Научная статья на тему 'Λαβύρινθος AND WORD-INITIAL LAMBDACISM IN ANATOLIAN GREEK'

Λαβύρινθος AND WORD-INITIAL LAMBDACISM IN ANATOLIAN GREEK Текст научной статьи по специальности «Языкознание и литературоведение»

CC BY-NC-ND
226
26
i Надоели баннеры? Вы всегда можете отключить рекламу.
Ключевые слова
λαβύρινθος / MYCENAEAN GREEK / LAMBDACISM / ANATOLIAN / LYGDAMIS / LINEAR A / МИКЕНСКИЙ ГРЕЧЕСКИЙ / ЛАМБДАЦИЗМ / АНАТОЛИЙСКИЕ ЯЗЫКИ / ЛИГДАМИС / ЛИНЕЙНОЕ A

Аннотация научной статьи по языкознанию и литературоведению, автор научной работы — Valério Miguel

The lexical pair formed by Mycenaean da-pu(2)-ri-toand later Greek λαβύρινθος presents a contrast between Linear B d and alphabetical λ in a position where one would expect to find a similar sound represented. This orthographic inconsistency has been taken as a synchronic fluctuation between /d/ and /l/, both optimal adaptations of what is assumed to be a non-Greek (Minoan) sound in da-pu(2)ri-to-. In turn, it has been proposed that this “special” and wholly theoretical sound, which according to some suggestions was a coronal fricative, was behind the Linear A d series. Here it is argued that there is actually no evidence that /d/ and /l/ alternated synchronically in Mycenaean Greek, and that therefore the /l-/ of λαβύρινθος is more likely the result of a later shift. Starting from this premise, it is hypothesized that λαβύρινθος derives from a form closer to Mycenaean da-pu(2)-ri-to-, an unattested *δαβύρινθος, that underwent a shift /d-/ > /l-/ in Southern or Western Anatolia. The proposed motivation is the influence of some local Anatolian language that prohibited /d/ word-initially. The same development is considered for λάφνη and λίσκος, which Hesychiusglossed as Pergaean (Pamphylian) forms of standard Greek δάφνη ‘sweet bay’ and δίσκος ‘discus, quoit’, and possibly also for the Cimmerian personal name Dugdammê/Λύγδαμις. Of course, this hypothesis has implications for our perception of the Linear A d series and certain open questions that concern the Aegean-Cypriot syllabaries.

i Надоели баннеры? Вы всегда можете отключить рекламу.
iНе можете найти то, что вам нужно? Попробуйте сервис подбора литературы.
i Надоели баннеры? Вы всегда можете отключить рекламу.

Текст научной работы на тему «Λαβύρινθος AND WORD-INITIAL LAMBDACISM IN ANATOLIAN GREEK»

Miguel Valério

University of Barcelona; mfgvalerio@outlook.com

Aapvpiv0o^ and word-initial lambdacism in Anatolian Greek

The lexical pair formed by Mycenaean da-pu(2)-ri-to- and later Greek Aa|3ug>Lv0oc; presents a contrast between Linear B d and alphabetical A in a position where one would expect to find a similar sound represented. This orthographic inconsistency has been taken as a synchronic fluctuation between /d/ and /l/, both optimal adaptations of what is assumed to be a non-Greek (Minoan) sound in da-pu(2)-ri-to-. In turn, it has been proposed that this "special" and wholly theoretical sound, which according to some suggestions was a coronal fricative, was behind the Linear A d series. Here it is argued that there is actually no evidence that /d/ and /l/ alternated synchronically in Mycenaean Greek, and that therefore the /l-/ of Aa|3ug>Lv0oc; is more likely the result of a later shift. Starting from this premise, it is hypothesized that Aa|3ug>LV0og derives from a form closer to Mycenaean da-pu(2)-ri-to-, an unattested *Sa|3ug>LV0og, that underwent a shift /d-/ > /l-/ in Southern or Western Anatolia. The proposed motivation is the influence of some local Anatolian language that prohibited /d/ word-initially. The same development is considered for Ad^>vr| and AioKog, which Hesychius glossed as Pergaean (Pamphylian) forms of standard Greek 5d^>vr| 'sweet bay' and SioKog 'discus, quoit', and possibly also for the Cimmerian personal name Dugdamme/ Auy5a^Lg. Of course, this hypothesis has implications for our perception of the Linear A d series and certain open questions that concern the Aegean-Cypriot syllabaries.

Keywords: Aa|3ug>Lv0og; Mycenaean Greek; lambdacism; Anatolian; Lygdamis; Linear A

1. Linear B da-puAi)-ri-to- , alphabetical Greek AapugivGog and the alleged /d/ ~ /l/ alternation in Mycenaean Greek

The pair formed by Linear B da-pu(i)-ri-to- and later Greek Aapupiv0og constitutes one of the most discussed sets of lexical items among Aegeanists.

The alphabetical form, Aapupiv0oc;, is first attested in the work of Herodotus (2.148), which dates from the 5th century BCE. It refers to a vast, partially underground mortuary complex located in Egypt, near the Lake Moeris. Later, the word appears in inscriptions from the temple of Apollo at Didyma, in Caria (ca. 218 BCE).1 There, it designates two stairwells of the temple that consisted of a double flight of steps (Montegu 1976: 304). If not for their own winding, these architectural features may have been termed AapuQiv0oc; because their ceiling was carved with a meander pattern (Fontenrose 1988: 38, n. 15). Generally speaking, the word came to refer to "a large building consisting of numerous halls connected by intricate and tortuous passages" (Liddell and Scott 1940), and accordingly Hesychius glossed the word as KoxAioaS^g Torcog 'spiral place' (Latte 1956). We can assume that a general sense of 'sinuous architectural feature' was what led Herodotus to use AapuQiv0oc; to describe the Egyptian complex, whereas the stairwells at Didyma received this designation either because of their shape or decoration.

1 See inscriptions no. Didyma 84 and 86 (McCabe 1985).

Journal of Language Relationship • Вопросы языкового родства • 15/1 (2017) • Pp. 51-59 • © The authors, 2017

The Linear B form was unveiled after the decipherment of the script in 1952. In the clay tablet KN Gg(1) 702 it is part of the phrase da-pu2-ri-to-jo po-ti-ni-ja 'to the Lady of D.' which registers the offering of a honey jar to a goddess named thus, alongside an identical gift 'to all the gods' (pa-si-te-o-i) (Ventris and Chadwick 1973: 310). The same female divine name appears also in a more fragmentary text, KN Oa 745, and a broken form da-pu-ri-to[ is attested in tablet KN Xd 140, most likely representing a variant spelling of da-pu2-ri-to-jo or a related form. Already Palmer (1955: 40) proposed to interpret da-pu2-ri-to-jo po-ti-ni-ja as *AapuQÍv0oio noTvia(i) 'Lady of Labyrinthos (dat.)'. In the Linear B tablets, the word po-ti-ni-ja /potnia/ 'mistress' is frequently preceded by an epithet and sometimes written as one word. When this is the case, the preceding word is thought to be usually a place name, or a noun in the genitive: a-ta-na-po-ti-ni-ja (MY Oi 701) /Athanai Potniai/ 'Lady (of) Athana (dat.)'; e-re-wi-jo-po-ti-ni-ja (PY Vn 48.3), of uncertain meaning, but possibly with a place-name in the genitive; si-to-po-ti-ni-ja (MY Oi 701) 'Lady of the Grain(s) (dat.)'; u-po-jo po-ti-ni-ja (PY Fn 187; Fr. 1225, 1236), again possibly with a toponym in the genitive (Ventris and Chadwick 1973: 545, 574; Aura Jorro 1985: 160; Trümpy 2001). Therefore, da-pu2-ri-to-jo is interpreted as the genitive of a place called da-pu2-ri-to-, closely comparable to AapúpivGog.

What the word meant by Mycenaean times (ca. 1450-1200 BCE) and what this place was exactly remains a matter of debate, the most notable suggestions being 'palace sanctuary' (Evans 1921: 6)2 and 'cultic cave, subterranean sanctuary' (Cagiano 1958: 48-52 and Faure 1964 apud Sarullo 2008; Montegu 1976: 304)3. Be it as it may, the problem that I would like to treat here is essentially phonological, not semantic. Since the 1950s, the indisputable equation of da-pu2-ri-to- and AapúpivGog has generated much discussion. The debate stems from the unexpected Linear B spelling of the initial sound of the word with d, which consistently represents /d/ in native Mycenaean Greek words, instead of the expected r, which transcribed both /r/ and the liquid /l/. The hypotheses that have been advanced to account for this inconsistency have ramifications for several open questions in Aegean and Anatolian studies, and this is the reason why they merit a reassessment.

2 Elsewhere (Valerio 2007) I have argued for a connection between Linear B da-pu2-ri-to- and the Linear A sequence du-pu2-re (but cf. also Biligmeyer 1989). Linear A du-pu2-re is attested as part of the compounded sequence (jj)a-di-ki-te-te-du-pu2-re in two inscribed stone libation vessels (PK Za 8 and 15) from the peak sanctuary at Petsophas (Palaikastro, Eastern Crete). In the Archaic period, the same site was home to the cult of Diktaian Zeus (in reference to the Mountain of Dikte, where according to later myths Zeus was born). Since the Linear B texts from Knossos mention a Cretan deity called di-ka-ta-jo di-we 'Diktaian Zeus' already in the Late Bronze Age, I suggested that Linear A (j)a-di-ki-te-te-du-pu2-re signified 'Master of/from Dikte' in Minoan, and du-pu2-re on its own 'master, ruler'. I also proposed, as a corollary, that a Minoan derivative of du-pu2-re, plus a suffix */-nth-/, was borrowed into Mycenaean Greek as da-pu2-ri-to- 'royal place > palace'. This proposal, however, echoed certain speculations of Evans (1921: 6), following Mayer and Kretschmer's idea that Aa|3upiv0og was etymologically linked to Zeus of Labraunda, a Carian locality (Kretschmer 1896: 404). By evoking also the word Adppug 'axe' (according to Plutarch the supposed Lydian source of Labraunda), Evans identified the Labyrinth with the Bronze Age 'palace sanctuary of Knossos'. Since this identification draws strictly on formal resemblance, it cannot demonstrate any association of Linear A du-pu2-re and Greek da-pu(2)-ri-to- ~ Aa|3upiv0og with the semantics of kingship (contra Valerio 2007), so I no longer favor this view (see now Valerio 2015 and the following note here).

3 This second interpretation is not too far-fetched, considering that the meaning 'sacred hypogeum' (or sim.) would account well for Herodotus' use of Aa|3upiv0og in his description of an Egyptian underground mortuary building. In this case, Linear A (j)a-di-ki-te-te-du-pu2-re might either refer to the 'Cave of Dikte' where the later Zeus was said to have been born, or mean 'Diktaian god' (cf. Linear B di-ka-ta-jo di-we), if da-pu(2)-ri-to- stems from a Minoan word that literally signified 'godly (place) > sanctuary'.

2. Suggestions of a special sound from an Aegean substrate language

Ventris and Chadwick (1973: 310, 538) proposed the identification of the Linear B place-name (sic) da-pu2-ri-to- with Aapupiv0oc; 'Labyrinth' with a question mark. Like Palmer (1955: 40) and Heubeck (1957: 151), they considered that, if the equation was true, then the d ~ A alternation must be the reflex of an "intermediate" sound of a non-Greek Aegean language. For Ya-kubovich (2002: 109), one possibility is that this foreign sound was identical with the coronal fricative /5/. Davis (2014: 204-210) builds on this suggestion and argues for an underlying Mi-noan phoneme /0/ that also possessed a voiced allophone /5/. Kassian (2010: 362, n. 31) considers a lateral affricate /ti/.

Lejeune (1958: 327-328) came up with a slightly different hypothesis, in which the different spellings are not taken as alternative scribal choices for spelling an alleged non-Greek

sound in Linear B, but rather the direct consequence of such sound. He linked the inconsistent

orthography of da-pu2-ri-to-jo and AapuQLV0o; to two intriguing features of the Mycenaean syllabary that meant the underrepresentation of two phonemic contrasts of Greek. On the one hand, Linear B possessed a single series for the two Greek liquids, the lateral /l/ and the rhotic /r/, transliterated as r by convention. On the other hand, the script did not mark the voicing opposition in any of the stop series, with the exception of the coronals, whereby d transcribed voiced /d/ and t was used for voiceless /t, th/. As an attempt to account for both "anomalies", Lejeune hypothesized that the model of Linear B, the Linear A script, possessed three series: r = /r/, d = /l/ and t = /t/, respectively. Still according to Lejeune, this Minoan /l/ possibly was very close to the Greek stop /d/ in terms of articulation. Thus, the Mycenaeans might have borrowed the theoretical Linear A d = /l/ for writing Greek /d/, while choosing to employ Linear A r for the two Greek liquids, /r/ and /l/. Lejeune supported this hypothesis by adducing some 'Aegean' words with interchanging 5 and A in their spellings that had intrigued scholars since the 19th century — particularly the name of Odysseus (Homeric D5uoo£u; and Ionian D5uoo^o; ~ OuAL^£u;/n;, D/QAuo(o)eu;, DAut(t)£u;, DAioeuc;4) and the pair 5d^vq ~ Ad^vn 'sweet bay', which will be of importance below.5 It is to be noticed that the hypothetical /l/ that Lejeune pondered as the sound behind Linear A d has also been proposed to be the first sound of a Minoan word borrowed into Mycenaean Greek as da-pm-ri-to-. Therefore, the tacit implication of his hypothesis is that, synchronically, Linear B d would represent a Minoan lateral (in borrowings and non-Greek onomastics) in addition to the Greek stop /d/.

Lejeune formulated his idea with utmost caution and underlined that it could not be demonstrated. In fact, he took notice of two possible counterarguments (Lejeune 1958: 328). On the one hand, certain Linear A sequences had close equivalents in non-Greek personal names in the Linear B tablets, and these pairs revealed a direct correspondence LA d > LB d (cf. e.g. Linear A ku-ku-da-ra vs. Linear B ku-ku-da-ro). On the other hand, Lejeune noticed two interesting developments in the Cypro-Greek syllabary, used on Cyprus during the first millennium (for the ancient local Greek dialect) and likely derived from Linear A indirectly, through the Cypro-Minoan syllabary (ca. 1500-1050 BCE). In Cypro-Greek, a syllabogram that was

4 See Liddell and Scott (1940) and Chantraine (1999: 775). Most of the forms with A occur in Greek vases and were compiled already in Kretschmer (1894: 146-147).

5 The pair Linear B fka-da-mi-ta ~ alphabetical Greek KaAa|jiv0a 'catmint, mint' (found in Furnee 1972 and reproduced e.g. in Kassian 2010: 362, n. 31) must be excluded as a potential example of Linear B d = alphabetical A. The form fka-da-mi-ta was the old reading of a Mycenaean sequence from tablet MY Ge 604. It has been corrected to ka-da-mi-ja and is rather to be compared to KdpSa|jov, Kap5a^ig 'garden cress (Lepidium sativum)' (Bennett 1958: 81, n. 5; Ventris and Chadwick 1973: 549). I am thankful to Maurizio Del Freo for the relevant references.

clearly derived from Linear A da was used as ta /da, ta, tha/. At the same time, Linear A ro was the likely source of Cypro-Greek lo /lo/, whereas the Greek syllable /ro/ was written with another series that represented exclusively /r/ and had no antecedents in Linear A. 6 These pieces of evidence imply that Linear A d transcribed a coronal obstruent and that r was its single liquid series, thereby making Lejeune's hypothesis very difficult.

3. Synchronic variation or change?

Often not underlined in treatments of da-pu(i)-ri-to- ~ AapupivGoc; is the fact that the pair does not represent a synchronic variation in spelling. Rather, da-pu2-ri-to- is separated from AapupivGoc; by more than seven centuries. As remarked above, the two forms have been taken as evidence of a /d/ ~ /l/ alternation in Mycenaean Greek under the influence of a substrate language, yet the evidence supplied in support of this notion all comes from much later alphabetical material, such as D5uaa£u; ~ DAuaaeu;. Conversely, no examples of Linear B words in which d = /d/ and r = /l/ interchange are available. For example, we never find **ra-pu(2)-ri-to- in Linear B. In other words, there is no proof that /d/ and /l/ ever alternated in words borrowed into Greek as early as the Late Bronze Age. We also lack any compelling example of interchanging r and d in Linear A > Linear B pairs of sequences.7 Moreover, there is no reason to assume that the Linear B d series represented anything other than /d/. Even when d was used in transcriptions of foreign names or loanwords, we should expect it to represent a foreign sound that was adapted to the Greek phonology and was pronounced by most native speakers as /d/. Thus, we ought to reconstruct Mycenaean da-pu(2)-ri-to- as /daphurinthos/ (see already Lejeune 1972: 57, n. 3). It must be noticed, incidentally, that there is no unequivocal evidence that Linear B p2, even in alternation with p, stood for anything other than the aspirated /ph/.8

Once it becomes clear that there is no indication of a synchronic variation of /d/ ~ /l/ in Mycenaean words, we must consider the possibility of a diachronic phenomenon. In other words, we must explore the idea that only in post-Mycenaean times did the /d/ of da-pu(2)-ri-to-shift to /l/, yielding AapuQivGo;.

4. Word-initial lambdacism in Greek forms from Anatolia?

It is interesting that the variant with A- first appears in the work of Herodotus, who was a native of Halicarnassus, in Caria (southwestern Anatolia). The reason I emphasize this is that Anatolia is also the source of the two glosses by Hesychius that are frequently cited as instances of 5 ~ A variations and comparanda for AapupivGoc; (e.g. Heubeck 1957: 152; Ventris

6 Recently, Steele (2014) also reevaluates Lejeune's hypothesis.

7 With Davis (2014: 206), we can take note of Linear A da-ri-da vs. Linear B ra-ri-di-jo 'of/belong to ra-ri-d-', but without any guarantee that the words compared are related.

8 As I have argued elsewhere (Valerio 2015: 332, n. 6; 2016: 201-202), there is a possible way to account for the discrepancy between Linear B p(2) = /ph/ and alphabetical |3 in da-pu(2)-ri-to- ~ Aa|3upivGoc;. One can entertain that two competing Mycenaean forms existed, /daphurinthos/ and */dawurinthos/, which reflected two different Greek adaptations of a foreign (Minoan) word containing a voiceless labial fricative. One suitable typological parallel is provided by Mongolian, which adapts Russian [f] variably as [ph], [pjh] or [w] in loanwords (Svantesson et al. 2005: 31). In this scenario, the theoretical Mycenaean */dawurinthos/ would have been the source of Aa|3upivGoc;. For w > |3, cf. Linear B mo-ri-wo-do /molivdos/(?) vs. later alphabetical Greek |j6Au|35oc; ~ |j6Ai|3oc; 'lead'.

and Chadwick 1973: 310): Aá^vq = Sá^vq 'sweet bay' and AíoKog = SioKog 'discus, quoit'. Both are attributed by the Hellenistic lexicographer to the speech of the inhabitants of Perge, a city in southern Pamphylia (Latte 1956). Is it a coincidence, then, that there is some connection to Anatolia in the three cases (AapupivGog Aá^vq and aloko^) in which lambdacism takes place in initial position?

A further similar case is presented by the name of a Cimmerian or Scythian warlord who raided parts of Anatolia in the 7th century BCE. The individual in question is mentioned in the Neo-Assyrian cuneiform sources as Dugdammé ~ Tugdammé, but as AúySa^ig in the writings of Kallimachos (3rd century BCE) and Strabo (Geo. 1.3.21, 1st century BCE) (Kuhrt 1987: 187). It is true that other historical figures existed, Greek and Carian, who bore the name AúySa^ig: an Olympic champion from Syracuse (mid-7th century BCE); a tyrant of Naxos (mid-6th century BCE); and two rulers of Halicarnassus, respectively the father and the son or grandson of Artemisia (Tokhtas'ev 2007: 611-612).9 This has raised suspicions that the name is Carian, not Cimmerian (Iranian?), but regardless of its popularity in Anatolian and Greek-speaking circles, this does not mean that it must be disconnected from the name of the Cimmerian chieftain, Dugdammé. We may compare the widespread use of the name of the ill-famed Hunnish ruler Attila, even today, in countries like Hungary and Turkey. Ultimately, the etymology and source of AúySa^ig may not be too relevant, as it is undeniable that the name circulated widely in Asia Minor. Thus Kuhrt (1987: 187), crediting an oral suggestion by S. Karwiese (1984), contemplates difficulties "in rendering the specific sound of an Anatolian language" as the cause for the change T/D- > A-. We may build on this suggestion and hypothesize that Dugdammé became *Lugdam(m)i (or similar) in an Anatolian language that prohibited initial /d/, and the latter form was then captured as AúySa^ig in the Greek alphabet.

We can do more to substantiate the hypothesis that AapupivGog Aá^vq, AíoKog and possibly AúySa^ig owe to an Anatolian tendency to realize word-initial /d-/ as /l-/. Lydian and Lycian both had phonotactic restrictions for /d/ in word-initial position (see Pedersen 1945: 42; Melchert 1993: 249, 252; Van den Hout 1995: 133) and we know that Lydian adapted the Aeolic Greek divine names Aa^áx^Q 'Demeter' and Aeúg 'Zeus' as lamétru and lews/lefs, substituting /d/ with /l/ (Melchert 1994: 335, with references).10 The suitable typological parallel from a contemporary language is provided by Yaqui, a Uto-Aztecan tongue of northwestern Mexico, which reportedly replaced foreign [d] with either [r] or [l] in words loaned from Spanish: cf. Yaqui lios < Spanish Dios 'God' (Estrada Fernández 2009: 834, 844-846).

In terms of articulation, the substitution of [l] for a dental stop is unsurprising. A lateral approximant is essentially a coronal sound articulated with occlusion, the latter being the most salient feature of stops (Ladefoged and Maddieson 1996: 182-183). We might expect Greek [d] to have been adapted as Lydian [t], which is its voiceless counterpart and exists as a phoneme in the latter language. Yet the preference for [l] over [t] seems plausible if we assume that the feature of voicing was perceptually favored by speakers of Lydian, or of any other Anatolian language, when adapting foreign [d]. From this perspective, [l] obviously has the advantage.11

I therefore find it likely that Pergaean AioKog and Aá^vq correspond to standard SioKog and Sá^vn as uttered by native speakers of a local Anatolian dialect with such phonotactic re-

9 I am thankful to Zsolt Simon for this reference.

10 It is to be noticed that already Popko (2008: 136) compares Dugdamme > Auy5a^Lg with the Lydian adaptation of Greek 5- as l-.

11 Compare for example Nahuatl, another Uto-Aztecan language of Mexico: Spanish Dios was borrowed as tios certainly because Nahuatl prohibited [l] in initial position; conversely, [l] was allowed medially, so we find Nahuatl expala for Spanish espada (cf. Bierhorst 1985: 122, 321).

strictions.12 Likewise, a virtual *5apupiv0oc;, cognate with Mycenaean da-pu(i)-ri-to-, may have been the Ionian form used in Western Anatolia, including Herodotus' native Caria, before a local variant Aapupiv0oc; emerged amongst speakers who had Greek as second language and an Anatolian dialect as their mother tongue. It would not be too surprising to find a secondary dialectal form such as Aapupiv0oc; in literary works (Herodotus) and official temple inscriptions (Didyma) instead of a more conservative *5apupiv0oc;. Languages often retain two distinct words with the same etymology, one standard, and the other dialectal but borrowed into standard speech. We may compare modern Spanish huelga [welya] 'strike; repose' vs. juerga [xwerya] 'spree, binge' (cf. Pountain 2003: 283), the latter originally a Western Andalu-sian form. Occasionally, dialectal forms may even completely replace standard ones.

Since Caria is the area connected with the first examples of Aapupiv0oc; and may be where the form emerged, we need to account for another possible obstacle. Unlike Lydian or Lycian d, the fricative nature of Carian d is much more uncertain. Adiego (2007: 245-249) argues that, as in the other two Anatolian languages mentioned, Anatolian /d/ may have become voiceless /t/ (written t) in initial position in Carian. In this case, d might represent a fricative as well. Examples of word-initial Carian d are scarce, not just in the Carian documentation, but also in the indirect evidence of Greek-written onomastics. Moreover, some of the existing examples could be the outcome of underrepresented initial vowels (cf. dquq = I5ayuyo;). Yet not all of the cases can beyond doubt be analyzed along these lines, and the evidence for Carian is not as positive as that of Lycian or Lydian (Adiego 2007: 245-246). Pending further discoveries, the lack of initial /d/ in Carian should be considered as unconfirmed. In any event, it remains possible that Aapupiv0oc; comes not from Caria, but rather from another Anatolian-speaking region in Southern or Western Asia Minor, for example Pamphylia or Lydia.

5. Final remarks

I hope to have shown that there is no compelling basis to assume that the contrast between Linear B d and alphabetical A in da-pu(i)-ri-to- ~ AapupivGoç reflects a "special" Minoan phoneme that underlay Linear A d — be it a coronal fricative /0/, a lateral affricate /ti/, or even a lateral coronal fricative Rather, it is much more economical to interpret AapupivGoç and other Greek forms as the result of a relatively late and localized shift /d-/ > /l-/, possibly as the result of contact with Anatolian languages during the 1st millennium BCE.

The implications are manifold. First, we should be wary of past etymologies of da-pu(i)-ri-to-~ AapupivGoç involving forms with initial /l-/, such as the Carian epithet AappavSéuç 'of Labraunda' and the alleged Lydian word Aàppuç 'axe' (Kretschmer 1896, Evans 1921), Greek Aàupa 'alley, lane, passage' (Smith 1859, Guntert 1932), and even Aàpipoç • poGuvoç 'hole' (Smith 1859), the latter from Hesychius' lexicon (see Latte 1956). As argued above, Linear B da-pu(2)-ri-to- most probably reflects a noun ^ydaphurintNos/, in turn related to Linear A du-pu2-re, a word of uncertain meaning but connected to the religious sphere.13 Secondly, there are important ramifications for our understanding of the Aegean and Cypriot syllabaries, namely

12 Dressel (1965: 187) and Brixhe (1976: 83, n. 16) associate the 5- > A- shift in Pergaean Ad^>vr| and AioKo; with the rhoticism (5 > p) of Pamphylian Greek, whereby /l/ and /r/ would be alternative outcomes of a weakened */d/. However, as Brixhe himself notes, rhoticism affects particularly intervocalic -5- in Pamphylian (and a similar shift occurred in Luwian as well). Thus, it cannot be evoked to account for the words in question here without making additional assumptions.

13 See nn. 2 and 3 above.

Linear A, Linear B and Cypro-Minoan (see Valerio 2015, 2016). It is a key point that there is now no serious obstacle to the notion that the Linear A d series transcribed a voiced coronal obstruent /d/.14 Finally, the present results are of consequence for questions relating to linguistic interactions between the Aegean and Anatolia in the 2nd millennium BCE.

Acknowledgements

Ignasi-Xavier Adiego, Zsolt Simon and Alexius Belov made useful suggestions to improve this article. As usual, the responsibility for the views expressed here and any remaining errors is mine alone.

References

Adiego, Ignasi-Xavier. 2007. The Carian Language: With an Appendix by K. Konuk, Leiden / Boston: Brill.

Aura Jorro, Francisco. 1985. Diccionario griego-español, Anejo I: Diccionario Micenico. Vol. I. Madrid: Consejo Superior de Investigaciones Científicas.

Bennett, Emmett L. 1958. The Mycenaean Tablets II. Transactions of the American Philosophical Society 48: 1-122.

Bierhorst, John. 1985. A Nahuatl-English dictionary and concordance to the Cantares Mexicanos with an analytical transcription and grammatical notes. Stanford: Stanford University Press.

Billigmeier, Jon-Christian. 1989. The Linear A Libation Formula Revisited. American Philological Association: 121st Annual Meeting: Boston, Massachusetts, December 27-30, 1989: Abstracts. Atlanta: Scholars Press, 109.

Brixhe, Claude. 1976. Le dialecte grec de Pamphylie. Documents et grammaire. Paris: Adrien-Maisonneuve.

Cagiano, Michaelangelo. 1958. Saççio sul labirinto. Milan: Vita e Pensiero.

Chantraine, Pierre. 1999. Dictionnaire étymologique de la langue grecque. Histoire des Mots. Nouvelle edition avec supplément. Paris: Klincksieck.

Davis, Brent. 2014. Minoan Stone Vessels with Linear A Inscriptions (Aegaeum 36), Leuven/Liège: Peeters.

Dressler, Wolfgang. 1965. Pamphylisch -S- zu -Q-: ein weiterer Substrateinfluß? Archiv Orientální 33(2): 183-189.

Estrada Fernández, Zarina. 2009. Loanwords in Yaqui, a Uto-Aztecan language of Northwestern Mexico. In M. Haspelmath and U. Tadmor (eds.), Loanwords in the World's Languages: a Comparative Handbook. Berlin: De Gruyter, 823-845.

Faure, Paul. 1964. Fonctions des cavernes crétoises (École Française d'Athènes, Travaux et mémoires 14). Paris: É. de Boccard.

Fontenrose, Joseph Eddy. 1988. Didyma: Apollo's Oracle, Cult, and Companions. Berkeley: University of California Press.

Furnée, Edzard J. 1972. Die wichtigsten konsonantischen Erscheinungen des Vorgriechischen. The Hague: Mouton.

Güntert, Hermann. 1932. Labyrinth, eine sprachwissenschaftliche Untersuchung. Heidelberg: C. Winters Universitätsbuchhandlung.

Heubeck, Alfred. 1957. Linear B und das äeäische Substrat. Minos 5: 149-153.

Kassian, Alexei. 2010. Hattic as a Sino-Caucasic language. Ugarit-Forschungen 41: 309-447.

14 For a more detailed argument, see Valerio (2016: 203—209, 226, 288—293). The more probable interpretation of Linear A d as a coronal obstruent agrees well with the evidence that signs ultimately derived from the Linear A d and t series appear to be merged in the Cypro-Greek t = /d, t, th/ series, likely as a result of developments in Cypro-Minoan. Steele (2014) also revisits the problem of the dental and liquid series in the Aegean-Cypriot syllabaries, along with Lejeune's hypothesis. In addition to the scenario upheld here, she considers another possibility: Linear B d and t were the product of the splitting of what was a single coronal series in Linear A; in parallel, this single series was inherited by Cypro-Minoan and then by Cypro-Greek. This hypothesis is less likely given the comparatively greater number of (probable) Linear A d and t signs in comparison to the counterparts in Cypro-Minoan and the Cypro-Greek syllabary. This apparent reduction is more suggestive of a merger of two coronal series on Cyprus.

Kretschmer, Paul. 1894. Die griechischen Vaseninschriften ihrer Sprache nach untersucht. Gütersloh: Bertelsmann.

Kretschmer, Paul. 1896. Einleitung in die Geschichte der Griechischen Sprache. Gottingen: Vandenhoed & Ruprecht.

Kuhrt, Amelie. 1987. Lygdamis. In: E. Ebling, B. Meissner, E. Weidner and D. O. Edzard (eds.), Reallexikon der As-syriologie und Vorderasiatischen Archäologie. B. 7(3/4): 186-189. Berlin/New York: De Gruyter.

Ladefoged, Peter, Ian Maddieson. 1996: The Sounds of the World's Languages. Oxford: Blackwell.

Latte, Kurt. 1956. Hesychii Alexandrini lexicon [recensuit et emendavit Kurt Latte], Vol. II: E-O. Copenhagen: Ejnar Munksgaard.

Lejeune, Michel. 1958. Mémoires de philologie mycénienne, Vol. 1. Paris: Centre National de la Recherche Scientifique.

Lejeune, Michel. 1972. Phonétique historique du mycénien et du grec ancien. Paris: Éditions Klincksieck.

Liddell, Henry G., Robert Scott. 1940. A Greek-English Lexicon. 9th edition, revised and augmented throughout by Sir Henry Stuart Jones with the assistance of Roderick McKenzie. Oxford: Clarendon Press.

McCabe, Donald F. 1985. Dydima Inscriptions. Texts and List. Princeton: The Institute for Advanced Study.

Melchert, H. Craig. 1993. Historical Phonology of Anatolian. Journal of Indo-European Studies 21: 237-257.

Melchert, H. Craig. 1994. Anatolian Historical Phonology (Leiden Studies in Indo-European 3). Amsterdam: Rodopi.

Montegu, John C. 1976. Note on the Labyrinths of Didyma. American Journal of Archaeology 80(3): 304-305.

Palmer, Leonard R. 1955. Observations on the Linear B Tablets from Mycenae. Bulletin of the Institute of Classical Studies 2/1: 36-45.

Pedersen, Holger. 1945. Lykisch und Hethitisch. Det Kgl. Danske Vidensk. Selskab Hist.-filol. Meddelser 30/4. Munskgaard.

Popko, Maciej. 2008. Völker und Sprachen Altanatoliens. Wiesbaden: Harrassowitz.

Pountain, Christopher. 2003. Exploring the Spanish Language. London: Arnold.

Sarullo, Giulia. 2008. The Cretan Labyrinth: Palace or Cave? Caerdroia 37: 31-40.

Smith, William (ed.). 1859. A Dictionary of Greek and Roman Antiquities. 2nd Edition. Boston: Little, Brown, and Co.

Steele, Philippa M. 2014. The /d/, /t/, /l/ and /r/ Series in Linear A and Linear B, Cypro-Minoan and the Cypriot Syllabary: Some Observations. Pasiphae VIII: 189-196.

Svantesson, Jan-Olof, Anna Tsendina, Anastasia Karlsson, Vivan Franzen. 2005. The Phonology of Mongolian. Oxford: University Press.

Tokhtas'ev, Sergey R. 2007. Der Name des kimmerischen Königs Lygdamis. In: J. Cobet (ed.), Frühes Ionien. Milesi-sche Forschungen 5: 607-612. Mainz: Philipp von Zabern.

Trümpy, Catherine. 2001. Potnia dans les tablettes mycéniennes: quelques problèmes d'interpretation. In: R. Laf-fineur, R. Hägg (eds.). Potnia, Deities and Religion in the Aegean Bronze Age: Proceedings of the 8th International Aegean Conference, Göteborg University, 12-15 April 2000: 411-421. Liège: Université de Liège.

Valério, Miguel. 2007. 'Diktaian Master': A Minoan Predecessor of Diktaian Zeus in Linear A? Kadmos 46: 3-14.

Valério, Miguel. 2015. Linear A du-pu2-re, Hittite tabarna and their alleged relatives revisited. Journal of Language Relationship 13/4: 329-354.

Valério, Miguel. 2016. Investigating the Signs and Sounds of Cypro-Minoan. Unpublished PhD Dissertation, University of Barcelona.

van den Hout, Theo P. J. 1995. Lycian Consonantal Orthography and Some of Its Consequences for Lycian Phonology. In: T. P. J. van den Hout, J. de Roos (eds.). Studio Historiae Ardens: Ancient Near Eastern Studies Presented Philo H. J. Houwink ten Cate on the Occasion of his 65th Birthday. 105-141. Istanbul: Nederlands Historisch-Archaeologisch Instituut.

Ventris, Michael, John Chadwick. 1973. Documents in Mycenaean Greek, 2nd ed. Cambridge: Cambridge University Press.

Yakubovich, Ilya. 2002. Labyrinth for Tyrants. Studia Linguarum 3(1) (GS. A. A. Korolev): 93-116. Moscow: Languages of Slavonic Culture.

Мигель Валериу. Лариро'Эод и анлаутный ламбдацизм в анатолийском греческом

В статье утверждается, что пара линейное B da-pu(2)-ri-to- ~ греч. Aa|3upiv0og не отражает синхронного варьирования /d/ и III в микенском греческом, и появление ll-l вызвано позднейшим переходом. Выдвигается гипотеза, что Aa|3upiv0og происходит от неза-свидетельствованного *Sa|3upiv0og, где переход начального /d/ в /I/ произошел в южной

или западной Анатолии под влиянием некоего местного языка, в котором начальный 1й1 был невозможен. То же объяснение предлагается для Лаф\т| и Люкод, которые Ге-сихий глоссирует как пергейские (памфилийские) формы греческих 5аф\т| «лавр» и бюкод «диск», и вероятно также для киммерийского имени Dugdammë/Лuy5aцLg.

Ключевые слова: Ла|3ира'0од; микенский греческий; ламбдацизм; анатолийские языки; Лигдамис; линейное А

i Надоели баннеры? Вы всегда можете отключить рекламу.