Henning Andersen ORCID: 0000-0002-9034-8419 University of California Los Angeles, USA andersen@ucla.edu Charlemagne and All the King’s Men. PS *korlĭ ‘king’? DOI: https://doi.org/10.3986/15.1.01 Slovenski jezik / Slovene Linguistic Studies 15/2023. 3–18. ISSN tiskane izdaje: 1408-2616, ISSN spletne izdaje: 1581-127 https://ojs.zrc-sazu.si/sjsls Slovenski jezik – Slovene Linguistic Studies 15 (2023): 3–18 @language: sl, en, de, it, fr @trans-language: sl, en, de, it, fr @publisher.id: id @doi: 10.3986/00.0.00 @article-type: 0.00 @article-category: category @pages: 3–18 @history-received: dd. mm. yyyy @history-accepted: dd. mm. yyyy *** Žurnal meta *** @issue: xx @volume: 15 @pub-year: 2023 @pub-date: dd. mm. yyyy *** Oprema *** @avtorji: Henning Andersen @running-header: Charlemagne and All the King’s Men. Henning Andersen (ORCID: 0000-0002-9034-8419) University of California Los Angeles, USA andersen@ucla.edu DOI: https://doi.org/10.3986/15.1.01 Charlemagne and a ll the King ’s m en . Ps * korlĭ ‘King ’? The generally accepted idea that the word for ‘king’ in the Slavic languages reflects German Karl, the name of Charlemagne (§1), does not account for all the relevant details of this etymon. This is at least one reason why it may not seem convincing (§2). In fact, the standard etymology is a product of an old-fashioned approach to etymology that is primarily impressionistic and often does not pay attention to phonological and morphological detail (§3). Thus, phonologically PS *korľ-ĭ does not correspond to Gm. Karl (§4.1). Morphologically it is a possessive adjective (§4.2), so it can hardly have originated as a fancy title for a foreign potentate, as is widely believed (§4.3–4). The question is sometimes asked whether this word was really Proto-Slavic. This question, at least, can be answered in the affirmative (§4.5). The main points of the exposition are summed up in the Conclusion (§5). Ključne besede: Common Slavic, Etymology, Language contact, Proto-Slavic, Semantic change Splošno sprejeta ideja, da beseda za ‘kralj’ v slovanskih jezikih odraža nemško Karl, ime Karla Velikega (§1), ne razloži vseh relevantnih podrobnosti tega etimona. To je vsaj eden izmed razlogov, zakaj se morda ne zdi prepričljiva (§2). Ta standardna etimologija je namreč rezultat zastarelega pristopa k etimologiji, ki je primarno impresionističen in pogosto ne posveča pozornosti fonetičnim in morfološkim podrobnostim (§3). Psl. *korľ-ĭ namreč fonetično ne ustreza nem. Karl (§4.1). V morfološkem oziru je svojilni pridevnik (§4.2), zato ga je težko izvajati iz finega naziva za tujega potentata, kot misli večina (§4.3–4). Včasih se zastavlja vprašanje, ali je bila ta beseda res praslovanska. Vsaj na to vprašanje je mogoče odgovoriti pritrdilno (§4.5). Glavni zaključki prispevka so povzeti v razdelku Conclusions (§5). Keywords: splošnoslovansko, etimologija, jezikovni stik, praslovanščina, pomenski prehod 4 Slovenski jezik – Slovene Linguistic Studies 15 (2023) 1 Introduction To the minds of most Slavists the etymology of PS *korľĭ was settled long ago. 1 But several details in its origin are not sufficiently clear to support a full and universally acceptable understanding of the word’s shape or meaning. The title of this communication points to three of them: the posited original shape of the etymon (*korlĭ?), its reconstructed meaning (‘king’?), and the status of this reconstruction (PS?). Many Slavists, perhaps, are content with the more or less laconic ac- counts offered by the handbooks . 2 Karl, nom propre de Charlemagne: sl. [PS] korljĭ ‘roi’, slavon [ChS] kraljĭ, s. [Srb.] krȃlj, slov. [Sn.] králj, tch. [Cz.] král, pol. [P] król, r. [R] koróľ. Cet example montre qu’une diphtongue entrée en slave à une date de peu antérieure aux premiers monu- ments de la langue a subi le traitement ordinaire. L’altération slave des diphtongues en r et l était donc récente à cette date. (Meillet 1965: 66). 3 [...] the Frankish victories in Pannonia impressed the Slavs so much that Charle- magne’s Germanic name Karl was borrowed into Slavic with the meaning ‘king’. Though originally adopted by central Slavic dialects only, this word eventually spread throughout Slavic: Sn. králj, Cz. král, P król, R koról’. (Schenker 1995: 11). 4 But to a critical reader of Slavic texts such very approximate accounts are not enough. A historian concerned with the relationship between 1 In the works cited here, the Slavic etymon is written variously as PS *korľĭ, *korljĭ, and *korl’ĭ. The first of these is my usage; the last is unfortunate, for the apostrophy is common as a marker of palatalization, and the reconstructed PS /ľ/ is not palatalized [l j ], but palatal [ʎ] (see §4.1). The etymon also occurs written PS *korlĭ, as in the title above, which is plain wrong (see §3). Its final vowel may be written equivalently as ĭ or ь. 2 Abbreviations. Cr. (Croatian), Cz. (Czech), CS (Common Slavic), Gm. (German), Gmc. (Germanic), Go. (Gothic), IPA (International Phonetic Association), lit. (literally), PS (Proto-Slavic), ChS (Church Slavonic), OCS (Old Church Slavonic), OR (Old Russian, Old East Slavic), P (Polish), R (Russian), Sn. (Slovene), Srb. (Serbian). 3 Bracketed insertions by the present author. 4 For good measure, let us add this from Wikipedia.org/wiki/Charlemagne: “Charles’ achievements gave a new meaning to his name. In many languages of Europe, the very word for „king“ derives from his name; e.g., Polish: król, Ukrainian: король (korol‘), Czech: král, Slovak: kráľ, Hungarian: király, Lithuanian: karalius, Latvian: karalis, Russian: король, Macedonian: крал, Bulgarian: крал, Serbo-Croatian Cyrillic: краљ/ kralj, Turkish: kral. This development parallels that of the name of the Caesars in the original Roman Empire, which became kaiser and tsar (or czar), among others.” Henning Andersen, Charlemagne and All the King’s Men. PS *korlĭ ‘king’ 5 Slavs and Franks, for example, might have serious questions about this word that are neither answered nor even implicit in the bland statements of the handbooks; see §2. And the reader who turns to an authoritative etymological dictionary such as Trubačev (1984: 82–89) will find a long-winded disquisition that offers a purely speculative origin story for the word and leaves essential questions regarding its form and meaning unanswered; see §3. It might be possible to do better; see §4. 2 what the critical reader needs to know In his recent history of Medieval Central Europe, Curta (2019: 2–4) shows that a critical reader who consults the Slavic handbooks for this crucial piece of linguistic evidence may come away with a story that is more like a fairytale than a scholarly account – as Curta puts it – [a] much repeated, but never demonstrated theory [which] purports that the common noun for ‘king’ [...] derives from Charlemagne’s name. (p. 2) It supposedly illustrates the metathesis of the liquids (the *tort formula with suffixal ictus), which could then be conveniently dated to the Carolingian age on the basis of Charlemagne’s first contacts with the Slavs. [...] In reality there is no connection between Charlemagne and the word for ‘king ’ used in most Slavic languages. (p. 3) The earliest reference to a native *korljĭ is in the Glagolitic inscription known as the Baška tablet, which is dated to 1100 or shortly after that. The word korolĭ [sic] appears four times in the Life of Methodius, and it is usually translated as ‘king’. Aleksander Brückner first noticed that the term was not a [common] noun, but a [...] name – Karl or Carolus used in reference to Frankish kings. 5 We cannot expect a non-linguist to imagine that historical linguistics, like other historical sciences, is essentially hermeneutic: Its practitioners are not engaged in “demonstrating theories”; they interpret available histor- ical data with the aim of creating as coherent as possible accounts of past events. In this business we don’t sweep away existing interpretations with unsupported claims that “in reality” things were different; we proceed, 5 Insertions by the present author. Brückner’s (1913) interpretation was first proposed by Bielowski (1864: vol. 1, 105, 116); see §4.3. 6 Slovenski jezik – Slovene Linguistic Studies 15 (2023) when we can, by amending or replacing existing interpretations that fail to account for available data, or which are not coherent. But the historian who wishes to use (or reject) linguistic evidence needs to know that language is primarily a spoken medium that can change through time independently of any textual attestation. So the fact that a given vocable is first attested in the 1100s does not say much about when its phonological shape or its meaning were codified. Regarding such a specific point as the Slavic “metathesis of the liquids”, it is useful to know that Old Church Slavonic provides a secure terminus ante quem for the regular South Slavic *tort metathesis. Its earliest texts reflect the usage of the missionaries Constantine and Methodius, native Slavic speakers born in the early 800s, whose writings document the *tort metathesis with dozens of examples. Since the *tort metathesis was a regular sound change, the posited *korľĭ would have changed – in any dialect in which it occurred – at the same time and with the same result as other lexical items containing such a liquid diphthong between consonants (cf. §1). With respect to the original meaning of *korľĭ, Schenker thought the Slavs were impressed with Charlemagne (cf. §1), whereas Curta does not think Charlemagne was such a popular figure that his name would enter funda- mental political vocabulary. This sort of speculation is not really useful in linguistic reconstruction. Indeed, as far as the “fundamental political vocabulary” is concerned, a basic rule in semantic reconstruction is not to assume that the modern meaning should be projected back to the time of origin: The first step in semantic reconstruction is to determine the reference potential of a reconstructed form. Once that is done one can attempt to form hypotheses about the referents that motivated its creation and about possible development(s) to its attested meaning(s). 3 what can we learn from the etymological dictionary It should be stated right away that the accounts of this word in Slavic etymological dictionaries typically are not based on a phonological or a morphological analysis. Henning Andersen, Charlemagne and All the King’s Men. PS *korlĭ ‘king’ 7 The etymological dictionaries show variations on the following themes: They state (i) that PS *korľĭ is the Slavic version of the Frankish Karl, or (ii) that it was borrowed or adopted from Frankish Karl, or (iii) that it was based on or goes back to Frankish Karl. 6 These interpretations of the Slavic etymon are based on its obvious similarity with the hypothetical source; they are impressionistic rather than analytic. The most extensive account in the twentieth century is Trubačev’s article on *korl’ĭ in ÈSSJa (1984: 82–89). After a survey of the existing liter- ature on the etymon (pp. 82–84) Trubačev explains how *korl’ĭ ‘king’ was created among the Polabian Slavs in the interaction between Slavs and Franks, especially after 780: Sozdalas’ blagoprijatnaja situacija dlja zaimstvovanija slavjanami bolee prestižnogo inojazyčnogo titula gosudarja. V ètoj funkcii bylo upotrebleno imja znamenitogo Karla Velikogo, [... kotoroe] bylo, konečno, u vsex polab[skix] slavjan na ustax ne odno desjatiletie [...] (p. 85). 7 Further: The meaning of the title may for a long time have been vague, and it may not initially have been used about Slavic leaders; e.g., the korolĭ mentioned in the Life of Methodius may have been the margrave of Bavaria (p. 85). 8 6 See, for instance, Brückner (1927), Vasmer (1953–1957), Sławski (1953–), Šanskij et al. (1975), Holub (1978), Bezlaj (1982), Mel’nyčuk (1989), Černyx (1993), Gluhak (1993), Machek (1997), Snoj (2015). Holzer (2011, s.v.) is an exception; see footnote 12. 7 “A situation arose that was favorable to the Slavs’ borrowing a more prestigeous, foreign title for the ruler. For this they used the name of the famous Charles the Great, ... [whose] name obviously was1 current among the Polabian Slavs for decades.” (p. 85) 8 Trubačev’s idea that the referent of korolĭ in the Life of Methodius was the margrave of Bavaria ignores the repeated conclusions (Bielowski 1864; Brückner 1913; Lunt 1966) that this is not the word for ‘ruler, king ’, but the name Carolus. Under the assumption that they represent PS *korľĭ, the text’s four korol- forms make no coherent sense; besides, their pleophony stands out as odd in a basically South Slavic text; and there is no other evidence that descendants of PS *korľĭ had the meaning ‘king’ and could replace CS kŭnędźĭ already in the 800s, when this text was written; it exists in Russian copies from around 1200. As argued since Bielowski (1864) it seems more reasonable to read these forms as scribal emendations of declensional forms of *karol-ŭ, presumably Carloman (Gm. Karlmann; who led an expedition against Moravia 858, was commander of the southeast frontier marches from 864, king of Bavaria from 876, and died 880 https:// en.wikipedia.org/wiki/Carloman_of_Bavaria). On this assumption, scribal emendations of *karolŭ, *karola, *karolu would have changed one stem vowel and the stem-final /l/ 8 Slovenski jezik – Slovene Linguistic Studies 15 (2023) Under the lemma *korl’ĭ, Trubačev consistently writes the reconstructed form as *korlĭ except where he quotes Lunt (1966) (p. 87). Trubačev’s preferred *korlĭ implies that the word originally was an i-stem, but no-one has ever argued for this, nor is there any evidence for it. All attestations, old and modern, in Slavic languages that have delension, consistently imply the word’s original jo-declension and, in languages that distinguish /l/ and /ľ/, stem-final PS */ľ/. Trubačev makes nothing of Lunt’s proposed reconstruction (*korl-j-ĭ) although it serves as his own article’s lemma. In the same connection he mentions the final accent of his PS *korlĭ, but the significance of this detail is not made clear (pp. 87–88). He also ignores Lunt’s interpretations of the earliest apparent and real attestations of *korľĭ. The article ends with considerations of the phonetic development of the metathesis. Trubačev acknowledges Mareš (1952) but follows Vondrák (1928) and older scholars in positing original dissyllabic forms, *kărăl’-ĭ or *korol’-ĭ, derived from the Frankish spelling variant Karal (p. 88–89), thus contradicting his own lemma. 4 data and interpretation As stated in §1, the question mark in the title of this paper implies three problems: (i) The original shape of the reconstructed PS *korľ-ĭ; (ii) the original meaning of this neologism; (iii) the question, raised by Lunt (1966) and Pronk-Tiethoff (2013, s.v.) – and perhaps others – whether this word can be reconstructed for Proto-Slavic. I will begin with the first of these, which falls into two parts, §§4.1–2. to produce the more familiar-looking East Slavic forms korolĭ, korolja, korolju; in addi- tion, a possessive *karoli (in the phrase *karol-j-i; ep[i]s[ko]pi.nom.pl) was changed to koroljevi (ep[i]s[ko]pi.). Incidentally, one more Karl (Charles the Fat, emperor 881–888) is referred to by name as Kralĭ at the end of the Proložnoe žitie Konstantina i Mefodija; its spelling with final ...ĭ is not informative since the text regularly writes final jer as ь (e.g., gen.pl книгь, preposition вь, dat.pl Бльгаромь, nom.sg Борысь). See textual analysis and references in Lunt (1966: 485–488); the four text loci are in Bielowski (1864: 1.105, 106, 111), also in Lavrov (1930: 74, 75, 77, 101), though without Bielowski’s construal. Henning Andersen, Charlemagne and All the King’s Men. PS *korlĭ ‘king’ 9 4.1 phonology The phonological correspondences across the Slavic language group (cf. Meillet, §1) leave no doubt that the original shape of this word contained a liquid diphthong. The correspondences are perfectly regular with the sole exception of Upper Sorbian kral (for expected *król; cf. Dybo 1963: 69); it is recognized as an intrusion from Czech (Trubačev 1984: 84). PS *korľ-ĭ is reconstructed with a stem-final /ľ/ (regardless of the inconsistent spellings in Cyrillic texts, including Trubačev 1984; see §3). This is important, for it refutes the long-accepted idea that PS *korľ-ĭ is a rendering, borrowing, or reflection of Gm. Karl. 9 In Slavic loanwords from Germanic, Gmc. /l/ is consistently rendered by PS */l/ not by */ľ/, whether before front vowel (1.a) or back vowel (1.b); see Pronk-Tiethoff (2013). 10 (1) a. PS *lixva ‘usury’, *lĭstĭ ‘cunning’, *lĭvŭ ‘lion’, *velĭblǫdŭ ‘camel’, *lěkŭ ‘medicine’, *xlěbŭ ‘bread’, *xlěvŭ ‘cow shed’, *stĭlędźĭ ‘coin’. b. PS *lugŭ ‘lye’, *lukŭ ‘onion’, *plugŭ ‘plow’, *kotelŭ ‘kettle’, *kŭbĭlŭ ‘tub, bushel’, *osĭlŭ ‘ass ’, *stĭklo ‘glass ’, *lagy ‘cask’, *pila ‘saw’, *xula ‘abuse’. The stem-final /ľ/ of PS *korľĭ points to an earlier cluster of base-final /l/ and suffix-initial /j/; see §4.2. 9 Lehr-Spławiński (1939) projected the scholarly Polish pronunciation of the extinct Polabian language back to the origin of PS *korľĭ and thought its ’soft l’ must render a German /l/; hence *korľĭ must have been coined at the Polabian border with Charle- magne’s realm and spread to other Slavic dialects from there. This hypothesis is based on a common, erroneous identification of palatal /ľ/ IPA [λ] with palatalized /l’/ IPA [l j ] as ’soft’. In Polish, PS */ľ/ merged with positional [l j ] variants of /l/, but there is no evidence that PS /ľ/ had merged with positional variants of /l/ in Polabian around 800; cf. Lunt (1966: 488). 10 In PS *bľudo ‘dish’ the /ľ/ is not from Gmc. /l/ but from Gmc. /j/; cf. Go. biuþs ‘platter’. One anonymous referee points to several toponyms in which Sn. /ľ/ corresponds to Gm. /l/. The referee suggests a likely difference in chronology between the acquisition of the appellatives and propria. The issue cannot be pursued here. 10 Slovenski jezik – Slovene Linguistic Studies 15 (2023) 4.2 morphology Almost all the Proto-Slavic masculine nouns acquired from Germanic are declined as o-stems or u-stems. Exceptions are a few secondary jo-declen- sion stems, results of the Progressive Palatalization (e.g., PS *koldędźĭ ‘spring, well ’, *stĭlędźĭ ‘coin’). The jo-declension noun PS *korľ-ĭ is unique in the corpus of Germanic loanwords; cf. Pronk-Tiethoff (2013, s.v.). If we assume PS *korľ-ĭ has a German model, it can only be a relative (’possessive’) adjective, derived from a base *korl- (Pre-PS *karl-; see footnote 11) by the morphologized mutation of the stem-final consonant /l/ → /ľ/ (earlier a suffix PIE *-jo-); cf. Lunt (1966); Trubačev (1984: 87). Some examples of OCS and Old Russian relative adjectives formed this way are in (2). 11 (2) a. OCS člověk-ŭ ‘man’ – člověčĭ ‘man’s, of man’, osĭlŭ – osĭľĭ ‘of an ass’, kozĭlŭ –kozĭľĭ ‘of a goat’, gospodin-ŭ – gospodiń-ĭ ‘lord’s’, mati, mater-ĭ.acc.sg – mateŕ-ĭ ‘mother’s’; a few names: Paul-ŭ – Pauľ-ĭ, ‘Paul ’s ’, similarly Simon-ŭ – Simoń-ĭ, Avra(a)m-ŭ – Avra(a)mľĭ, Pilatŭ – Pilaštĭ (Diels 1932: 108–112). b. OR gospodj-ĭ – gospož-ĭ ‘lord’s’, gospož-a ‘lord’s spouse, lady’, knjazj-ĭ – knjaž-ĭ ‘prince’s’, knjaž-a ‘prince’s spouse, princess’, Vsevolod-ŭ – Vsevolož-ĭ ‘Vsevolod’s ’, Vsevolož-a ‘Vsevolod’s spouse’; Novgorod dial.: gospodj-in-ŭ – gospož-a ‘lord’s spouse’, knjazj-ĭ – knjaž-ĭ mužĭ ‘deputy; lit. prince’s man’, kyjanjin-ŭ – kyjanjinj-ĭ ‘Kievan’s’, Boislav-ŭ – Boislavlj-ĭ, Davyd-ŭ – Davyž-ĭ (Sreznevskij 1958, s.vv.; Zaliznjak 2004, word-index). In brief, simple phonological and morphological analysis of the recon- structed stem *korľ- determines that it is not a Slavicized version of Gm. Karl, but a Slavic relative or possessive adjective possibly derived from Karl. Since the base PS *korl- is monosyllabic, we might expect the posses- sive suffix allomorph -ĭj-; but in the earliest attested period the ideally complementary distribution of the alternants -j- ~ -ĭj- is clouded (cf. OCS 11 The examples in (2.a) are normalized and transliterated; in (2.b) palatalization (sharping) is indicated (with superscript j ) where appropriate. Henning Andersen, Charlemagne and All the King’s Men. PS *korlĭ ‘king’ 11 ovĭč-ĭ ~ ovĭč-ĭj-ĭ ‘sheep’s, OR kŭnjaž-ĭ – knjaž-ĭ ‘prince’s ’; Vaillant 1974: 429). As for the desinential accent of *korľ-ĭ (cf. R koról’ – koroljá, Srb. krȃlj – králja), the accent of the base form *Korlŭ is not attested, and we are limited to observing that the adjective’s accent is consistent with a derivative from an accentless (enclinomenon) base (traditionally a.p. c; cf. Dybo 2000: 120). 4.3 semantics The morphological analysis of the word leads straight to an understanding of its original literal reference potential: If the relative adjective *korľ-ĭ was derived from a man’s name and meant ‘Karl’s’, it could characterize anything or anyone that was connected with that individual. In a phrase such as PS*korľĭ mǫžĭ – parallel to OR knjažĭ mužĭ in (2.b) (cited from Birchbark text 109, Zaliznjak 1995: 235) – it could refer to a person as one of Karl’s men, a representative, deputy, or vassal. 12 In frequent use, such a phrase would be prone to ellipsis of its semantic head mǫžĭ and reanal- ysis of *korľĭ as a noun. Note the feminines OR gospoža ‘lady’, knjaža ‘princess, lady’, Vsevoloža ‘wife of Vsevolod’ in (2.b), which illustrate substantivized relative adjectives. Substantivized, *korľĭ would retain the semantic value of the former phrase, ‘representative of Karl’. Trubačev thought that the original meaning of *korľ-ĭ was vague. It is better to recognize that its meaning was precise, but broad. Assuming that Karl was the name of a ruler – say, a Frankish king – ‘Karl ’s (man)’ could refer to a Frankish person of secular authority – civil or military – or ecclesiastic, without regard to rank, a detail for which there was no Slavic terminology, and which would likely be unimportant to the common person. So, any of Karl’s men might be known locally by his name but could be referred to by a local variant of the generic PS *korľĭ, first with the meaning ‘Karl ’s man’, later narrowed in reanalysis, functional as ‘the Frankish head man, governor, or regent ’ or hierarchical as ‘the vassal, duke, or local rex, 12 “[…] we have here […] a derivative from the proper name, a possessive formed with the familiar formant -j-. *korljĭ was then ‘Karl’s’, hence ‘Karl’s man, representative, deputy’.” (Lunt 1966: 488). Holzer (2011, s.v.) recognizes the possessive suffix but proposes no semantic interpretation. 12 Slovenski jezik – Slovene Linguistic Studies 15 (2023) king’. The time and context in which the expression’s meaning became reanalysed as ‘ruler, king ’ is unknown; the earliest attestations are Croatian and Serbian from the 1100s (Lunt 1966). Regardless where the word was created, it may well have been propagated across the Slavic regions before the final semantic change to ‘king’ occurred. 4.4 space and time Assuming that PS *korľĭ is a derivative of the name of one of the Frankish kings, it may have been created in any of the contact zones of Slavs and Franks, perhaps in the West Lechitic or Sorbian areas in the northwest, where the Franks fought with the Slavs ’ Saxon neighbors repeatedly in the 700s, or – perhaps more likely – in the southeast of the Frankish area of expansion, between Bavaria in the west and Pannonia in the east, between Bohemia and Moravia in the north and Friuli, Istria, and Dalmatia in the south. There were enough dukes, counts, margraves, and other men representing Frankish authority in these regions, from Charles Martel’s Bavarian campaigns in the 720s to Charlemagne’s repeated advances in this region later in the century (Curta 2019: 63–65, 101–105). Historians who are interested in this matter, and who have specialist knowledge to contribute, should feel invited to point to areas where the creation of this Slavic term would have been likely. But actually such an obvious label for a Frankish person of authority could have been created independently in several places. Nor is it surprising that – once coined – it would become propagated and widely adopted across the Slavic lands. From a Slavic point of view, the Frankish kings, especially Charlemagne, were major actors on the geopolitical stage, and their campaigns, whether in the northwest or the southeast, could hold significant implications for Slavic communities. Regarding the propagation of PS *korľĭ ‘Frankish person of authority’, it is reasonable to assume a difference between areas in which it spread from person to person and was in common use, and more distant areas in which it was of little or no practical utility and might be known mainly to members of governing circles. Henning Andersen, Charlemagne and All the King’s Men. PS *korlĭ ‘king’ 13 While there is no precise answer to the question Where was it coined? the question When? has a fairly determinate answer. First, the extralinguistic context for its coinage may well have been the 700s, during Charles Martel’s rule (714–741) or Charlemagne’s (768–814). Assuming an early substantivization of PS *korľĭ to ’head man, governor’, it might have been political events after Charlemagne’s death that prompted a semantic shift from functional to hierarchical meaning based on usage equivalent to that of western words for ‘head of state, ruler, king’. Secondly, the regular phonological correspondences show that variants of *korľĭ became established across the Slavic-speaking territories either before or during the period when the ‘*tort metathesis’ was in progress. This was a phonetically gradual change. 13 Undoubtedly it was enacted through stages of stylistic variation. And even though it produced different regional outcomes, at any time during its progression Slavic speakers would have had no difficulty identifying the stylistic variation in the speech of their neighbors with their own ideal (underlying) representations. One early attestation of the metathesis is the oft-cited Veleti name Dragovit, recorded in 789 (Schenker 1995: 46). It gives a chronological hint, but since it was recorded by a non-Slav, it is uncertain which stage in the phonetic progression of the sound change it reflects. Besides, it can be assumed that the change proceeded earlier in some regions than in others (cf. footnote 13 ). 4.5 Is * k o r ľ ĭ proto-slavic? Pronk-Tiethoff (2013, s.v.) hesitates to reconstruct *korljĭ as Proto-Slavic, being that it is such a recent neologism. Lunt (1966) had similar qualms. This hesitation implies a terminological problem and a conceptual confu- sion. 13 It is notable that the ‘metathesis’ intersected with the change of Pre-PS /ă/ > /o/ (part of ‘the qualitative differentiation of long and short vowels ’). In South and Central Slavic it was a ‘*tart metathesis’: [karľ- > kar a ľ- > k a rāľ- > krāľ-], Cr. králj, Cz. král; in Northwest Slavic, by contrast, [korľ- > kor o ľ- > k o rōľ- > krōľ-], P król. In East Slavic, the ‘metathesis ’ stalled, producing pleophony: [korľ- > kor o ľ- ~ k o roľ- > koroľ-], R korol’; cf. Timberlake (1986); Bethin (1998: 47–52). Van Ginneken (1935: 59–66) illustrates similar regular changes in Dutch dialects in Brabant and Limburg; cf. Weijnen (1966: 265). 14 Slovenski jezik – Slovene Linguistic Studies 15 (2023) The term Proto-Slavic is used in two senses: (i) PS 1 is a technical term. Think of its reference as an inventory of reconstructed roots and affixes, lexemes, wordforms and fixed expressions, as well as morphosyntactic and sentence-syntactic rules, all established by the comparative method; when these are cited they are labeled PS. This inventory “ne se laisse amener à aucun état historique déterminable dans le temps et dans l ’espace” (Meillet 1965: 11). 14 Although timeless it comprises elements of different age; some items are its Indo-European heritage, some are prehistoric native forma- tions, some are borrowings from identifiable or unidentifiable sources, still others are later neologisms of various kinds. Sorting the inventory into such categories can only be done through secondary comparisons of Proto-Slavic reconstructions with non-Slavic languages. Thus, in our example, the Slavic words for ‘king ’ present regular correspon- dences and are reconstructed as PS *korľ-ĭ ‘Karl ’s ’. Secondary comparison with Germanic helps us identify the source of its base as Gm. Karl. This and the regular correspondences help us hypothesize the time at which it was coined. But this additional information does not remove the reconstructed *korľ-ĭ ‘Karl’s’ from the Proto-Slavic inventory. (ii) In the second sense, Proto-Slavic (PS 2 ) is used loosely (by some) about the Slavic spoken in the period around its first written attestation. This usage appears based on the naive idea that PS 1 is a reconstructed language, a sort of pre-stage of Old Church Slavonic. Scholars who are accustomed to this loose usage may feel that Proto-Slavic (PS 1 ) reconstructions are somehow ‘older’, and that new words in the language (PS 2 ) should perhaps not be assigned reconstructed forms. The only cure for such feelings is a clear distinction between (i) Proto- Slavic, the inventory of reconstructions, to which scholars add new items and revisions as time goes by, and (ii) the spoken language of the medieval Slavs with its dynamic development and gradual geographical differenti- ation; for a detailed discussion of this distinction, see Andersen (1986). It 14 “ […] cannot be assigned to any historically defined point in time or place in space [HA]”. The abstract nature of a proto-language is not due to any difficulty in dating it or locating its ‘homeland’. It is a logical matter that derives from its creation and the use to which it is put. It is created by the comparative method, and its asterisked elements refer to actual elements of the languages from which it has been derived. Henning Andersen, Charlemagne and All the King’s Men. PS *korlĭ ‘king’ 15 would be a good idea to use the term Proto-Slavic just for the inventory of reconstructions (PS 1 ) and choose another name for the living language, a name that could be modified with chronological adjectives such as early or late and geographical ones, such as southern or northwestern. Some writers have called it Common Slavic (CS). One could perhaps call it (Spoken) Medieval Slavic (SMS). Whereas this language (PS 2 ) was spoken in time and space – in medi- eval Slavic communities – Proto-Slavic (PS 1 ) is not a language, but a meta-language, that is, a language used to speak about another language or languages. It is a scholarly creation whose elements are never spoken, except as examples in scholarly presentations and debates. It is, technically speaking, a scientific meta-language (cf. Hjelmslev 1969: 120). 5 conclusion In this paper we have witnessed the disbelief with which a non-linguist may encounter the results of Slavic historical linguistics, and we have seen some of the basic information without which such results cannot be understood, let alone evaluated (§2). But we have also seen an illustration of the traditional impressionistic approach in Slavic etymology, which begins with a hypothetical solution and then does not know what to do with essential details that call for linguistic analysis (§3). The phonological and morphological sections (§§4.1–2) showed how a purely linguistic analysis of a given expression may conclude with its semantic minimum, its literal refential potential. The pragmatic interpretation of the neolo- gism hypothesizes a socio-cultural context for its creation and subsequent semantic development (§4.3). In this case both the regular phonological correspondences and the context posited for its origin contributed to some understanding of the time and place(s) in which it may have been coined (§4.4). The introduction posed three questions that for some time have needed to be clarified. (i) The phonological shape of the etymon is PS *korľ-ĭ, not *korlĭ, as has often been claimed (§§4.1–2). 16 Slovenski jezik – Slovene Linguistic Studies 15 (2023) (ii) Its original meaning was not ‘king’. It was coined as a possessive adjective ‘Karl ’s ’. Usage in phrases such as PS *korľĭ mǫžĭ ‘Karl ’s man (representative, deputy)’ prompted substantivization as ‘Frankish man in authority, governor’, which with time and through changes in usage eventuated in ‘ruler, king’ (§§4.3–4). (iii) PS *korľ-ĭ ‘Karl’s’ is a lexeme in the reconstructed Proto-Slavic lexicon. Its shape is warranted by the regular phonological correspon- dences in the Slavic words for ‘king’. Its posited original meaning ‘Karl’s’ is ‘warranted by its morphological composition; its modern meaning ‘king’ arose through understandable semantic developments subsequent to its creation (§4.5). references Andersen, Henning. 1986. Protoslavic and Common Slavic: questions of periodiza- tion and terminology. Michael S. Flier, Dean S. Worth (eds.), Slavic Linguistics, Poetics, Cultural History. In honor of Henrik Birnbaum on his sixtieth birthday, 13 December 1985, 67–82. (= International Journal of Slavic Linguistics and Poetics 31/32.) Columbus, Ohio: Slavica. Bethin, Christina Y. 1998. Slavic Prosody. Language change and phonological theory. Cambridge: University Press. Bezlaj, France. 1982. Etimološki slovar slovenskega jezika, vol. 2. K–O. Ljubljana: Mladinska knjiga. Bielowski, August. 1864. Monumenta historiae Poloniae/Pomniki dziejowe Polski, vol. 1. Lwów: Polska Akademia Umiętności. Brückner, Aleksander. 1913. Die Wahrheit über die Slavenapostel. Tübingen: Mohr. Brückner, Aleksander. 1927. Słownik etymologiczny języka polskiego. Cracow: Kra- kowska spółka wydawnicza. Curta, Florin. 2019. Eastern Europe in the Middle Ages (500–1300). (Brill Companion to European History, 18.) Leiden: Brill. Černyx, P. Ja. 1993. Istoriko-etimologičeskij slovar’ sovremennogo russkogo jazyka 1–2. Moscow: Russkij jazyk. Derksen, Rick. 2008. Etymological Dictionary of the Slavic Inherited Lexicon. Leiden: Brill. Dybo, V. A. 1963. Ob otraženii drevnix količestvennyx i intonacionnyx otnošenij v verxnelužickom jazyke. L. È Kalnyn’ (ed.), Serbo-lužickij lingvističeskij sbornik. Moscow: Akademija Nauk SSSR. 54–83. Dybo, V. A. 2000. Morfologizovannye paradigmatičeskie akcentnye sistemy. Tipologija i genezis. Moscow: Jazyki russkoj kul’tury. Henning Andersen, Charlemagne and All the King’s Men. PS *korlĭ ‘king’ 17 Trubačev, Oleg. N. 1984. Etimologičeskij slovar’ slavjanskix jazykov. Praslavjanskij leksičeskij fond, vol. 11. Moskow: Nauka. Ginneken, Jacobus van. 1935. Ras en taal. Amsterdam: Noord-Hollandsche uitgevers- maatschappij. Gluhak, Alemko. 1993. Hrvatski etimološki rječnik. Zagreb. Hjelmslev, Louis. 1969. Prolegomena to a Theory of Language. Translated by Francis J. Whitield. Madison: University of Wisconsin Press. Holub, Josef. 2 1978. Stručný etymologický slovník jazyka českého. Praha: Statní peda- gogické nakladatelství. Holzer, Georg. 2011. Glasovni razvoj hrvatskoga jezika. Zagreb. Lavrov, P. A. 1930. Materialy po istorii vozniknovenija drevnejšej slavjanskoj pis’men- nosti. Leningrad: Akademija nauk SSSR. Lehr-Spławiński, Tadeusz. 1927. Pochodzenie i rozpowszechnienie wyrazu król w polszczyźnie i w innych językach słowiańskich. Prace filologiczne 12: 44–53. Lunt, Horace G. 1966. Old Church Slavonic *kraljь? Orbis scriptus. Festschrift für D. Tschiževskij. Munich: Sagner. 484–489. Machek, Václav. 3 1971. Etymologický slovník jazyka českého. Praha: Československá akademie věd. Mareš, František V. [1952] 1969. Diachronische Phonologie des Ur- und Frühslavischen. (Slavistische Beiträge, 40.) Munich: Otto Sagner. Meillet, Antoine. 2 1965. Le slave commun. Paris: Champion. Melnyčuk, O. S., Bilodid, I. K., Kolomijec, V. T., Tkačenko, O. B. 1989. Etymolohičnyj slovnyk ukrajins’koji movy, vol. 3. Kiev: Naukova dumka. Pronk-Tiethoff, Saskia. 2013. The Germanic Loanwords in Proto-Slavic. Amsterdam: Rodopi. Schenker, Alexander M. 1995. The Dawn of Slavic. An introduction to Slavic philology. New Haven: Yale University Press. Sławski, Franciszek. 1953–. Słownik etymologiczny języka polskiego. Cracow: To- warzystwo Miłośników Języka Polskiego. Snoj, Marko. 3 2016. Slovenski etimološki slovar. Ljubljana: Založba ZRC. Sreznevskij, Izmail I. 1958. Materialy dlja slovarja drevnerusskogo jazyka po pis’men- nym pamjatnikam 1–3. Sankt-Peterburg. Šanskij, N. M., V. V. Ivanov, T. V. Šanskaja. 1975. Kratkij etimologičeskij slovar’ russkogo jazyka. Moscow: Prosveščenie. Timberlake, Alan. 1986. The metathesis of liquid diphthongs in Upper Sorbian. M. S. Flier, D. S. Worth (eds.). Slavic Linguistics, Poetics, Cultural History. In Honor of Henrik Birnbaum on his Sixtieth Birthday 13 December 1985. (= International Journal of Slavic Linguistics and Poetics 31–32, 1985.) 417–430. Trubačev, O. N. 1984. *korl’ь. Ètimologiceskij slovar’ slavjanskix jazykov. Praslavjanskij leksičeskij fond vol. 11. Moscow: Nauka. 82–89. Vaillant, André. 1974. Grammaire comparée des langues slaves vol. 4. La formation des noms. Paris: Klincksieck. 18 Slovenski jezik – Slovene Linguistic Studies 15 (2023) Vasmer, Max. 1953–1957. Russisches etymologisches Wörterbuch 1–3. Heidelberg: Winter. Translated: O. N. Trubačev, 1964–1973. Ètimologičeskij slovar’ russkogo jazyka 1–4. Moscow: Progress. Vondrák, Vácslav. 1928. Vergleichende slavische Grammatik, vol. 1. Lautlehre und Stammbildungslehre. Göttingen: Vandenhoeck und Ruprecht. Weijnen, A. 2 1966. Nederlandse dialectkunde. Assen: Gorcum. Zaliznjak, A. A. 2 2004. Drevnenovgorodskij dialekt. Moscow: Jazyki russkoj kul’tury. Prispelo februarja 2022, sprejeto maja 2022. Received February 2022, accepted May 2022. Acknowledgements I am grateful to Ms. Justina Bandol for drawing my attention to Curta (2019) and to two anonymous referees, whose friendly comments helped me make the exposition more precise in several places. summary charlemagne and All the King’s men. ps *korlĭ ‘king’? The article is motivated by the historian F. Curta’s rejection of the traditional etymology of PS *korľĭ and the unsatisfactory treatment of this etymology in the authoritative Ètimologoičeskij slovar’ slavjanskich jazykov ed. by O. N. Trubačov. It shows that phonological and morphological analysis lead to a Proto-Slavic possessive adjective *korl-j-ĭ with an original referential value ‘Karl’s (representative)’, substantivized as ‘Frankish person of authority’, eventually > ‘regent, king’ (as first suggested by Lunt 1966). The Proto-Slavic status of *korľĭ, which some have doubted, is shown to be not an empirical but an important theoretical issue. Karel Veliki in vsi kraljevi možje. psl. *korlĭ ‘kralj’? Prispevek sta spodbudili F. Curtova zavrnitev tradicionalne etimologije psl. *korľĭ in nezadovoljiva obravnava te etimologije v avtoritativnem Ètimologičeskij slovar’ slavjanskich jazykov, ki ga ureja O. N. Trubačov. Prikazuje, da fonetična in morfološka analiza kažeta na praslovanski svojilni pridevnik *korl-j-ĭ s prvotno nanašalno vrednostjo ‘Karlov’, posamostaljen kot ‘Frankovska oseba z avtoriteto’, kasneje > ‘vladar, kralj’ (prvič predlagal Lunt 1966). Praslovanski status rekonstrukcije *korľĭ, v katerega se je dvomilo, se izkaže ne za empirični, temveč pomemben teoretični problem.