Further Thoughts on Linear B po-re-na, po-re-si, and po-re-no-

§1. Opinions have varied and swayed regarding the interpretation of the Linear B term po-re-na. Whatever meaning is assigned, many would draw the forms po-re-si and po-re-no- into their interpretation of po-re-na, and vice versa. In this investigation I begin with the interpretation of po-re-na that appears most probable and reconsider po-re-si and po-re-no- on the basis of both internal and comparative evidence.

§2. Pylos tablet Tn 316 has been the object of close scrutiny since the early days of Linear-B scholarship and remains a document of interest for various reasons, not least of which is the recurring verbal phrase of lines 2–3 on the front of the tablet and of 1–2, 5, and 8 on the reverse: do-ra-qe, pe-re, po-re-na-qe, a-ke. The document reads as follows:

The tablet can be translated in the following way. My interpretation of the verbal phrase do-ra-qe, pe-re, po-re-na-qe, a-ke has been informed by Willi (1994–1995), who rightly recognizes, I believe, a recurring coordinated syntagm of a primitive Indo-European pattern, and, especially, by Nagy (2015 and 2017), who realizes that a formulaic parallel to this lexical concatenation is expressed in Iliad 23.509–513:

§3. In the formulaic phrase do-ra-qe, pe-re, po-re-na-qe, a-ke the form po-re-na is here interpreted as the infintive phorēnai (φορῆναι) ‘for the carrying.’ This interpretation follows the observations and suggestions of Ventris and Chadwick 1956:285 (“Though one might logically expect this last word [i.e. po-re-na] to be an unattested noun meaning something like ‘cup-bearer’, it is possible that it merely represents φορῆναι ‘to carry’.”); [5] Chantraine 1973:497 (“L’infinitif porena, qui semble attesté à Pylos [Documents, p. 285], répond exactement à l’homérique φορῆναι.”; see also p. 505); and especially Willi 1994–1995 and Nagy 1994–1995, revised and expanded in Nagy 2015 and Nagy 2017:§§100–122. [6] The infinitive phorēnai (φορῆναι) ‘for the carrying’ functions within the syntax of the tablet as, in effect, the dative of a verbal nominal, reflecting early Indo-European usage. [7] The sense is thus ‘X carries gifts and takes Y for the carrying’, where the referent of X in context is likely to be understood as Pylos, and Y refers to an unnamed individual whom the agent ‘takes’ and who is given the task of ‘carrying’ the specified vessels. The conjunction of verbs denoting conveyance that is seen here—pherein and agein (φέρειν καὶ ἄγειν) ‘to carry/bear’ and ‘to take/drive’—represents a Greek reflex of a primitive Indo-European syntagm in which the coordination of *b^her- and *h₁aǵ- (or *h₂eǵ-) expresses (respectively) the ‘carrying’ of portable plunder and the ‘taking’ or ‘driving away’ of animals or people. As this description of the Indo-European formula suggests, the actions captured by the phrase fall within the sphere of warrior activity. [8] In his studies of Pylos tablet Tn 316 Nagy (2015:§§8–9, 17; 2017:§§100–103) underscores the parallel co-occurrence of these Greek verbs in Iliad 23.512–513, which, as we noted, he offers as a parallel for the procedures described in the tablet. In the epic the setting of the action is provided by the funeral games for Patroclus. The victory in the chariot race at the games has gone to the powerful warrior Diomedes; the prize is a tripod with handles and a slave woman. Diomedes’ Argive companion and charioteer, [9] the ‘mighty Sthenelus’ (ἴφθιμος Σθένελος), takes possession of the prize and gives to his ‘comrades in arms’ (ἑταῖροι) (1) the woman ‘to take’ (ἄγειν) and (2) the tripod ‘to carry off’ (φέρειν), while he unyokes the horses. In his comments on the lines, Eustathius (Commentarii ad Homeri Iliadem 4.773) notes that here the use of agein (ἄγειν) ‘to take’ and pherein (φέρειν) ‘to carry off’ subscribes to the conventional practice of differentially applying the former verb to an ‘animate’ object (ἔμψυχος) and the latter to an ‘inanimate’ (ἄψυχος).

§4. While the recent interpretations of Willi and Nagy align with those of Chadwick 1956 and so on, by the date of the publication of the 1961 international colloquium on Mycenaean studies in Racine, Wisconsin, [10] Chadwick (1964:23) has changed his mind regarding the interpretation of po-re-na, identifying it as an “acc[usative] pl[ural],” hence, a nominal. [11] In the same volume, both Lejeune and Georgiev offer views on the form. Like Chadwick, Lejeune (1964:92) advocates for a nominal interpretation of po-re-na: “Du point de vue de la forme, po-re-na peut être soit l’accusatif (sg. ou pl.) d’un nom en –νᾱ, soit l’accusatif pl. d’un nom en -νον.” Georgiev, however, reads the form as the infinitive (1964:128)—“po-re-na = Hom. φορῆναι” (citing Ventris and Chadwick 1956:285 and Bartoněk 1959:121)— and in doing so reverses his earlier (1956) interpretation of po-re-na as a noun *phorēn (*φορην; see below). In his review of this collection of papers (i.e. Bennett 1964), Palmer (1965:315) has harsh words for Georgiev: “It is regrettable to see po-re-na still quoted as an athematic infinitive, although it has long been recognised to be a noun … ,” citing only himself (Palmer 1955) for this “long-recognized” view. [12] In his Interpretation of Mycenaean Greek Texts (first published 1963) Palmer had glossed po-re-na as “‘defilements’ (?),” [13] as in Palmer 1955 (p. 10). In his review of the Bennett volume, however, Palmer (1965) proposes a different sense, writing that (p. 320) “it would seem most plausible to take the word as referring to cult objects which can be ‘incensed’ and girded (?).” The second element of this compound description is secondary to the form po-re-no-zo-te-ri-ja (“acceptable as a festival name” [p. 317]) on Pylos tablet Un 443 + 998 (discussed below), on which tablet there also occurs the form tu-ru-pte-ri-ja that Palmer links to the root thu- and which he views as fundamentally meaning ‘to incense’ (see his discussion on pp. 316–322; tu-ru-pte-ri-ja is most commonly understood to spell struptēriā, later στυπτηρία, ‘alum’). Clearly there has been a good bit of opinion switching in the matter of the meaning of po-re-na.

§5. In the second edition of Documents in Mycenaean Greek (Ventris and Chadwick 1973:460–461) Chadwick proposes that “the po-re-na [of Tn 316] must be the ten persons who are led to the rite; though no Greek word provides an interpretation it may seem appropriate to translate as victims.” Several other investigators who regard po-re-na as a noun, both before and since Ventris and Chadwick 1973, have advocated for the sense ‘victims’. [14] Palaima can be numbered among those who contend for a nominal interpretation, though he states (1999:454) that he finds “no compelling reason why po-re-na has to refer to human victims,” [15] but does not dismiss the possibility that such is the proper reading. Palaima allows the potentiality that the sense of the term could be ‘porteur’, [16] thus connecting po-re-na with the root of the verb phorēnai (φορῆναι). As we have just seen, in 1956 Georgiev had interpreted po-re-na as accusative of a noun *phorēn (*φορην)—a view that he subsequently abandoned. Among those who view the form as a nominal, this morpho-lexical analysis of po-re-na has, however, been that one most widely held, if sometimes tentatively (thus, Palmer 1969:267: “the morphological analysis of *φορενα is unclear”). [17]

§6. As we have seen, in his investigation of the recurring phrase do-ra-qe, pe-re, po-re-na-qe, a-ke, Nagy, along with Willi, contends for the reading of po-re-na as the infinitive phorēnai (φορῆναι). In doing so Nagy responds to objections to his arguments that appear in Palaima 1996–1997 and 1999 (see also Palaima 2011:66), pointing out the significant problem that an envisioned o-grade nominal *phorēn (*φορην) would be of a type without morphological parallel in Greek (Nagy 2015 §§ 20–25), a hobbling hardship undoubtedly reflected in the vacillations and uncertainties that characterize the treatments of earlier investigators. While advocating for this interpretation Palaima too acknowledges the difficulty (1999:454n57): “The o-grade treatment in *po-re = *φορην is problematical no matter whether one interprets the word ‘actively’ as ‘he/she who carries’ or ‘passively’ as ‘he/she/it who/which is brought.” [18] Contrast with an aberrant *phorēn (*φορην) the expected e-grade seen in phernē (φερνή), Aeolic pherena (φέρενα), [19] denoting ‘dowry; bridal gift’—i.e. that which a bride brings. To explicate the meaning Joannes Tzetzes, Exegesis in Homeri Iliadem 115.71, makes recourse to the feminine participle hai pherousai (αἱ φερούσαι) ‘those bearing’: φερένας καὶ φρένας τὰς φερούσας νοῦν. Doric shows a form pherna (φερνά), attested at Epidaurus (IG IV², 1 40.6–7; IG IV², 1 41.7–8) and used to signify the portion of an offering that is dedicated to a deity—consistent in sense with use of the formula of offering presentation of Pylos tablet Tn 316, in which the infinitive phorēnai (φορῆναι) occurs.

§7. In arguing that po-re-na spells the infinitive phorēnai (φορῆναι), Nagy (2015 §4) draws attention to two important dialect considerations: (1) the refashioning of finite verbs terminating in – éō (-έω) as athematic forms ending in – ēmi (-ημι) and (2) the formation of the corresponding infinitives in -ēnai (-ῆναι) represent an innovation associated with the Arcado-Cypriot dialect group of the first millennium BC. [20] This Arcadian dialect feature is visible in the Homeric Kunstsprache: both Eustathius, in his commentary on the Iliad, and Homeric scholia draw explicit attention to the athematic finite verb phorēmi (φόρημι) and its relationship to the infinitive of the form phorēnai (φορῆναι). [21] That the Greek of the Linear B documents shares the innovation—as suggested by phorenai (φορῆναι)—gives evidence of a particular dialect relatedness between that second-millennium linguistic system and Arcado-Cypriot. This specific observation is consistent with an otherwise endorsed view of the palpable closeness of Mycenaean Greek and Arcado-Cypriot. [22] Nagy is careful to suggest (2015:§5) that it is the standard dialect of Mycenaean that is especially closely related to Arcado-Cypriot—the Mycenaean dialect which has been called Normal Mycenaean, as opposed to Special Mycenaean. Nagy’s is an important proposal, for it spells out that whatever Special Mycenaean is—in terms of affiliation with the known first-millennium BC dialects—it is not the dialect ancestral to Arcado-Cypriot. [23]

§8. Thebes tablet Of 26 preserves the form po-re-si, an apparent dative plural, which has been commonly cited in support of a nominal interpretation of po-re-na. [24] Tablet Of 26 records consignments of wool (ku LANA) preceded by a grammatical directive (either an allative in -de or a dative-case form) that signals the recipient:

Toward identifying the recipients in these lines we can rewrite them with partial translation in the following way:

As the translation indicates, allative expressions appear to occur twice, when the wool is destined for a do (commonly interpreted as ‘house’, i.e. dō [δῶ])—that of Ko/Qi-de-wa and that of Di-u-ja-wo. These forms have been construed as genitives, given the context provided by this and related tablets: Ko/Qi-de-wa is not otherwise attested in the Mycenaean documents; Di-u-ja-wo is most likely genitive plural of Diwyarwos (Δίϝyαρϝος) ‘priest of Diwia’. [25]

§9. Commonly, however, recipients appearing on the Thebes Of tablets look to be marked by the dative case. Palaima (1996–1997:308–309) points out that the use of allative expressions (and locatives) alongside datives is frequently found in Mycenaean allocation records, particularly those dealing with matters of cult. [26] In the instance of the Thebes Of series, dative singular morphology can be identified with reasonable confidence in the case of the names written Pa-pa-ra-ki (Of 25); A-re-i-ze-we-i (Of 37); and Qa-ra2-te (Of 38). [27] The dative plural ma-ri-ne-we-ja-i, a derivative of a man’s name, or possibly of a theonym, [28] ‘to the women of Ma-ri-ne-u’, is found twice in these materials (Of 25; Of 35).

§10. Del Freo and Rougemont (2012:270), extrapolating from Hiller (1987:245–246), are most likely correct in proposing that various feminine appellatives appearing in the Of tablets are to be read as dative singular (as opposed to nominative plural, in light of the absence of secure nominative plurals “of rubric” in the Of series, in contrast to the presence of secure dative plurals). [29] This nominal set consists of at least the following: (1) a-ka-i-je-ja, a derived adjectival form found twice on Of 27, modifying du-qo-te-ja and pu ₂ -ke-qi-ri-ne-ja (these being perhaps names of women); [30] (2) a-pi-qo-ro, identifying an ‘attendant’, [31] and a-ra-ka-te-ja ‘spinner’ on tablet Of 34, the two contrasting as, in some sense, ne-wa versus pa-ra-ja, ‘new’ versus ‘old’, respectively; [32] (3) te-pe-ja, on tablet Of 35, denoting a maker of a kind of cloth (te-pa, a densely woven fabric) and written sequentially after Ko-ma-we-te-ja, the theonym (Komāwenteiā) found on the reverse of Tn 316, naming one of those deities to whom gifts are carried; (4) no-ri-wo-ki-de on Of 36 (line 1) apparently designating a woman who produces another variety of cloth; [33] (5) also on Of 36—the nominal a-ke-ti-ra ₂ ‘decorator’, listed twice as a recipient of wool, once (line 2) in conjunction with the allative phrase po-ti-ni-ja, wo-ko-de ‘to the woikos of Potnia’; Chadwick suggests that in the Mycenaean documents woikos (in the allative wo-ko-de), in opposition to do (in the allative do-de), may refer to the “house of a deity” or “temple”. [34]

§11. Beyond these, other recipients are probably recorded in the dative case on tablets of the Of series from Thebes. Succinctly, this set consists minimally of the following men’s names: Pu2-re-wa (Of 26); [35] Su-me-ra-we-jo (Of 26); [36] Pi-ro-pe-se-wa (Of 28); [37] I-da-i-jo (Of 28); [38] Ku-ru-me-no (Of 33) [39] —and perhaps also Ne-e-to (Of 38) and Ṇẹ-a2-ri-da (Of 39). [40]

§12. In addition to Potnia and Komāwenteiā (and possibly Ma-ri-ne-u), yet other deities appear in the Thebes Of series, and are likely marked as dative recipients. On tablet Of 28 wool is consigned E-ra, most probably ‘for Hera’, whose name is seemingly modified by an epithet ke-o-te-ja. The fragmentary tablet Of 31 appears to preserve allative phrases in lines one and two, followed in the third line by the form E-ma-a ₂ ‘for Hermāhās’ (Hermes)—yet another god for whom gifts are reported as carried on Tn 316.

§13. This brings us to a closer consideration of the form po-re-si of tablet Of 26. We have seen now that it co-occurs with two allative phrases and two probable datives of proper names (Pu ₂ -re-wa and Su-me-ra-we-jo). In light of the evidence regarding the marking of recipients of wool on Of tablets in the dative case, po-re-si must certainly be read as dative, and this seems to have been the default parsing for most investigators. [41] Those who interpret po-re-na as a nominal on Tn 316 (rather than as an infinitive) would see in po-re-si a dative plural of that same nominal. Hiller (2011:182), for example, writes that po-re-si “is obviously the dat. (pl.) of po-re-na, designation of persons (victims, bearers of gold vessels?) who appear in clearly religious function on the Pylos ‘pantheon tablet’ Tn 316.” [42] In order to understand the term as a dative denoting ‘ones who bear [gold vessels etc.]’ it is not, however, necessary to interpret po-re-si as a form of the problematic, conjectured noun *phorēn (*φορην).

§14. As we have seen, it is the Arcadian dialect of the first millennium BC that is crucially significant in elucidating the morphology of second-millennium po-re-na. Verbs that end in -éō (-έω) in most dialects appear as athematic verbs in -ēmi (-ημι) in Arcadian, with corresponding infinitives formed in -ēnai (-ῆναι). Linear B po-re-na can be understood to spell this infinitive phorēnai (φορῆναι). In his discussion of Arcadian verb morphology, Dubois (1988:143) draws attention to an accusative singular participle kuensan (κυενσαν; IPArk 34.12), comparing the Attic inscriptional correspondent kuōsan (κυõσαν; SEG 33:147.39, 44), participle of the thematic contract verb kueō (κυέω) ‘to bear in the womb, be pregnant with’. The Arcadian participle kuensan (κυενσαν) clearly points to an athematic finite verb *kuēmi (*κύημι). [43] In other words:

Correspondingly, Linear B po-re-si on Thebes tablet Of 26 must certainly spell the dative plural participle phor-en-si (φορ-εν-σι), [44] recording an allotment of wool assigned ‘to/for those who carry’. Compare, with Ionic thematic morphology, the participle, for example, of Iliad 8.89, where the charging chariot steeds of Hector are described as θρασὺν ἡνίοχον φορέοντες (phoreontes) Ἕκτορα ‘those that carry the recklessly bold charioteer, Hector’.

§15. Aeolic shares with Arcadian the -ēmi (-ημι) athematic inflection of verbs that terminate in -éō (-έω) in Attic-Ionic and elsewhere (but not the corresponding infinitive in -ēnai [-ῆναι]). Thus, Alcaeus fr. 41.10 (Lobel and Page) preserves ]phoren[t]es [ (]φόρεν[τ]ες [) ‘ones carrying’, a nominative plural participle of phorēmi (φόρημι). One would have expected Aeolic *phorēntes (*φόρηντες) and perhaps the reading should be emended accordingly. Regardless, the significance of this form for understanding Mycenaean po-re-si can hardly be overstated. The context of Alcaeus’ participle is one having cult indications, with surrounding fragmented references to notions agnai (ἄγναι) ‘pure’ (l. 7); iran (ἴραν) ‘sacred’ (l. 9); oin[o]n (οἶν[ο]ν) ‘wine’ (l. 11); kitharis (κίθαρις) ‘lyre’ (l. 14); te]menos lakhois[a (τέ]μενος λαχοισ[α) ‘having obtained a temenos’ (l. 17); k]oruphan polēos (κ]ορύφαν πόληος) ‘peak of the polis’ (l. 18); Aphrodita (Ἀφρόδιτα) ‘Aphrodite’ (l. 19). We surely find ourselves here enmeshed in a lexical web of Lesbian ritual [45] that continues among its constituent strands an ancestral Mycenaean cult term for those whose role it is ‘to bear’—phorēnai (φορῆναι)—offerings for the gods. The quasi-official nature of the participle’s use in Mycenaean cult terminology is revealed by the alternation of po-re-si with forms such as po-ti-ni-ja wo-ko-de ‘to the woikos of Potnia’ and names of divine and priestly recipients in the Of series as discussed above. Compare here the term ka-ra-wi-po-ro (κλαϝι-φόρος) ‘key bearer’, naming a sacred officiant at Pylos.

§16. Regarding the short vowel of the reading ]phoren[t]es [ (]φόρεν[τ]ες [) of this fragment, compare later spelling variation seen in three Lesbian inscriptions: (1) IG XII,2 15.18 from Mytilene (ca. 193 BC) shows a genitive plural participle katoikēntōn (κατοικήντων), from athematic *katoikēmi (*κατοίκημι), Attic katoikeō (κατοικέω) ‘to settle, dwell in’; (2) SEG 36:750.17, also from Mytilene, but earlier (ca. 340–330 BC), [46] attests a short-vowel variant of the participle, katoikentōn (κατοικέντων); (3) IG XII Suppl. 692.23 from Eresos (second century BC) similarly preserves katoikenṭ[ō]ṇ (κατοικέντ̣[ω]ν̣). [47] Some conditioned shortening of the suffixal vowel is perhaps suggested in the context created by the participial morphology by the later fourth century. [48]

§17. There is at least one—possibly two—additional Linear B form(s) in which a purported nominal po-re-na has been judged to play a role. As mentioned above Pylos tablet Un 443 + 998 records the entry po-re-no-zo-te-ri-ja, followed by a specification of a measure of wool. The tablet reads as follows:

We saw that Palmer (1965) segments po-re-no from the syntagm/compound and connects the remaining morphology (zo-te-ri-ja) with that family of terms headed by the verb zōnnumi (ζώννυμι) ‘to gird’, identifying po-re-no-zo-te-ri-ja as the name of a festival. [49] Many investigators have proceeded likewise. [50]

§18. Towards making sense of Linear B po-re-no (in po-re-no-zo-te-ri-ja), it is important to take account of the Sanskrit derivative nominal bharaṇa-. [51] The Sanskrit verbal root bhar- (from PIE *b^her-) ‘to bear, carry’ is cognate with Greek pher- (φερ-)—thus, Sanskrit bharati, Greek pherō (φέρω), the finite verbs. With the Sanskrit thematic nominal bharás ‘bearing’ (adjective) and bháras ‘a bearing away, plunder’ (noun), from e-grade *b^her-, compare Greek phorós (φορός) ‘bearing’ and phóros (φόρος) ‘payment, tribute’ (i.e. ‘that which is brought’), from o-grade *b^hor-. With these o-grade forms of Greek compare Sanskrit bhāra- ‘burden; labor; bulk’. For Greek e-grade nominal derivatives consider pher-ma (φέρ-μα) ‘fetus; harvest’ (Sanskrit bharman- ‘support; nourishment’), in addition to the above-mentioned phernē (φερνή), Aeolic pherena (φέρενα) ‘dowry; bridal gift’, and pherna (φερνά), denoting the deity’s portion.

§19. Sanskrit bharaṇa- is derived from bhar- by means of the suffix -ana-, descended from a primitive Indo-European formant *-e/ono-. The Sanskrit suffix produces two morpho-semantically distinct formations, depending on accent placement, both of which constitute nominals with conspicuous verbal qualities. Thus, on the one hand, when the accent falls on the root, a neuter noun is derived that signifies the result of an action: for example, vácana– ‘word’ (from vac- ‘to speak’), káraṇa– ‘deed’ (from kr̥- ‘to make, do’). On the other hand, when the accent falls on the -ana- suffix (i.e. –aná-), the nominal derived serves as an adjective or “agent noun”: for example, vacaná– ‘speaking’, karaṇá– ‘active, skilled’ (i.e. ‘doing’). [52] Avestan provides evidence of the inherited formant as well: for example, ham-ərəna- ‘battle’ beside Sanskrit sam-áraṇa- ‘battle’ (from r̥- ‘to go towards, attack’; cf. Greek or-nu-mi [ὄρ-νυ-μι] ‘to incite, rush on’); varəna- ‘choice, belief’ beside Sanskrit varaṇá- ‘choosing’. [53]

§20. The formant is clearly of primitive Indo-European origin, leaving reflexes outside of Indo-Iranian. [54] Comparable formations occur regularly and plentifully in Slavic built with the e-grade of the Indo-European formant (i.e. *-eno-) and functioning as past passive participles, as in Old Church Slavic nes-enŭ ‘carried’. [55] The formant similarly survives in Germanic, typically attested as reflexes of the o-grade, though the e-grade variants are widely, if not commonly, preserved. In his examination of the variable survival of *-eno- in Germanic, Nielsen (1992:641–642) [56] identifies past participles of this e-grade form from across the Germanic family, such as the following: Old English binumine ‘taken away’, forsleginum ‘struck down’; Old Frisian fendsen ‘caught’, hwendsen ‘hung’; Old Norse gripinn ‘seized’, tekinn ‘taken’; Early Runic faikinaz ‘threatened’, haitinaz ‘called’; Gothic fulgins ‘secret’ (i.e. ‘hidden’), aigin ‘property’ (i.e. ‘owned’); and probably Old High German abasnitine ‘cut off’ and Old Saxon bismitin ‘soiled’ and kumin(a) ‘come’. [57]

§21. Morphologically, Sanskrit bharaṇa-, denoting ‘bearing’; ‘the act of bearing (in the womb), bringing; (hence) payment’, [58] suggests a Greek o-grade cognate phoreno- (φορενο-). Sanskrit bháraṇa- surfaces in the Rig Veda—in hymn 10.36, “an extremely obscure hymn” dedicated to All Gods, but principally a song in praise of Agni. [59] The locative bharaṇe occurs in the sixth stanza, in a reference to gods carried within a womb (seemingly the womb of Agni; see 6c–d): bharaṇa- here appears to signify the fetal ‘burden’ of the womb, [60] or else the action of bearing a fetus (i.e. of being in the condition of pregnancy). [61] If, in fact, po-re-no-zo-te-ri-ja denotes a festival, and if Linear B po-re-no- spells phoreno- (φορενο-), o-grade equivalent to bharaṇa-, as appears probable, then the festival so named could literally (componentially) denote something like a ‘girding of fetus-bearing’, that is, ‘of childbearing’. For semantic similarity within Greek compare the, chiefly, o-grade forms phor-a (φορ-ά) ‘gestation; productiveness’, phor-as (φορ-άς) ‘fecund; brood-mare’, phor-imos (φόρ-ιμος) ‘fertile’, pher-ma (φέρ-μα) ‘fetus’. [62] The e-grade structural equivalent survives in Aeolic pherena (φέρενα), ‘dowry; bridal gift’, the exact cognate of Sanskrit bharaṇa-. [63] One might possibly understand the po-re-no-zo-te-ri-ja as a communal celebratory event dedicated to clothing women in a way that makes an outward declaration of a fetus borne within them. Such a ‘girding of childbearing’, one might imagine, may simply be an “un-girding” or a girding with something other than a conventional belt (on pregnant women depicted as wearing unbelted garments on archaic Greek votive plaques see Lee 2012:26–28).

§22. More likely, however, po-re-no-zo-te-ri-ja would name a festival at which women who had given birth were ritually and symbolically re-girded following birth. The anthropological primitive associating untying and unbinding with childbirth has been carefully explored by Bettini (2013:69–82, see especially pp. 70–74), who draws attention to how a woman’s act of ungirding as labor begins became, in antiquity, a metaphor for birthing, as seen, for example, in Callimachus Hymns 4.209 (λύσατο δὲ ζώνην ‘she loosened her belt’) and 4.222 (μίτρην ἀναλύεται ‘she is undoing her girdle’), used of Leto (mother of Apollo and Artemis). The lexical concatenation of luō + zōnē (λύω + ζώνη), as in 4.209, finds expression in the adjective lusizōnos (λυσίζωνος) ‘loosening the belt’, used as an epithet of Eileithyia in her role as goddess who comforts and brings women through childbirth (Theocritus Idylls 17.60; Cornutus De natura deorum 73; Orphic Hymns 2.7–9), and similarly of Artemis (Libanius Epistulae 371.4; Hesychius Λ 1443). [64] The metaphor is encountered in the form luein mitrēn (λύειν μίτρην) ‘to loosen the girdle’ in the Argonautica of Apollonius Rhodius (1.288): a scholiast on the line explains that ‘women giving birth for the first time loosen their girdles and dedicate them to Artemis; for which reason there is also a temple of Artemis Lusizonos in Athens’. [65] Lee (2012:33–36; 2015:213–214), following Morizot 2004, draws attention to a fourth-century BC votive plaque from Echinus that depicts worshippers of Artemis presenting an infant before an image of the goddess, with a variety of gowns—votive offerings, seemingly—shown as suspended within the goddess’ shrine. She also notes the practice of women offering various garments to Artemis in her sanctuary at Brauron. [66] On textile dedications to deities recorded in the epigrams of the Greek Anthology see Table 32 in Brøns 2016. [67] Seven such dedicatory epigrams are specified as occasioned by childbirth: in each of these instances the recipient deity is either Eileithyia (three times) [68] or Artemis (four times); [69] and garments offered include belts, undergarments, breastbands, hairbands, chitons, pepla—among still other items, including, commonly, sandals, the loosening of which is conspicuous in the sympathetic context of easy birthing (see Bettini 2103:71–74). A cult setting for Pylos tablet Un 443 + 998 (on which po-re-no-zo-te-ri-ja occurs) is suggested by the occurrence of the name Ka-pa-ti-ja (Karpathiā) on line three (see above), naming a woman who contributes a large quantity of barley, perhaps for the celebration of the festival; a cult official—a ka-ra-wi-po-ro (κλαϝιφόρος) ‘key bearer’—of the same name appears on Pylos tablets Eb 338 + fr. and Ep 704. [70]

§23. The apparent mention of alum (tu-ru-pte-ri-ja) in line one of tablet Un 443 + 998 could possibly be pertinent in regard to cult and fertility. As is well known, alum could be used in the process of coloring wool in order to make the dye take, [71] and perhaps the mention of alum here is significant only in that way. Perhaps wool was required to be dyed a particular color for use in a rite associated with the po-re-no-zo-te-ri-ja festival. But alum, a strong astringent, has other uses, notably in the realm of healing. For example, in the Hippocratic corpus its use in various mixtures is prescribed as a styptic in the treatment of ulcers (e.g. De ulceribus 12, 14), of hemorrhoids (De haemorrhoidibus 7, 8), and of fistulae or a prolapsed rectum (De fistulis 3, 7, 9, 10). But of greater relevance in the present context are prescribed gynecological usages; thus: (1) if a woman has had difficulty conceiving, vaginal insertion of wool soaked in an unguent and alum is prescribed in order to promote pregnancy (De natura muliebri 53); (2) the same condition of infertility can addressed with a mixture of alum, bull’s gall, and burnt deer horn (De mulierum affectibus 225); (3) among prescribed pessaries is one of Egyptian alum wrapped within wool, and another of butter, alum, and honey applied to a linen cloth (De natura muliebri 97).

§24. While Linear B po-re-no- can be reasonably read as phoreno- (φορενο-), cognate with Sanskrit bharaṇa-, the evidence for survival of the primitive Indo-European suffix *-eno- in Greek is meager. [72] We have already noted the Aeolic feminine pherena (φέρενα), denoting that which a bride brings. Aeolic thus continues not only the Mycenaean participle po-re-si (Lesbian ]φόρεν[τ]ες [) ‘ones carrying’) but an e-grade form of Mycenaean po-re-no- as well.

§25. Reflexes of o-grade *-ono- appear to be slightly more common. The following examples can be identified: (1) kl-ono-s (κλ-όνο-ς) ‘confused motion; throng’ (beside kelomai [κέλομαι] ‘to urge, exhort’), from Proto-Indo-European *kel- ‘to drive, set in motion’; [73] (2) thr-ono-s (θρ-όνο-ς) ‘seat; oracular seat’, beside Linear B to-no (i.e. thor-no-s), as well as to-ro-no-wo-ko (probably thr-ono-worgoi ‘seat makers’), from an Indo-European root *d^her- ‘to hold firm’, also source of, inter alia, Sanskrit dharaṇa- ‘supporting; support’ (from *d^her-eno-) and dharma- ‘what is established; law’ (from *dher-mn̥-); [74] (3) khr-ono-s (χρ-όνο-ς) ‘time’, of uncertain origin; [75] (4) amp-ekh-ono-n (ἀμπ-έχ-ονο-ν; and ampekhonē [ἀμπεχόνη]) ‘shawl; clothing’ (i.e. ‘that which encloses’), from Proto-Indo-European *seǵ^h– ‘to hold fast’ (i.e. Greek ekhō [ἔχω]), compare Sanskrit abhi-ṣah- ‘to overwhelm’, abhi-ṣā-ta- ‘gained’(Rig Veda 5.41.14); (5) possibly phth-ono-s (φθ-όνο-ς) ‘malice’. [76] Probable feminine nominals include these: (6) hēd-onē (ἡδ-ονή) ‘enjoyment, pleasure’, from primitive Indo-European *sweh₂d- ‘to be sweet, pleasant’; compare Sanskrit svād-ana-, ‘tasting’ (neuter noun) and ‘making savory’ (adjective); [77] (7) per-onē (περ-όνη) ‘pin (of a buckle, etc.)’ (beside peirō [πείρω] ‘to pierce’), from Proto-Indo-European *per- ‘to pass through’, [78] and compare Sanskrit par-aṇa- ‘crossing’ and pār-aṇa- ‘bringing over’ (adjective) and ‘carrying through’ (neuter noun); (8) bel-onē (βελ-όνη) ‘needle’, from Proto-Indo-European *g^wel- ‘to throw; pierce’; [79] (9) ak-onē (ἀκ-όνη) ‘whetstone’, from Proto-Indo-European *h₂eḱ- ‘sharp’. [80]

§26. In its relative rarity in Greek the *-e/ono- suffix fundamentally parallels the status of the related formant *-no-. Primitive Indo-European -*no- competes with, or otherwise varies with, the more common suffix *-to- as a formant used to derive verbal adjectives. Ancestral *-to- and *-no- both survive in Sanskrit, in which language the distribution shows some sensitivity to phonological context, [81] and are there used to form past participles, [82] as in Avestan as well. [83] The reflex of *-no- appears in Sanskrit past participles such as chinná- ‘cut’, kīrṇá- ‘scattered’, bhugná- bent’, pūrṇá- ‘full’, and so on (about 70 examples); [84] much more common is the survival of *-to-, as in gatá- ‘gone’, śrutá- ‘heard’, srutá- ‘flowed’, tatá- ‘stretched’, hatá- ‘slain’, and so forth. A similar imbalance of frequency holds in Avestan. [85] Greek too continues *-to-, using it—somewhat as in Indo-Iranian, though not identically—to form verbal adjectives; thus Greek cognates survive for each of the Sanskrit forms just cited: batós (βατός) ‘passable’, klutós (κλυτός) ‘renowned’, hrutós (ῥυτός) ‘flowing’, tatós (τατός) ‘able to be stretched’, phatós (φατός) ‘slain’ (respectively). In addition, in Sanskrit, as in Avestan, *-no- also forms a few adjectives that are not attached to verb paradigms as productive synchronic formations—as well as some nouns. For example, Sanskrit yajñá- ‘worship, devotion, prayer’ exists alongside a past participle iṣṭá-, from yaj- ‘to worship, consecrate’; uṣṇá- ‘hot’, beside uṣṭá-, ‘burnt’, from uṣ- ‘to burn’. The Sanskrit reflexes of *-no- thus also bifurcate morpho-semantically, though not in exact parallel to those of *-eno-.

§27. While Greek reflexes of *-to- are plentiful, the comparable use of *-no- is attested by only a relatively few forms, [86] much as in Indo-Iranian. In Greek *-no- gives rise chiefly to, again, verbal adjectives. [87] One of the most conspicuous examples is provided by hagnos (ἁγνός) ‘sacred, holy’, matched precisely by the Sanskrit substantival cognate yajñá- ‘worship, devotion, prayer’, occurring with great frequency in the Rig Veda, and by Avestan yasna- ‘sacrifice’. The etymon is a primitive root *yaǵ-, and both Greek and Indo-Iranian attest verbal reflexes as well: Greek hazomai (ἅζομαι) ‘to stand in awe of’; Sanskrit yajati ‘to worship; offer’; Avestan yazaite ‘to honor’. [88] Greek hagnos (ἁγνός) ‘sacred, holy’ is archaic and clearly a form inherited from the cult language of an earlier Indo-European moment, as are its Indo-Iranian cognates. It is regularly used attributively with theonyms, as with Artemis (i.e. Ἄρτεμις ἁγνή etc.), [89] Demeter, [90] Persephone/Kore, [91] Apollo/Phoebus, [92] Zeus, [93] Pallas, [94] Hestia, [95] Helios, [96] Cypris, [97] Muses, [98] Nereids, [99] Hades, [100] Eumenides. [101] The form hagnos (ἁγνός) is well attested early—Homer (though only in the Odyssey), Hesiod, the Homeric Hymns, lyric, Pindar, and Aeschylus all know it. At some moment the -no- morphology of the form was repudiated, in a sense, and a by-form hagios (ἅγιος) was created; the nonce form is well attested from Herodotus onward, [102] though without replacing hagnos (ἁγνός) in antiquity. The two terms can co-occur, as in Isidorus Hymns to Isis 3.2: ‘O holy (hagnos [ἁγνός]) Isis—holy (hagios [ἅγιος]), great, greatly-named Deo’—and found even within the same phrase structure, as in Orphic Hymns 41.7 ‘revealer of the holy (hagios [ἅγιος]) marriage-bed of holy (hagnos [ἁγνός]) chthonic Zeus’. In each of these instances hagnos (ἁγνός) continues its conventional, and undoubtedly liturgical, attributive function.

§28. With hagnos (ἁγνός) compare the nearly synonymous Greek semnos (σεμνός) ‘revered, holy’. Semnos (σεμνός), first attested in the Homeric Hymns and lyric, [103] is a -no- derivative of the verb root seen in sebomai (σέβομαι) ‘to worship; feel shame’, from ancestral Indo-European *tyeg^w– ‘to withdraw in awe’. [104] The Greek verb is of common origin with Sanskrit tyajati ‘to stand back from something’, having a *-to- past participle tyaktá- ‘left’, [105] with which compare the formal Greek equivalent septos (σεπτός) ‘august’ (earliest in Aeschylus Prometheus Bound 812). The reflexes of *yaǵ- and *tyeg^w– form a tight semantic set, and the two roots appear to have their beginning in a period of Graeco-Indo-Iranian linguistic and cultural unity. The utilization of the rare -no- suffix in the derivation of semnos (σεμνός) could perhaps be attributed to morpho-semantic influence of hagnos (ἁγνός) on a lexeme with which it shares cult usage; though that a common ancestral -no- formation was eliminated in Indo-Iranian in favor of -to- is equally probable, if not more so, in light of Sanskrit phonological sensitivities and given the divergence in function of primitive *-no- exhibited between Greek and Indo-Iranian. In parallel with hagnos (ἁγνός), the adjective semnos (σεμνός) is frequently used as an attributive modifier of divine names: thus we find semnos (σεμνός) so used with, for example, the theonyms Athena/Pallas, [106] Poseidon, [107] Zeus, [108] Demeter, [109] Persephone, [110] Artemis, [111] Apollo/Phoebus, [112] Hera, [113] Aphrodite/Cypris, [114] Dione, [115] Prometheus, [116] Thetis, [117] Charities, [118] Hades, [119] Erinyes/Eumenides. [120]

§29. To this set of two Greek –no– formations can be added deinos (δεινός) ‘terrible’, derived from the root of the verb deidō (δείδω), originating in a perfect *de-dwoi-a (*δε-δϝοι-α), from Indo-European *dwei- ‘to fear. [121] Sanskrit preserves, with an -s- extension of the root, the finite verb dveṣṭi ‘to hate, be hostile’, producing a -ta- past participle dviṣṭa- ‘hated’; compare the Avestan verb dvaēš– ‘to be hostile’. The Sanskrit past participle appears in the Rig Veda (9.73.5) in the compound Indradviṣṭa- ‘hated by Indra’, describing those spiritually hostile ones who are destroyed by a sacred sound associated with the god Varuṇa and with the sacrificial material called Soma. As with hagnos (ἁγνός) and semnos (σεμνός), the -no- derivative deinos (δεινός) can be used as an attributive modifier of theonyms: Ares/Enyalius, [122] Charybdis, [123] Athena, [124] Artemis, [125] Aphrodite/Cypris, [126] Phoebus, [127] Persephone, [128] Keres, [129] Eris, [130] Styx, [131] Erinyes, [132] Hades, [133] Echidna. [134] In addition, a recurring syntagm of Greek is deinos/deinē theos (δεινὸς / δεινὴ θεός) ‘terrible god(dess)’, found especially in epic and tragedy and used to denote a variety of deities, most often female: Athena/Pallas, [135] Calypso, [136] Circe, [137] Thetis, [138] Hecate, [139] Eris, [140] Nymphs, [141] Keres, [142] Erinyes, [143] Rhea, [144] Nemesis, [145] Echo, [146] Apollo, [147] Zeus, [148] Dionysus, [149] Helios, [150] Hypnos and Thanatos, [151] Triton, [152] Ares, [153] ‘Grief’ deified, [154] unnamed gods. [155] Compare with the syntagm deinos/deinē theos (δεινὸς / δεινὴ θεός) ‘terrible god(dess)’ the compound adjective theou-dēs (θεου-δής) made with the related formant –dēs (-δής) and referencing the proper holding of the gods in fear. [156] It recurs in formulaic lines at Odyssey 6.120–121, 8.575–576, 9.175–176, and 13.201–202, in which theou-dēs (θεου-δής) stands in opposition to the descriptor oude dikaioi (οὐδὲ δίκαιοι) ‘not just’, and aside from that found at 19.109 and 364.

§30. In a brief but wide-ranging article, Singh (1995:257–258) succinctly contends that in epic usage deinos (δεινός) belongs to the “domain of the sacred and not the profane” (perhaps with a nod to Benveniste). [157] He takes note of the occurrence of the phrase deinos theos (δεινὸς θεός) at Iliad 4.514, to which, as we have just seen, several other examples could be added, and also of the phrase ‘terrible portents of the gods’ (δεινὰ πέλωρα θεῶν) at Iliad 2.321 (with reference to birds being devoured by a serpent that was then turned to stone, all in the midst of a sacrifice). He mentions too the existence of the similar genitival phrase deinoio pelōrou (δεινοῖο πελώρου); this observation could also be fleshed out a bit. The phrase is found several times as a syntagmatic unit in Homeric and Hesiodic poetry; serpentine contexts are typical. It occurs at Iliad 5.741–742 with reference to the head of the Gorgon upon the aegis (ἐν δέ τε Γοργείη κεφαλὴ δεινοῖο πελώρου | δεινή τε σμερδνή τε, Διὸς τέρας αἰγιόχοιο), [158] head of a ‘terrible monster’, and a kephalē (κεφαλή) ‘head’ which is itself further characterized by the adjective deinē (δεινή) in coordination with its (near) synonym smerdnē (σμερδνή, term to which we shall soon return); for the syntagm compare Odyssey 11.634 and Shield 223–224 (of the Gorgon), Theogony 856 (of Typhoeus; cf. 825 and 829), but also Odyssey 10.168 (of a great stag). Singh also draws attention to the recurring phrase deina … epea (δεινὰ … ἔπεα) ‘terrible words’ within formulaic lines at Iliad 5.439, 16.706, and 20.448, though fails to note what must be a significant factor for his claim, namely that the first two of these (and compare 16.787–789) are the terrible enunciations of Apollo—speech acts, in effect, that repulse Greek warriors—and in the remaining instance (20.448) it is, in a case of poetic inversion, Achilles who speaks such words to an Apollo-adumbrated Hector. We should also add for consideration Odyssey 8.405–412, lines in which Euryalus, in a ritual setting of gift-giving, invokes the ‘Storm winds’ (ἄελλαι) to seize and bear away any epos … deinon (ἔπος … δεινόν) ‘terrible word’ that had been ‘uttered’ (the archaic verb bazō [βάζω]) against Odysseus. Finally, Singh mentions Iliad 2.755, in which line we read: ‘for it [the Titaressus] [159] is a branch of the Styx, terrible water of oath’ (ὅρκου γὰρ δεινοῦ Στυγὸς ὕδατός ἐστιν ἀπορρώξ). We can add to this the recurring formulary of Iliad 15.37–38, Odyssey 5.185–186, and Homeric Hymn to Apollo 85–86, in which the water of the Styx is equated to ‘both the greatest and most terrible oath for the blessed gods’ (ὅς τε μέγιστος | ὅρκος δεινότατός τε πέλει μακάρεσσι θεοῖσι). Benveniste is again pertinent as he emphasizes that the oath of the gods—this most terrible oath—is not a divine speech act but instead an act that finds expression in the very waters of the Styx that are poured, “being a material invested with baneful powers”: [160] the material which is the terrible god equates to a terrible act of cult speech. [161] We can remind ourselves that Hesiod, Theogony 776, calls the goddess of these waters deinē Stuks (δεινὴ Στύξ) ‘terrible Styx’, and in the line preceding names her stugerē theos athanatoisi (στυγερὴ θεὸς ἀθανάτοισι) ‘hateful to the immortals’, playing stugerē (στυγερή) ‘hateful’ and deinē Stuks (δεινὴ Στύξ) phonically and semantically—and in fact etymologically—off of one another, both stugerē (στυγερή) and the goddess’ name being derived from the same primitive etymon, which is also source of the -no- adjective stugnos (στυγνός), to which we now turn.

§31. Occupying similar semantic territory as deinos (δεινός), but collectively far less commonly attested than either hagnos (ἁγνός), semnos (σεμνός), or deinos (δεινός) individually, are three other -no- derivatives: (1) stugnos (στυγνός) ‘hated, horrible’; (2) smerdnos (σμερδνός) ‘terrible (to perceive)’; and (3) phriknos (φρικνός) ‘dreadful, terrible’. This last is a hapax legomenon (Hesychius Φ 886). Of the remaining two, stugnos (στυγνός) appears earliest in Archilochus fr. 171.1 and occurs frequently in tragedy, as in, for example Aeschylus Persians 286 and 976, used in Persian characterizations of Athens, and 472, of the deity that caused the Persian destruction. [162] Of common origin with the name of the river Styx ‘Hateful’– called deinē Styx (δεινὴ Στύξ) by Hesiod (Theogony 776), as we have just seen [163] —stugnos (στυγνός) is conventionally linked to the Proto-Indo-European etymon *(s)teu- ‘to move forcefully’, a root showing various consonant extensions among its wide-ranging reflexes—a *-g- in the instance of stugnos (στυγνός). [164] Compare the Sanskrit nasal-infix form tuñjati ‘to strike’. The Sanskrit noun tuja– denotes ‘thunderbolt’; in Rig Veda 6.26.4 an obscure figure called “Tuji the ritual enunciator” is said to have been aided by Indra (see also Rig Veda 10.49.4). An -áya- form of the verb, tujayant-, appears in Rig Veda 7.104.7, used of Indra and Soma driving off demons. The second of these -no- formants, Greek smerdnos (σμερδνός) ‘terrible (to perceive)’, is comparatively rare; more common is the by-form smerdaleos (σμερδαλέος). [165] As we have already noted, Homer uses smerdnos (σμερδνός) to describe the head of the Gorgon upon the aegis (Iliad 5.741–742)—here in conjunction with deinos (δεινός)—and twice, formulaically, of the battle roaring of Ajax (Iliad 15.687, 732). Hesychius (Σ 1232) glosses smerdnos (σμερδνός) as deinos (δεινός). This adjective smerdnos (σμερδνός) has been traced to the Indo-European etymon *smerd- ‘to pain’, [166] a variant of the root *merd- ‘to scrape, rub away’, source of Sanskrit mr̥dnāti ‘to crush; destroy’: the intensive marmartti occurs in the Rig Veda, in hymn 2.23.6, as Br̥haspati, here virtually identical to Indra, is invoked imperatively to employ his ‘seizing calamity’ (duchunā harasvatī) against one that would set a snare for the poet. [167]

§32. Standing in approximate semantic opposition to the preceding set is the -no-adjective terpnos (τερπνός) ‘pleasing, pleasurable’. We find it early in Semonides fr. 7.53, of the absence of anything pleasing among weasels; Mimnermus fr. 1.1, of the pleasures of Aphrodite; Tyrtaeus fr. 12.38, of enjoying life’s pleasures before descending into Hades’ realm; Sappho fr. 160.1 (L-P), of pleasurable songs; Theognis 1.256, of the pleasure of gaining what one loves. It is frequent in Pindar. With terpnos (τερπνός) compare Sanskrit tr̥ptá- ‘satisfied’, as in Rig Veda 7.38.8, either of chariot horses or the divine warrior Maruts. [168] Compare also the compound asu–tr̥p- ‘enjoying (i.e. taking) another’s spirit (asu-)’, as in Rig Veda 10.14.12, of the hounds of Yama (god of the dead) and 10.87.14, of the demonic element—or in the sense ‘enjoying life’, as in Rig Veda 10.82.7, of certain hymn-chanting priests who live too well and gorge on offerings. [169] The Greek and Sanskrit forms are descended from Indo-European *terp- ‘to take pleasure’, having reflexes in Germanic and Balto-Slavic as well. [170]

§33. A further example of the Greek -no- formant is provided by steg(a)nos (στεγ[α]νός) [171] ‘enclosing’, from Proto-Indo-European *(s)teg- ‘to cover’. [172] The Greek adjective (steganos [στεγανός]) is first attested in Aeschylus Agamemnon 358 in a prayer to Zeus and Nyx, used in describing the net the gods had cast over Troy. Compare the Sanskrit past participle sthagita- ‘hidden’, used of the goddess Sarasvatī in a metaphorical expression referring to failed verbal expression. [173] With this compare particular Greek uses of the finite verb stegō (στέγω) to denote enunciatory suppression, as in Tiresias’ prophetic words at Oedipus Tyrannus 341: ‘These things [i.e. ἔπη ‘words’; l. 340] will come to pass, even if I myself hide them in silence’ (ἥξει γὰρ αὐτά, κἂν ἐγὼ σιγῇ στέγω). [174] Hesychius (Σ 1681) glosses neuter steganon (στεγανόν) as signifying ‘to hide words and not to proclaim them’ (στέγειν τοὺς λόγους καὶ μὴ ἐξαγγέλλειν). A reflex of *(s)teg-no- also survives in Latin tignum, [175] archaic term denoting materials used in constructing a building or vineyard (Festus p. 364M) [176] and providing the derivative tigillum ‘beam’. The derivative occurs notably in the rite of the sororium tigillum, referring to the sacred beam beneath which Horatius passed in making expiation for the shedding of his sister’s blood, and used more broadly in putting away the warrior’s combat fury following battle. [177] Both the particular and the general case entail attempted suppression of powerful potentialities, polluting and destructive, in a context of cult operations. Germanic and Celtic reflexes of *(s)teg- are also attested.

§34. Two of the three most frequently attested Geek verbal adjectives in -no-, semnos (σεμνός) ‘revered, holy’ and hagnos (ἁγνός) ‘sacred, holy’, belong centrally to the realm of worship. [178] The (by far) most frequently attested, deinos (δεινός), is also routinely employed in a way that parallels the use of semnos (σεμνός) and hagnos (ἁγνός) as language descriptive of the gods, modifying attributively the names of divine beings. It also participates in the recurring and long-lived syntagm deinos/deinē theos (δεινὸς / δεινὴ θεός) ‘terrible god(dess)’. We have, moreover, seen good reasons for identifying deinos (δεινός) as central to the lexicon of cult. Stugnos (στυγνός) ‘hated, horrible’ and smerdnos (σμερδνός) ‘terrible (to perceive)’ function synonymously, at times in conjunction with deinos (δεινός), and display linguistic ancestry that is at home in the realm of religious speech. Much the same can be said of steg(a)nos (στεγ[α]νός) ‘enclosing’.

§35. Preservation of archaic -no- morphology in the instance of these several forms must surely be another expression of the tendency of early Indo-European languages to cling to the ancestral lexicon of religion and cult speech with particular tenacity, as observed by Vendryes a century ago (“un nombre assez considérable de mots qui se rapportent à la religion et notamment à la liturgie du culte, au sacrifice” [1918:266]). Vendryes is principally concerned with the languages of the eastern and western edges of the Indo-European expansion area (Italic, Celtic, and Indo-Iranian) and the primitive priestly classes that characterize the societies of those regions, though in at least one instance (p. 270) he brings a parallel from that investigative domain to bear on a problem of Greek, as he argues for the etymological commonness of hieros (ἱερός) ‘holy᾽ and hieros (ἱερός) ‘lively᾽ (Sanskrit iṣira-), contra Boisacq, comparing Old Irish nóeb ‘holy’ (*noib-o-) and niab, Welsh nwyf, ‘excitation’ (*neib-o-). A related historical and social linguistic dynamic must be operative in the relic preservation of primitive Graeco-Indo-Iranian and Proto-Indo-European *-no- lexical morphology in Greek, one in which ancestral priestly formulae and enunciations provide a model for efficacious religious linguistic structures, even if inherited Indo-European priestly structures have been modified among the Greeks (at least among post-Mycenaean Greeks).

§36. If Linear B po-re-no is to be rightly understood as phoreno- (φορενο-), cognate with Sanskrit bharaṇa-, a form providing a trace preservation of the ancestral formant *-eno-, the cause of that preservation must similarly lie in the use of the term in sacred phrasing. No less than po-re-si, po-re-no- must belong to the Mycenaean lexicon of cult—as it self-evidently does, to the extent that the compound po-re-no-zo-te-ri-ja has been rightly understood to be the name of a religious festival. Moreover, many students of Linear B would identify yet an additional example of phoreno- (φορενο-) being used in a parallel way.

§37. The brief and broken inscription of Pylos tablet Ua 1413 (from a series containing the state-banquet documents) inventories in its first line a consignment of cloth: 7 units of *146 cloth and 1 unit of *166+WE, with a break following. The second line begins ro-u-si-jo, a-ko-ro ‘field of Lousos’, referencing the environs of one of the major cities in the vicinity of the Pylos palace. This locational descriptor is followed by a single and incomplete form, po-re-no-tu-ṭẹ[. The form has been aggressively and “almost universally restored” [179] to read po-re-no-tu-ṭẹ[-ri-ja, with the second element of the form, *tu-te-ri-ja, understood as θυ(σ)τήρια, [180] denoting an element of offering; and, thus, for those who interpret po-re-na as victims, the restored po-re-no-tu-ṭẹ[-ri-ja signals the ‘sacrifice of victims’. [181] The form po-re-no-tu- ṭẹ[ has, as with po-re-no-zo-te-ri-ja, been interpreted as the name of a festival. [182] Palaima (1999:455), urging caution, notes that according to Aristarchus (p. 455n62) “in Homer θύειν [thuein] is used of offering and burning but never of slaughtering victims in sacrifice (σφάξαι [sphaksai]).” Palaima also calls attention to Plato’s Euthyphro 14C, in which Socrates is made to say that the act of thuein (θύειν) is that of making a gift to the gods. We might note that Plato’s Socrates contrasts this act with that of eukhesthai [εὔχεσθαι] ‘to pray’, which is asking something from the gods. If in fact po-re-no-tu-ṭẹ[ were properly restored as po-re-no-tu-ṭẹ[-ri-ja we would likely see a reference to an offering made in conjunction with phoreno- (φορενο-) in the sense ‘childbearing’. Such a ritual offering, whether undertaken in order to promote conception and ensure healthy fetal development and safe childbirth or in thanksgiving for these, belongs to the same sphere of activity as the presentation of votive vestments of pregnancy to Artemis that we considered above in our discussion of po-re-no-zo-te-ri-ja. In fact, the action described by the hypothesized form *po-re-no-tu-te-ri-ja could itself entail the offering of such textile items. Is the po-re-no-tu-ṭẹ[-ri-ja a festival at which such vestments were offered? As we observed above, various votive objects offered to such ends are well attested in post-Mycenaean Greece, as are prayers of thanksgiving for aid in birth. [183] In the documents of Mycenaean Greece the birth goddess Eileithyia is mentioned four times, and in three of these instances (Knossos tablets Od 714, 715, and 716) she is mentioned in conjunction with a consignment of wool.

§38. In conclusion, the interpretation of Linear B po-re-na as a Mycenaean athematic infinitive phorēnai (φορῆναι) is consistent not only with an Indo-European syntagmatic pattern and a linguistic feature of the closely related Arcadian dialect, but is also consistent with the analysis of po-re-si as a dative plural, specifically phor-en-si (φορ-εν-σι), the dative plural of the athematic participle of phorēnai (φορῆναι), of a type attested in both Arcadian and Aeolic. The interpretation of one form informs that of the other. Po-re-no-, in contrast, preserves a primitive Indo-European morphology. Read as a nominal phoreno- (φορενο-), cognate with Sanskrit bharaṇa-, the Mycenaean form is likely inherited from the liturgical language of Proto-Graeco-Indo-Iranian tradition, if not from that of an earlier period.