to pour’, LIV² 405f.); this idea merely reflects the completely outdated state of knowledge customary with Udolph. That PIE *leiH- and *uleikᵘ⁻ have any common origin is fundamentally excluded. A PIE root *leikᵘ⁻ ‘low, depression’ that could somehow be indirectly connected with ‘to flow, liquid’ (as, for example, Merlingen 1978: 60) is likewise a phantasm or derives from outdated views: this root is generated by grouping the name Lech with the family around Latvian liekns; the name Lech would then—although this is not developed there—ultimately have meant ‘the low one, the shallow one.’ In light of the discussion above in 3.3, especially in 3.3.1, this etymology appears to be of little use.

⁵¹ Ultimately, this is a concealing term that spares the user from having to decide whether to regard pre-individual-language (in this case pre-Celtic) formations as Proto-Indo-European or ‘Old European’ (that is, somehow western Proto-Indo-European) in Krahe’s sense.

Straipsniai / Articles

165

HARALD BICHLMEIER

recte ‘to moisten’, one would have to assume that the pre-Celtic Indo-European language from which Celtic adopted the name was such that it had lost the labiovelar or changed it into a velar (as, for example, the satem languages did) and had simplified initial *ulto *l(as Baltic, Slavic, and Italic did). Here one enters an area in which scarcely anything can any longer be secured by linguistic argumentation.⁵²

Moreover, this raises the question: if the river name is pre-Celtic (“IE pre-linguistic”), can the ethnic name Licates / Likates still be called Celtic, as it is in the quotation? Or is it Celtic on a pre-Celtic basis, because it contains the (Celtic?) suffix (*)at?

Apart from that, the connection apparently made in the quotation between PIE *leiH- ‘to pour, flow’ (LIV² 405f., Pokorny 1959: 664f.) and PIE *uleiku̯- ‘to moisten’ (LIV² 696f.; not in Pokorny 1959) is downright adventurous—and is not even found in this form in the Udolph (1990: 115f.) cited there.

4.7. Let us return once more to the other two etymologies:

4.7.1. First, the alleged root *h₁leik- ‘to bend’:⁵³

The most important literature containing this etymology is probably IEW 669; Snyder 1967: 161 (“The basis is probably IE *leik- ‘to bend’. Celtic origin of these names... is... less probable.”);⁵⁴ without reference to IEW; Anreiter (2008: 150).

That the etymology starting from this root must be rejected was already demonstrated above in § 3.4.1.

4.7.2. The ‘stone’ etymology is represented primarily in the French-language literature, which, however, was neither received nor cited by Wiesinger, Greule (2019) or the earlier works cited there. In the French-language literature it is moreover generally presented as the sole one for the Lech

52 Moreover, this view nota bene implies that, before the speakers of the language from which Celtic later emerged, speakers of another Indo-European idiom must have immigrated into the Alpine foothills—in other words, one would consequently have to reckon with at least two waves of Indo-European immigration: one perhaps in the 2nd millennium BCE (whose speakers would have coined the hydronym—that would then, in classical terminology, probably be ‘Old European’) and the one with the speakers who later developed into the Celts in the 1st millennium BCE.

53 Both etymologies are mentioned in Sims-Williams (2006: 185, n. 78), without a decision being made.

54 Snyder (1965: 187f.) on the other hand still offers both solutions, the one leading to Proto-Celtic *lik(k)- ‘stone’ and the one leading to PIE *leik- ‘to bend’ (after IEW 669), but already seems to favor the latter. He is categorical, however, in rejecting (for phonological reasons) an etymologization based on the root *uleiku̯-.

166

Acta Linguistica Lithuanica LXXXIII

Anmerkungen zu den Flussnamen lit. Lieḱe, Lieḱā und ihrer Sippe

mentioned, with it usually being pointed out that the river name had simple Proto-Celtic *k, whereas the compared lexemes had geminated Proto-Celtic *kk (e.g. DLG^² 200; DLG^³ 201; Delamarre [2007: 177]; Delamarre [2012: 117]).

DTNG 2, s.v. “licco-, licca ‘flat stone, slab, cliff, rock’ / ‘slab, flat stone, cliff’” now apparently implicitly assumes that the geminate was fundamentally present but usually not written in the attestations, with the exception of Licca in Venantius Fortunatus. While this is no longer mentioned here, the geminate is regarded in DLG^² 200; DLG^³ 201 (and in older literature) as “expressive” (as still in IEW 832, Pokorny 1960: 170, but also Bichlmeier 2010: 43).^⁵⁵

4.7.3. This assessment is astonishing, since the solution for this geminate had already been recognized by Stokes (1893 passim, here: 170) and, following him, confirmed by Lühr (1985 passim, here: 293, 337): In Proto-Celtic there was a sound law comparable to Kluge’s law in Germanic, which caused groups of PIE *p/t/k + *n to yield Proto-Celtic *pp/tt/kk, while those arising from PIE *bʰ/dʰ/gʰ + *n yielded Proto-Celtic *bb/dd/gg.

Stokes (1893) already gives the probably correct etymology: The Celtic ‘stone’ word in the appellative vocabulary goes back to PIE *pₗ̥knā-, in modern notation PIE *p̥l̥k-néh₂- > Early Proto-Celtic *φlikkā- > Proto-Celtic *likkā- > Old Irish lecc ‘slab’, Welsh llech ‘slab’, Breton lec’h ‘gravestone’. The assumption of the zero grade in the reconstruction is necessary only if the name Lech, with its attestations, is to be connected with the form /lik(k)-/.

4.7.5. Otherwise, the following applies: These Celtic words could also be traced to Celtic *(p)lekkā, an originally full-grade form. A further problem with this approach is that, judging from the meanings of the other words derived from this root, this must certainly be a semantic special development in Celtic, in which, starting from the actual meaning ‘flat’, ‘surface’, the restriction to ‘stone surface’ took place (cf. Bichlmeier 2010: 43). In other words, even pre-Proto-Celtic *plek-nḗk could lead to the attested Celtic appellatives.

4.7.6. What has been said thus yields the following: If one starts from PIE/pre-Proto-Celtic *p̥lk-néh₂- > Early Proto-Celtic *φlikkā- > Proto-Celtic *likkā, one can

^⁵⁵ For personal names containing the element Licc, see also OPEL 3: 26. – For further possibilities for the emergence of geminates in Continental Celtic, cf. de Bernardo Stempel 2010. – Isaac 2004 views this solution skeptically (> Possibly Celtic Elements, s.v. lico-); but since there he considers only the etymology, which must be excluded, according to Anreiter (et al.), from the root *(h₁)leik- ‘to bend’ as an alternative, this entry provides no new arguments that would genuinely call the Celtic character of the lexeme into question.

Straipsniai / Articles

167

HARALD BICHLMEIER

the appellative forms and the attestations of the hydronym can be explained together, but one must assume that the geminate was often not written in the hydronym. Alternatively, one can posit PIE/pre-Proto-Celtic *pl̥k-néh₂- > Early Proto-Celtic *φlikkā- > Proto-Celtic *likkā for the appellatives (and one hydronym attestation?), but PIE/pre-Proto-Celtic *pl̥k-éh₂- > Early Proto-Celtic *φlikā- > Proto-Celtic *likā for the hydronym attestations.

4.7.7. Since, as already explained above, the basic meaning of the root PIE *plek was probably ‘flat’ (cf. perhaps Latvian plaks ‘flat’, if from pre-Proto-Baltic *pl̥ok-o- [if not from *pl̥ak-o- ← *pl̥h₂k-o-; see below 4.7.9], Lithuanian plãkanas ‘flat’, as well as several other connections in Pokorny 1959: 831f.⁵⁶),⁵⁷ a no-derivation PIE/pre-Proto-Celtic *pl̥k-nó should likewise most probably be understood as having this meaning. And, as already presented by Bichlmeier (2010: 44), there seems to be no alternative but to assume the following semantic developments: for the appellatives, in any case probably ‘flat’, substantivized then ‘surface’ > ‘stone surface’ > ‘slab’ > ‘gravestone’, for the hydronym ‘the flat one, the shallow one’, whereby, secondarily—on the assumption that the semantic development attested outside Gaulish also took place in the pre-Alpine Gaulish—the meaning ‘slab’ may have been read back into it. A meaning ‘stony’ or the like may have arisen, but cannot for the time being be derived from what has been said and, in any case, would not be morphologically reflected in a derivational process assignable to one of these semantic shifts.

4.7.8. Yet a scenario that initially appears complicated should be presented, one that could explain precisely this meaning ‘stony’ for this hydronym:

If one takes as the starting point an ablauting proterodynamic n-stem PIE/pre-Proto-Celtic *plék-n̥ n. (or *plék-ōn m.), gen. *pl̥k-én-s → *pl̥k-n-és ‘flatness, surface’, or an amphikinetic n-stem PIE/pre-Proto-Celtic *plék-n̥ n. (or *plék-ōn m.), gen. *pl̥k-n-és, an adjective PIE/pre-Proto-Celtic *pl̥k-n-ó derived from it with the possessive PIE *-ó would mean precisely ‘having a surface/flatness’, that is, ‘flat’. If the semantic development

⁵⁶ Cf., with a different etymology (to the PIE root *pleh₂k), LitIEW 602f.; LEV 2: 58; SEJL 467f.; ALEW 2: 785f.; SEJL² s.v. plãkanas; LED 978f.

⁵⁷ Whether the root was actually the result of a process of root extension, as implied in the same place, is irrelevant to the further discussion in our case. If this supposition is correct, one should probably start from a primordial root *pel, which would then have been extended, among other ways, on the one hand to *pl-eh₂- ‘broad, to spread out’, and on the other to *pl-ek- ‘flat’. Here, however, one enters the realm of glottogonic speculation.

168

Acta Linguistica Lithuanica LXXXIII

Anmerkungen zu den Flussnamen lit. Lieḱe, Lieḱā und ihrer Sippe

If the development from ‘surface’ to ‘stone (surface)’ had already taken place in Proto-Celtic, a meaning ‘stony’ could thus also be motivated.

The same applies if one starts from a men-stem Proto-Indo-European/Proto-Celtic *pl̥k-mn̥ n., Gen. *pl̥k-mén-s, Instr. *pl̥k-mn̥-éh₁ ‘flatness, surface’; here, in addition, the simplification of the three-consonant cluster in the instrumental⁵⁸ *-kmn- > *-kn- (still in Proto-Celtic), which can also be observed elsewhere, would have had to take place and become generalized throughout the word’s entire paradigm, so that the possessive adjective derived from it would likewise result as Proto-Celtic *pl̥k-mn-ó- > *pl̥k-n-ó.

4.7.9.Finally, however, attention should be drawn to one more problem connected with the positing of the Proto-Indo-European root *plek: This may indeed underlie some of the formations compiled in IEW 831f.; yet for most of the forms cited there, a Proto-Indo-European root *pleh₂k- ‘to strike’ (LIV² 485) is more likely to be assumed. On its basis, the aforementioned Baltic cognate group around Latvian plaks, Lithuanian plākanas ‘flat’ (< **[flattened] by striking’) can likewise be explained by positing, with renewed vocalism, plak-< *plək- (instead of Proto-Indo-European *pl̥h₂k- > Proto-Baltic *pīlk-/*pūlk).⁵⁹

If the root is indeed (and exclusively) to be posited as Proto-Indo-European *pleh₂k, the question arises whether, on this basis, the derivation of the Proto-Celtic/Gaulish form *libkā is still possible. According to the derivational processes presented above, a preform Proto-Celtic *pl̥h₂k-(m)n-ó- would have to be posited.

Yet this form, too, would ultimately lead precisely to the result presented above, namely Proto-Celtic *libk-o, since this form displays a structure that provides the input for the operation of the ‘Wetter rule’:

According to the ‘Wetter rule’, a laryngeal was lost before a combination of stop and nonsyllabic sonorant when the final syllable was stressed. Cf. Neri (2017: 343), who specifies the ‘Wetter rule’ as follows:

“Laryngeal loss occurred regularly (i.e. as an exceptionless sound law) only in the position after an unstressed vowel or unstressed syllabic resonant and before any sequence of at least one obstruent + nonsyllabic resonant or semivowel (VHK R/UṾ > VK R/UṾ, RHK R/UṾ > RK R/UṾ or R K R/UṾ). The sound change took place at a time when the zero grade and the coloring of a short e had already occurred, but still before the general language-specific laryngeal loss, which was accompanied by compensatory lengthening of the preceding sonorant segment, and is therefore to be classified as Proto-Indo-European.”

⁵⁸ One of the most striking examples of this is provided by Sanskrit áśman- ‘stone’: Nom. áśmā, Gen. aśnáḥ, Instr. aśnā.

⁵⁹ Cf. the literature in footnote 56.

Straipsniai / Articles

169

HARALD BICHLMEIER

with compensatory lengthening of the preceding sonorant segment, and is therefore to be classified as Proto-Indo-European.”

In other words: the development, including that starting from the root form to be assumed as underlying the majority of the attested word forms, is regular and leads to the attested result: Proto-Indo-European *pl̥h₂k-(m)n-ó- > Proto-Celtic *pl̥k-(m)n-ó- > Early Proto-Celtic *pl̥ik-nó- > Proto-Celtic/(Proto-)Gaulish *likko.

4.8. The following conclusions emerge from the review of all the possibilities undertaken here:

4.8.1. That the semantically quite suitable Proto-Indo-European root *uleikū- ‘moisten’ played a role in etymologizing the name of the Lech as a Celtic name is ruled out. If the Lech has a name assigned before the Celtic period, it is theoretically possible that this root underlies it, but this requires several additional assumptions, such as that sound changes must have taken place which otherwise cannot be demonstrated for Central Europe or the Alpine region (but occur only in Italic or in satem languages such as Baltic and Slavic).

4.8.2. That a Proto-Indo-European root *h₁leik- ‘bend’ played a role is likewise ruled out. It simply never existed: the only (Baltic) appellative pair claimed for it must be explained differently (it probably belongs to the Proto-Indo-European root *uleikū- ‘moisten’). The other presumed evidence for the root’s existence, consisting purely of names, is worthless, since purely onymic material can never be used to infer any meanings of anything.

4.8.3. On the basis of the Proto-Indo-European root *plek- ‘flat’ or *pleh₂k- ‘strike (flat)’, both the forms of the hydronym and the Celtic appellatives can be derived without difficulty. The Proto-Celtic geminate continued in the attested word and name forms arose regularly from Proto-Indo-European *-kn-́ > Proto-Celtic *-kk-. It cannot for the time being be determined whether the hydronym attestations without a geminate are based on a different stem formation (Proto-Indo-European *pl̥(h₂)k-ó-/éh₂-) or merely on an imprecise rendering of the Celtic word. A meaningful interpretation for a river name results if an adjectival no-stem Proto-Celtic *pl̥(h₂)k-nó- ‘flat’ (> ‘shallow’) is assumed; the meaning ‘the stony one’, already assumed fairly often in the past, can be derived if Proto-Celtic *pl̥(h₂)k-(m)n-ó- is understood as a possessive ó-derivation from a (me)n-stem Proto-Indo-European/Proto-Celtic *plé(h₂)k-(m)n n., *pl̥(h₂)k-(m)͂n- ‘flatness, surface’ > ‘stone surface’ > ‘stone (surface)’. Depending on how far the semantic development of the (me)n-stem had progressed in (Late) Proto-Celtic / Pre-Alpine Gaulish by the time of adjective formation

170

Acta Linguistica Lithuanica LXXXIII

Anmerkungen zu den Flussnamen lit. Lieḱe, Lieḱā und ihrer Sippe

by then, the derivation meant either ‘having flatness’, ‘having stone surfaces / stone slabs’, or ‘having stones’ = ‘stony’. ⁶⁰

4.8.4. It will probably never be possible to determine what the name’s exact meaning was.

5. RESULT

The investigation presented here thus yields the following: the Proto-Indo-European root *(h₁)leik cannot be attested appellatively and consequently probably never existed. If it did exist, it did so only in onymic word forms, and its meaning therefore cannot be determined.

Thus, for the cognate group of hydronyms around Lithuanian Lieḱe, Lieḱā, the most likely assumption is that the Proto-Indo-European root *uleikᵘ- ‘moisten’ forms the basis. This may also underlie Slavic hydronyms such as Croatian Lika, Slovak Likava, etc.—both if these are genuinely Slavic formations and if they are merely Slavicizing transformations of pre-Slavic formations.

For the name of the Lech, by contrast, a derivation from the Proto-Indo-European root *uleikᵘ- ‘moisten’ is impossible; such an explanation requires special assumptions that have no parallels anywhere. Rather, the name arose from a purely Celtic formation, most likely based on a preform Proto-Indo-European *pl̥(h₂)k-(m)n-ó- > Proto-Celtic *(φ)likko- ‘stony’; alongside this, there may have been a parallel formation Proto-Indo-European *pl̥k-ó- > Proto-Celtic *(φ)liko- (or Proto-Indo-European *pl̥(h₂)k-ó- → Proto-Celtic *(φ)liko- [with renewed vocalism]).

It follows that the connection, frequently encountered in older literature, between Lithuanian Lieḱe, Lieḱā and the Celtic hydronym *Lik(k)o/ā is untenable. The two names (or name groups) are based on formations from two different Proto-Indo-European roots—the results in the individual languages merely resemble each other.

This is therefore a classic case of ‘false friends’—or of the pernicious effect of the ‘siren of sound-alikeness’.

⁶⁰ The possibility of deriving this meaning ‘stony’ for the attested name forms at all was judged considerably more skeptically in Bichlmeier (2010: 44); in light of the considerations presented here, that skepticism is no longer justified.

Straipsniai / Articles

171

HARALD BICHLMEIER

REFERENCES

ALEW – Altaltaiisches etymologisches Wörterbuch (ALEW). Vol. 1: A–M. Vol. 2: N–Ž. Edited under the direction of Wolfgang Hock with the collaboration of Elvira-Julia Bukevičiūtė and Christiane Schiller by Rainer Fecht, Anna Helene Feulner, Eugen Hill and Dagmar S. Wodtko. Vol. 3: Indexes and indices. Edited under the direction of Wolfgang Hock with the collaboration of many others by Rainer Fecht. (Studien zur historisch-vergleichenden Sprachwissenschaft 7), Hamburg: Baar, 2015.

Anreiter Peter 2005: The stratification of Tyrolean hydronyms. – Greule Albrecht, Janka Wolfgang, Prinz Michael (eds.): Hydronyms in Bavaria and Austria. 3rd Colloquium of the Working Group for Bavarian-Austrian Onomastic Research. Regensburg, 27/28 February 2004. (Regensburger Studien zur Namenforschung 1), Regensburg: Edition Vulpes, 35–49.

Anreiter Peter 2005a: A wishful thought: deeper interdisciplinary anchoring of onomastics. – Brendler Andrea, Brendler Silvio (eds.): Name research tomorrow. Ideas, perspectives, visions, Hamburg: Baar, 13–21.

Anreiter Peter 2008: [Review of:] Reitzenstein 2006. – Austrian Name Research 36, 145–153.

Anreiter Peter, Haslinger Marialuise, Roider Ulrike 2000: The Names of the Eastern Alpine Region mentioned in Ptolemy. – Parsons David N., Sims-Williams Patrick (Hrsgg.): Ptolemy. Towards a linguistic atlas of the earliest Celtic place-names of Europe. Papers from a workshop, sponsored by the British Academy, in the Department of Welsh, University of Wales, Aberystwyth, 11–12 April 1999, Aberystwyth: CMCS.

Bichlmeier Harald 2010: Bavarian-Austrian place and hydronyms from an Indo-Europeanist perspective. – Blätter für oberdeutsche Namenforschung 46 (2009[2010]), 3–63.

Bichlmeier Harald 2011: Modern Indo-European linguistics vs. traditional onomastics (Part 2). – Sabe, Drau, Zöbern. – Ziegler Arne, Windberger-Heidenkummer Erika (eds.): Methods of name research. Methodology, methods and practice, Berlin: Akademie Verlag, 63–87.

Bichlmeier Harald 2012: Notes on the terminological problem of ‘Old European hydronymy’, together with Indo-Europeanist additions to the name of the Elbe. – Beiträge zur Namenforschung, Neue Folge 47/4, 365–395.

Bichlmeier Harald 2013: Some Indo-Europeanist notes on the presumed derivational basis of the place name Leipzig: the Proto-Germanic river name *Libō or the Proto-Germanic area name *Libia- (with an excursus on the name of the Rhön and an appendix with further considerations on the name of the Elbe). – Namenkundliche Informationen 101/102, 49–75.

172

Acta Linguistica Lithuanica LXXXIII

Notes on the Lithuanian river names Liėkė, Liekȧ and their cognate group

Bichlmeier Harald 2017: Some notes on the oldest layer of the hydronyms of Pannonia 1. Sá́va/Száva/Save. [Some notes on the oldest layer of the hydronyms of Pannonia—Part I: Sá́va/Száva/Save.] – Nė́vtani Értesítő 38 (2016[2017]), 131–142.

Bichlmeier Harald 2019: Morphological and semantic analysis of the hydronyms of western South Aukštaitija (Part 2). – Acta Linguistica Lithuanica 80, 195–263.

Bichlmeier Harald 2020: The Morphology of Lithuanian Toponyms – Differences and Parallels of Different Classes of Names. – Acta Linguistica Lithuanica 82, 118–139.

Bichlmeier Harald 2020a: Addenda and corrigenda to two recent accounts of the etymology of the river name Lech. – Austrian Name Research 48. [in press]

Bichlmeier Harald, Blažek Václav 2014: Labe—to the sources of the hydronym. – Linguistica Brunensia 62/2, 17–28.

Bichlmeier Harald, Blažek Václav 2015: Elbe—on the sources of a hydronym. – Acta Linguistica Lithuanica 71 (2014[2015]), 125–146.

Bilkis Laimutis 2008: Formation of Lithuanian helonyms. Suffixed and suffixal helonyms, Vilnius: Institute of the Lithuanian Language.

Bräuer Herbert 1961: Slavic linguistics. I: Introduction, phonology. (Sammlung Göschen 1191/1191a/1191b), Berlin: Walter de Gruyter & Co.

de Bernardo Stempel Patrizia 2010: Geminates in Continental Celtic. – Stüber Karin, Zehnder Thomas, Bachmann Dieter (eds.): Proceedings of the 5th German-language Celtic Studies Symposium. Zurich, 7–10 September 2009. (Keltische Forschungen, Allgemeine Buchreihe 1), Vienna: Praesens, 65–87.

Delamarre Xavier 2007: Celtic personal names in classical epigraphy. Nomina Celtica antiqua selecta inscriptionum, Paris: Edition Errance.

Delamarre Xavier 2012: Celtic place names of ancient Europe. (− 500 / + 500). Dictionary, Paris: Edition Errance. (Collection les Hesperides).

DLG² – Delamarre Xavier 2003: Dictionary of the Gaulish language. A linguistic approach to Continental Old Celtic. Preface: Pierre-Yves Lambert, ²rd revised and expanded edition, Paris: Edition Errance.

DLG³ – Delamarre Xavier 2018: Dictionary of the Gaulish language. A linguistic approach to Continental Old Celtic. Preface: Pierre-Yves Lambert, ³th revised and expanded edition, Paris: Edition Errance.

DTNG – Delamarre Xavier 2019–2021: Dictionary of Gaulish nominal stems. / A Dictionary of Gaulish Nominal Stems. Volume I: Ab/Ixs(o)-. Volume II: Lab/Uxs-pero-. (Collection Linguistique 5, 6), Paris: Éditions Les Cent Chemins.

Straipsniai / Articles

173

HARALD BICHLMEIER

EDBIL – Derksen Rick 2014: Etymological Dictionary of the Baltic Inherited Lexicon. (Leiden Indo-European Etymological Dictionary Series 13), Leiden, Boston: Brill.

EDIV – Cheung Johnny 2007. Etymological Dictionary of the Iranian Verb. (Leiden Indo-European Etymological Dictionary Series 2), Leiden, Boston: Brill.

EDPC – Matasović Ranko 2009: An Etymological Dictionary of Proto-Celtic. (Leiden Indo-European Etymological Dictionary Series 9), Leiden, Boston: Brill.

Holzer Georg 2007: Historical grammar of Croatian. Introduction and phonological history of the standard language. (Schriften über Sprachen und Texte 9), Frankfurt, Bern etc.: Peter Lang.

IEW – Pokorny Julius 1959: Indo-European etymological dictionary. Vol. 1, Munich, Bern: Francke.

Isaac Graham R. 2004: Place-Names in Ptolemy’s Geography. CD-ROM, Aberystwyth: CMCS.

LED – Smoczyński Wojciech 2018: Lithuanian Etymological Dictionary, ed. by Axel Holvoet, Steven Young with the assistance of Wayles Browne, Berlin, Bern, Bruxelles, New York, Oxford, Warszawa, Wien: Peter Lang.

LEV – Karulis Konstantīns 1992: Latvian etymological dictionary. 2 vols., Riga: Avots. [^²001: Reprint in 1 vol.]

LitEW – Fraenkel Ernst 1962–1965: Lithuanian etymological dictionary. 2 vols. (Indogermanische Bibliothek: 2nd series, dictionaries), Heidelberg: Carl Winter Universitätsverlag.

LIV^² – Rix Helmut et al. 2001: Lexicon of Indo-European verbs. The roots and their primary stem formations. Second revised and expanded edition, Wiesbaden: Reichert.

Lühr Rosemarie 1980: Cases of double consonance in Celtic. On the question of their genesis. – Sprachwissenschaft 10, 274–346.

Merlingen Weriand 1978: On a semantic branching in the Indo-European lexicon. – Indogermanische Forschungen 83, 40–106.

Neri Sergio 2017: Wetter. Etymology and sound law. (Culture Territori Linguaggi 14), Perugia: University of Perugia.

NIL – Wodtko Dagmar S., Irslinger Britta, Schneider Carolin 2008: Nouns in the Indo-European lexicon, Heidelberg: Carl Winter Universitätsverlag.

OPEL – Onomasticon Provinciarum Europae Latinarum.

Lőrincz Barnabás, Redő Franciscus (eds.) 1994: Vol. I: Aba – Bysanys, Budapest: Archaeolingua. (Archaeolingua 3).

174

Acta Linguistica Lithuanica LXXXIII

Notes on the Lithuanian river names Liėke, Liėka and their cognate group

Lőrincz Barnabás (ed.) 1999: Vol. II: Cabalicius – Ixus, Vienna: Research Society for Viennese Urban Archaeology.

Lőrincz Barnabás (ed.) 2000: Vol. III: Labareus – Pythea, Vienna: Research Society for Viennese Urban Archaeology.

Lőrincz Barnabás (ed.) 2002: Vol. IV: Quadratía – Zures, Vienna: Research Society for Viennese Urban Archaeology.

Pokorny Julius 1960: Celtic etymologies. – Zeitschrift für celtische Philologie 65, 161–171.

Prinz Michael 2007: Regensburg – Straubing – Bogen. Studies on medieval name transmission in the eastern Bavarian Danube region. Part One: uncompounded names. (Materialien zur bayerischen Landesgeschichte 20), Munich: Bavarian Academy of Sciences.

Raybould Marilynne E., Sims-Williams Patrick 2009: Introduction and Supplement to the Corpus of Latin Inscriptions of the Roman Empire containing Celtic Personal names, Aberystwyth: CMCS.

Reitzenstein Wolf-Armin Frhr. von 2006: Lexicon of Bavarian place names. Origin and meaning. Upper Bavaria, Lower Bavaria, Upper Palatinate, Munich: C. H. Beck.

Repanšek Luka 2016: Celtic heritage in the toponymy of the southeastern Alpine region. (Linguistica & philologica 33), Ljubljana: Založba ZRC.

Savukynas Bronys, Vanagas Aleksandras, Vitkaukas Vytautas, Vosylyvtė Klementina, Erm anytė Irena 1963: Gazetteer of the rivers and lakes of the Lithuanian SSR, Vilnius: State Publishing House of Political and Scientific Literature.

Schmid Wolfgang P. 1975: Iranian word studies 1. – Indogermanische Forschungen 80, 80–84.

Scheungraber Corinna, Grünzweig Friedrich E. 2014: The ancient Germanic toponyms and non-Germanic toponyms of Germania. A handbook of their etymology. Edited using a bibliography by Robert Nedoma by Hermann Reichert. (Philologica Germanica 34), Vienna: Fassbaender.

Schramm Gottfried 1981: Conquerors and natives. Geographical loan names as witnesses to the history of southeastern Europe in the first millennium AD, Stuttgart: Anton Hiersemann.

SEJL – Smoczyński Wojciech 2007: Etymological dictionary of the Lithuanian language, Vilnius: Vilnius University.

SEJL² – Smoczyński Wojciech 2018: Etymological dictionary of the Lithuanian language. Second, revised and substantially expanded edition. n.p. Version of 20 January 2018 (www.

Straipsniai / Articles

175

HARALD BICHLMEIER

rromanes.org/pub/alii/Smoczyński W. Słownik etymologiczny języka litewskiego.pdf [accessed in March 2018]).

SEJL^² – Smoczyński Wojciech 2019: Etymological dictionary of the Lithuanian language. Second, revised and substantially expanded edition. Editorial collaboration: Michał Oślon. n.p. Version of 9 March 2019 (access: www.roman es.org/pub/alii/Smoczyński W. Słownik etymologiczny języka litewskiego [pdf; accessed in March 2019]).

SEJL^² – Smoczyński Wojciech 2019a: Etymological dictionary of the Lithuanian language. Second, revised and substantially expanded edition. Editorial collaboration: Michał Oślon. n.p. Version of 24 September 2019 (access: www.roman es.org/pub/alii/Smoczyński W. Słownik etymologiczny języka litewskiego [pdf; accessed in October 2019]).

SEJL^² – Smoczyński Wojciech 2020: Etymological dictionary of the Lithuanian language. Second, revised and substantially expanded edition. Editorial collaboration: Michał Oślon. n.p. Version of 21 June 2020 (access: www.roman es.org/pub/alii/Smoczyński W. Słownik etymologiczny języka litewskiego.pdf [accessed on 28 June 2020]).

Snyder William H. 1967: On the evidence of river names for prehistory and early history. – Beiträge zur Namenforschung, Neue Folge 2, 146–164.

Stokes Whitley 1893: On the assimilation of pretonic n in Celtic suffixes. – Zeitschrift für celtische Philologie 2, 167–173.

Udolph Jürgen 1990: The position of Poland’s hydronyms within Old European hydronymy. (Beiträge zur Namenforschung, Neue Folge, Beiheft 31), Heidelberg: Carl Winter Universitätsverlag.

Vanagas Aleksandras 1970: Formation of Lithuanian hydronyms, Vilnius: Mintis.

Vanagas Aleksandras 1981: Etymological dictionary of Lithuanian hydronyms, Vilnius: Mokslas.

Vanagas Aleksandras 1981a: Semantics of Lithuanian hydronyms. – Studies in Lithuanian onomastics = Issues in Lithuanian linguistics XXI), 4–153.

Vanagas Aleksandras 1988: Lithuanian hydronyms, Vilnius: Mokslas.

Walde Alois 1927–1932: Comparative dictionary of the Indo-European languages. Edited by Julius Pokorny, Vol. 3, Berlin, Leipzig: de Gruyter. [Reprint Berlin: de Gruyter 1973].

Wiesinger Peter 2014: On the gender of hydronyms in Bavarian. – Rampl Gerhard, Zipser Katharina, Kienpointner, Manfred (eds.): In Fontibus Veritas. Festschrift for Peter Anreiter on his 60th birthday, Innsbruck: innsbruck university press, 665–691.

176

Acta Linguistica Lithuanica LXXXIII

Notes on the Lithuanian river names Liėkė, Liėką and their cognate group

Wiesinger Peter, Greule Albrecht 2019: Bavarians and Romans. On the relationship between the early medieval ethnic groups from the perspective of linguistics and onomastics. Cartography: Michael Schefbäck, Tübingen: Narr Francke Attempto.

Some notes on Liėkė, Liėką and other cognate river names

ABSTRACT

The river names Liėkė, Liėką and other hydronyms with the same root have so far appeared in studies of “Old European” hydronymy. They are traditionally etymologically connected with the Proto-Indo-European root *(h₁leik- ‘turn, bend’, although this root probably never existed, since it has not survived in the appellative lexicon of the Indo-European languages. If such a root *(h₁leik did exist, it cannot be associated with any meaning. It might be more reasonable to derive both Lithuanian river names from the Proto-Indo-European root *uleikū- ‘become wet, moisten’. Connections with Baltic lexical material are possible but difficult to substantiate. It must likewise be acknowledged that the hydronym Lech, which has quite often been invoked in explanations based on the roots mentioned, cannot be derived from the Proto-Indo-European root *(h₁leik-, since that root probably never existed, nor from Proto-Indo-European *uleikū-, since there are no phonological grounds for connecting them. It is most likely a Celtic formation whose meaning would be associated with ‘stony, stony ground’: Proto-Indo-European *pl(h₂)r̥-(m)néh₂- > Early Proto-Celtic *φlikrā- > Proto-Celtic *likkā. It thus becomes clear that Lithuanian Liėkė, Liėką, etc., and German Lech are etymologically unrelated.

Submitted on 2 September 2020.

HARALD BICHLMEIER

Saxon Academy of Sciences in Leipzig, Jena Office: Etymological Dictionary of Old High German Zwätzengasse 12a D-07743 Jena [email protected]

Martin Luther University Halle-Wittenberg Institute of Oriental Studies Department of Indo-European Studies and General Linguistics Ludwig-Wucherer-Str. 2 D-06108 Halle (Saale) [email protected]

Straipsniai / Articles

177