6. The Name of Achilles: Questions of Etymology and “Folk-Etymology”*

6§1 In his book on the language of the Linear B tablets, Leonard R. Palmer explained the etymology of the name of Achilles, Ἀχιλ(λ)εύς, as a shortened variant of a compound formation *Akhí-lāu̯os, built from the roots of ἄχος ‘grief’ and of λαός ‘host of fighting men, folk’, morphologically parallel to such “Caland” compounds as Homeric κυδι-άνειρα and Οἰδι-πόδης. [1] The posited morphological shortening from *Akhílāu̯os to Ἀχιλ(λ)εύς, with optional doubling of the last consonant in the shortened variant, is paralleled by such forms as Χαρί-λαος and Χάριλλος (cf. also Φιλεύς vs. Φιλλεύς). [2] What follows is a reassessment of Palmer’s explanation.

6§2 In my own work on the name of Achilles, I agreed with Palmer’s reconstruction of *Akhílāu̯os, offering further evidence on the two distinct {131|132} levels of linguistics and poetics. [3] The linguistic evidence was primarily morphological, with a few additions to the examples already adduced by Palmer. [4] The poetic evidence came mainly from the formulaic system attested in the Dichtersprache of the Homeric Iliad and Odyssey. [5]

6§3 First of all, we may note that the noun ἄχος ‘grief’ is a functional synonym of πένθος ‘grief’ in the Homeric Dichtersprache: for example, the personal grief of Achilles over Briseis is ἄχος at Iliad I 188, XVI 52, 55 and πένθος at Iliad I 362; his grief over Patroklos is ἄχος at Iliad XVIII 22, XXI 47 and πένθος at XVIII 73; likewise, the collective grief of the Achaeans is ἄχος at Iliad XVI.22 and πένθος at IX 3. [6] This thematic parallelism between ἄχος and πένθος is pertinent, I argued, to the morphological parallelism between Palmer’s reconstructed “Caland” compounds *Akhí-lāu̯os and *Penthí-lāu̯os, matching respectively the shortened “Caland” forms Ἀχιλ(λ)εύς and Πένθιλος. [7] Second, I argued at length that the poetic evidence of the Homeric Dichtersprache reveals “a pervasive nexus” between ἄχος and Ἀχιλ(λ)εύς, which is “integrated in the inherited formulaic system and hence deeply rooted in the epic tradition.” [8]

6§4 This statement is quoted, with approval, by Gary Holland, who then goes on to summarize my overall interpretation of the Iliad along the lines of this etymology:

6§5 Despite his agreement on the level of poetics, Holland has two objections on the level of linguistics. First, he suggests that the thematic nexus between {132|133} ἄχος and Ἀχιλ(λ)εύς may be a matter of “folk-etymology,” not etymology: “the preponderance of ἄχος and its derivatives may simply be due to a folk-etymological association of the word with the name of Achilles on the part of the epic poet(s), and not to an actual etymological connection” (highlighting mine). [10] Second, he suggests that my translation of the “Caland” compound *Akhí-lāu̯os, ‘whose λαός has ἄχος’, “seems wrong for this compound type,” because “dependent noun compounds are used very infrequently as the basis for bahuvrīhi or possessive adjective compounds.” [11]

6§6 It is easier to begin with the second objection, if I am right in thinking that it is based on a misunderstanding. All along, I interpreted the reconstructed “Caland” compound *Akhí-lāu̯os as ‘whose host of fighting men is sorrowful [= grieving]’, where the syntactical function of the first component is indeed that of a possessive adjective. [12] Intending to convey a diathetical neutrality in the adjectival component, [13] which I am here rendering as ‘sorrowful [= grieving]’, I devised the translation ‘whose lāu̯ós [λαός] has ákhos [ἄχος = sorrow, grief]’. [14] Similar translations can be applied to other “Caland” compounds, as with κυδι-άνειρα ‘whose men are κυδροί’, that is, ‘whose men have κῦδος’; also, Οἰδι-πόδης ‘whose feet are swollen’, that is, ‘whose feet have swelling = οἶδος’ (in this case, the “Caland” simplex with suffix -ρός, alternate of the compound formant οἰδι-, is not attested).

6§7 Holland’s second objection raises a more important question, which is central to this presentation: how to distinguish an etymology from a “folk-etymology.” The latter term is misleading, I suggest, if it leads to the assumption that the only “genuine” etymology in comparative linguistics is one where a given reconstructed form can be traced all the way back to the parent language of the given languages being compared. According to such an assumption, a reconstruction like *Akhí-lāu̯os would be a “false” etymology if it cannot be traced back to “proto-Indo-European.”

6§8 The term “folk-etymology” implies another, even more misleading, assumption: that any etymologically “wrong” derivation of one given form from another is purely a synchronic phenomenon. True, a functioning or living connection between a given set of forms that had once been unconnected must be assumed to have a starting point at some given synchrony. Still, any synchrony is destined to become, moving forward in time, simply a cross-section in the diachrony of language. As we reconstruct {133|134} a given language forward in time, what may count as a “wrong” connection in an earlier cross-section can become a “right” connection in a later cross-section, from the standpoint of the evolving structure of that language. Here I refer to the classic work of Emile Benveniste on the necessity of combining synchronic with diachronic methods in the establishment of etymologies. [15]

6§9 In the case of a form like Ἀχιλ(λ)εύς, the question is not whether it had always been connected with the forms ἄχος and λαός. What matters instead is whether this connection is “deeply rooted,” as I have described it, in the formulaic system of Homeric Dichtersprache and whether it can be traced far back enough in time to reach the remote stage when “Caland” formations were still a productive mechanism in the Greek language.

6§10 Moving diachronically forward, by the time we reach even the earliest attestations of the Greek language, we find that the “Caland” mechanism is already residual, clearly no longer productive: only such vestiges as κυδι-άνειρα vs. κυδρός are left. [16] What remains productive, however, as I argued, is the actual Dichtersprache that had preserved “Caland” formations like *Akhí-lāu̯os vs. Ἀχιλ(λ)εύς and *Penthí-lāu̯os vs. Πένθιλος.

6§11 Such a Dichtersprache, however, can be considered a system in its own right, capable of generating, analogically, such non-“Caland” formations as Χαρί-λαος vs. Χάριλλος, Σθενέ-λαος vs. Σθένελος, Νείλεως (Ionic, from *Nehélāu̯os, apparently attested in the Linear B tablets as ne-e-ra-wo) vs. Νηλεύς (non-Ionic, from *Neheleús), Ἰόλαος vs. Ἰόλη and Ἰόλεια (implying a corresponding *Ἰολεύς), Περίλαος vs. Πέριλλος. [17] Still other non-“Caland” types that could have been generated by the Dichtersprache along the lines of *Akhí-lāu̯os and *Penthí-lāu̯os include Πρωτεσί-λαος (Iliad II 698, etc.), Χαιρεσί-λαος, Πενθεσί-λεια. [18]

6§12 With reference to Πενθεσί-λεια, Holland remarks: “Although πένθος means ‘pain’ synchronically in Greek, further connections within Indo-European are semantically difficult.” [19] I draw attention to his use here of “synchronically,” since his purpose is to argue that seemingly related forms, such as πενθερός ‘relative by marriage’, are to be derived from the Common Greek root *penth- ‘bind’ (as in πεῖσμα ‘rope’; the Indo-European root is *bhendh-, as in Sanskrit bandh-), so that Πενθεσί-λεια should mean ‘binding the λαός’ rather than ‘paining the λαός’. [20] {134|135}

6§13 The problem is, Holland’s use here of “synchronically” implies that there is just one level of synchrony for the meaning of ‘grief’ or ‘pain’—as if any previous level would default diachronically to the meaning of ‘bind’. And yet, the possibility of reconstructing earlier levels of synchronicity for πένθος in the sense of ‘pain’ becomes open-ended if the root is derived from Common Greek *ku̯enth- ‘suffer’ (cf. Lithuanian kenčiù, Irish cēssaim), as opposed to Common Greek *penth- ‘bind’. [21]

6§14 It would be preferable in this case, I suggest, to keep in mind not the diachrony of the root πένθ- but also the synchronicity of a Dichtersprache that could generate, along with a morphological and thematic parallelism of ἄχος vs. πένθος, a morphological and thematic parallelism of *Akh(es)í-lāu̯os vs. *Penth(es)í-lāu̯i̯a. These parallelisms converge in the epic tradition of a mortal combat between the male warrior Ἀχιλ(λ)εύς and the female warrior Πενθεσί-λεια, as reflected in the Aithiopis (Proclus summary p. 105.22 Allen).

6§15 Up to now, I have offered linguistic arguments in support of Palmer’s explanation of Ἀχιλ(λ)εύς. In terms of my argumentation, however, Palmer’s explanation “will not carry conviction unless we can show that the meaning of *Akhí-lāu̯os is intrinsic to the function of Achilles in myth and epic.” [22] In a later work, Palmer himself quoted and gave his approval to this formulation. [23] After he quotes my formulation about *Akhí-lāu̯os, Palmer goes on to say:

6§16 To restate my original formulation: “the ἄχος of Achilles leads to the μῆνις of Achilles leads to the ἄχος of the Achaeans.” [25] As I also argued, the {135|136} ἄχος experienced by warriors in the epic Dichtersprache is formulaically the converse of κράτος: that is, the λαός or ‘host of fighting men’ is conventionally described as having κράτος when they win, ἄχος when they lose. [26] It is crucial to note in this context Benveniste’s demonstration that the semantics of κράτος are driven by a “zero-sum” mentality: the very fact that one of two sides gets κράτος necessitates that this side is thereby the winner and the other side, the loser. [27] Moreover, the thematic polarity of κράτος / ἄχος is mirrored by the morphological parallelism of Ἀχαιός / κραταιός, embedded in the formulaic system of the Homeric Dichtersprache, and the very name of the λαός, that is, the Ἀχαιοί, is synchronically derived from ἄχος—at least, within the framework of this Dichtersprache. [28]

6§17 How, then, could it happen that the naming of this host of fighting men was driven by a negative concept, as encoded in the word ἄχος? My answer centered on both the ritual and the mythological aspects of warfare, as viewed within the epic tradition. [29] Palmer asks a similar question about the naming of a hero like Achilles: it can only happen, he answers, if the very idea of *Akhí-lāu̯os, ‘whose λαός has ἄχος’ had been generated by the themes of myth. [30]

6§18 And yet the name of Achilles is “attractively identified,” as Palmer puts it, in the Linear B tablets: in the text of Pylos tablet Fn 70.2, a list of names in the dative includes a-ki-re-we, to be read as Akhil(l)ēwei. [31] As I commented on this attestation, “we must be ready to assume that the mythopoeic name of Ἀχιλ(λ)εύς inspired the naming of historical figures called Ἀχιλ(λ)εύς.” [32] Palmer comments on my comment: “In fact, it is at the very least unlikely that any parent would have bestowed such a name on his son unless its {136|137} inauspicious overtones had been masked by its occurrence as a heroic name in a famous story.” [33] If Palmer’s “chain of reasoning,” as he calls it, is correct, “then the Pylian record may be construed as implying that a version of the ‘Wrath of Akhilleus’ was current at the time of the destruction of Pylos.” [34]

6§19 All this is not to rule out an etymological connection, proposed by Holland, between the intermediate reconstructed Greek form *Ἀχιλος and “proto-Germanic” *Agilaz, from which the Old Norse name Egill can be derived. [35] Still, even though Holland allows for the possibility of an earlier reconstructed Greek form *Akhí-lāu̯os, the acceptance of a Germanic cognate *Agilaz leaves us with morphological as well as semantic problems that are unresolved. [36] In another connection, Palmer once called attention to “the first rule of etymology,” attributed to Franz Skutsch: “Look for Latin etymologies first on the Tiber.” [37] That “rule” is applicable to the name of Achilles. {137|138}