Introduction
Among the evidence for the regular outcome of *rÌ¥ in Ionic-Attic, the root of καÏÏεÏá½¹Ï, κÏαÏá½»Ï, κÏá½±ÏÎ¿Ï and related forms is of crucial importance. Several formations have doublets, the most prominent ones being καÏÏεÏá½¹Ï ~ κÏαÏεÏá½¹Ï, κÏá½±ÏÎ¿Ï ~ κάÏÏοÏ, and κÏá½±ÏιÏÏÎ¿Ï ~ κάÏÏιÏÏοÏ. In chapter 1, we have seen that the forms with âαÏâ are troublesome for accounts arguing for a regular development *rÌ¥ > âÏαâ in Proto-Ionic, in view of the full grade attested in Aeol. κÏá½³ÏÎ¿Ï and Ion. κÏá½³ÏÏÏν. Indeed, by a process familiar from chapter 4, forms like κÏá½±ÏÎ¿Ï have secondarily introduced an analogical zero grade κÏαÏâ from a related adjective, replacing the original full grade κÏεÏâ. This is a very important argument for viewing âαÏâ as the regular reflex.
However, the analysis of doublets like καÏÏεÏá½¹Ï ~ κÏαÏεÏá½¹Ï is complicated by two issues that must be clarified before we draw this conclusion. One approach to such doublets has been to claim that κÏαÏâ and καÏÏâ were freely interchangeable allomorphs.1 This is in my view an ad hoc strategy designed to save the idea of a regular development *rÌ¥ > âÏαâ. For one thing, it does not explain why doublets are attested only for a subset of the attested formations. For instance, the adjective κÏαÏÎ±Î¹á½¹Ï has no by-form *καÏÏαιόÏ, in spite of the fact that using κÏαÏÎ±Î¹á½¹Ï in hexametrical Greek necessitated the use of muta cum liquida scansion, an uncommon phenomenon in Homer. Moreover, some variants appear to be limited to specific genres: for example, κÏαÏεÏá½¹Ï never occurs in prose.
Apparently, then, doubles with καÏÏâ beside κÏαÏâ could be created only under certain conditions. One main goal of this chapter is to show, by a fresh etymological analysis of the evidence, in which respective formations the root shapes κÏαÏâ and καÏÏâ originated and by which mechanisms they spread. This analysis is reinforced by a close consideration of synchronic derivational relations: once we take the lexical meanings of base form and derivative into account, we may infer with more plausibility that certain formations were analogically or even artificially created in the epic language. For instance, we will see that the epic form κάÏÏÎ¿Ï is only used as an adjectival abstract meaning âstrength, violenceâ (i.e. the fact of being καÏÏεÏá½¹Ï ~ κÏαÏεÏá½¹Ï), and that it never means âpowerâ or âsuperiorityâ, the default meanings of κÏá½±ÏοÏ. This strengthens the idea that κάÏÏÎ¿Ï was derived from καÏÏεÏá½¹Ï within Epic Greek.
A second question is: how certain are we about the original shape of the full grade? Two different etymologies have been proposed. The mainstream view holds that κÏá½±ÏÎ¿Ï is related to the Indo-Iranian noun meaning âwill-powerâ (Ved. krátuâ m.); an alternative proposal (Lamberterie 1990: 346â353) compares Goth. hardus âhardâ and derives the Greek and Germanic words from PIE *kertâ âcutâ. The second proposal is problematic in view of the root shape of forms like Aeol. κÏá½³ÏοÏ, but Benveniste (1969: II, 71â83) suggested that the Greek group may be a conflation of both roots, and thus have a dual etymological origin. Benveniste based this suggestion on the semantics of καÏÏεÏá½¹Ï (and its doublet κÏαÏεÏá½¹Ï), the adjective that seems to have served as the derivational basis of most other attested formations.
If Benvenisteâs thesis were correct, it could be assumed that the vocalization of *rÌ¥ (and hence the emergence of two root allomorphs καÏÏâ and κÏαÏâ) took place at a time when these roots were still lexically distinct. Thus, if we wish to utilize forms like καÏÏεÏá½¹Ï as evidence for the regular reflex of *rÌ¥, we must be able to exclude a dual etymological origin as proposed by Benveniste, or at least to render this idea unlikely. For this purpose, a careful study of the semantics of the attested forms will be necessary.
5.1 Semantics and Etymology
In this section, I will first review the existing etymologies and their problems, and explore the semantics of καÏÏεÏá½¹Ï ~ κÏαÏεÏá½¹Ï in more depth. The analogical spread of κÏαÏâ and καÏÏâ across the attested formations will be studied in section 5.2.
5.1.1 The Competing Etymologies
There is no generally accepted etymology for κÏαÏεÏá½¹Ï and related forms. Since the early days of Indo-European studies, the Greek lexical family has been compared to two different formations.2 On the one hand, the epic adjective κÏαÏá½»Ï has been equated with Goth. hardus âhardâ and its Germanic cognates, the root of which is PIE *kert-. On the other hand, it has been compared to an isolated Indo-Iranian masculine noun, Ved. krátuâ m. âwill-power, resolveâ and Av. xratu- âid.â, a u-stem to a different PIE root of the shape *kretâ.3
The connection with Goth. hardus is advocated by Lamberterie (1990: 323â350). On the basis of an extensive discussion of the Greek attestations and their semantics, he claims that the basic meaning of κÏαÏεÏá½¹Ï and κÏαÏá½»Ï can be reconstructed as âhard, firm, solidâ.4 He proposes (1990: 349) to derive the Greek and Germanic forms from the verbal root PIE *kert- âto cutâ.5 The semantic development would lead from âcuttingâ to âsharpâ and then, independently in Greek and Germanic, to âhardâ.6 However, there are serious problems with this etymology. For one thing, the root of the Germanic adjective (quasi *kortúâ) has a different full grade slot compared to Aeol. κÏá½³ÏοÏ, Ion. κÏá½³ÏÏÏν.7 Moreover, as I will argue below, it is unlikely that âhard, solid, firmâ is the basic meaning of the Greek group.
The second etymology, a comparison of the adjective κÏαÏá½»Ï with Ved. krátuâ, has been advocated by Strunk (1975). In the framework of internal derivation, this proposal is nowadays accepted without hesitation.8 It is to be noted, however, that κÏαÏá½»Ï is attested only in the epic formula κÏαÏá½ºÏ á¼ÏγεÏÏόνÏηÏ, the meaning of which cannot be determined with certainty (the widely-used translation âstrongâ may well be off the mark). Strunk (1975: 269â270) argued that κÏαÏá½»Ï in this formula must have referred to a characteristic trait of Hermes, such as cleverness, quick wits, or dexterity, and he claims that an older meaning of PIE *krétuâ was âmagical powerâ. However, even if we suppose that magic played a role in Indo-Iranian *krátuâ, this is rather implausible for the Greek concept of κÏá½±ÏοÏ.9
There is also a morphological problem: the side-by-side existence of an adjective (κÏαÏá½»Ï) and a derivationally isolated noun (Ved. krátuâ) requires an explanation. Strunk envisaged two possible ways to connect them. On the one hand, he considers the possibility that the Indo-Iranian noun was originally an adjective. He deems this unlikely, because Ved. krátuâ is inflected according to a different accent and ablaut paradigm (gen. kratváḥ, ins. kratvÄÌ) compared to the adjectives in âá½»Ï (gen. âέοÏ). Strunk therefore suggested that κÏαÏá½»Ï may have originally been an abstract noun. Assuming that the formula κÏαÏá½ºÏ á¼ÏγεÏÏόνÏÎ·Ï is comparable to cases like βίη ἩÏακληείη âthe Heracleid forceâ (for âHeraclesâ) and that it had the meaning âthe Argos-killing powerâ, he suggests that κÏαÏá½»Ï could be reinterpreted as an adjective in such instances. However, since we do not really know the underlying synchronic meaning of κÏαÏá½ºÏ á¼ÏγεÏÏόνÏηÏ, this is mere speculation at best. Moreover, it would be highly problematic for the entire âCaland systemâ of κÏαÏεÏá½¹Ï, the most extensive of its kind in Greek, to be based on one single form which is itself seen as the product of a reanalysis.
Nowadays, the similarity between κÏαÏá½»Ï and Ved. krátuâ is often accounted for within the framework of internal derivation, under the assumption that Ved. krátuâ reflects a so-called acrostatic paradigm, *krót-uâ / *krét-u-.10 However, κÏαÏá½»Ï beside Ved. krátuâ is in reality one of the very few examples for the supposed derivational scheme. Moreover, the semantic connection between these two items, though possible, is not as clear-cut as some scholars make us believe. It therefore seems best to suspend judgment on this point.
In Van Beek 2013: 151â155, I have proposed a new etymology according to which καÏÏεÏá½¹Ï ~ κÏαÏεÏá½¹Ï âpowerful; firmâ is related to Vedic Årathi âto loosenâ, Åithirá- âlooseâ. Semantically, the idea is that in the Rigveda, Åithiráâ still means âlooseâ in the sense âagile, moving freelyâ. While Åithiráâ may refer to the unbridled motion of horses and the Maruts, κÏαÏεÏá½¹Ï in Homer often denotes the powerful motion of warriors, animals or weapons, and hence their capacity to have impact or to deal damage. Thus, âloose, unrestrained, unbridledâ would have developed into âfierce, violentâ and hence âpowerful, firmâ (see below for further discussion of the semantic developments in Greek). In terms of phonology and morphology, the etymology is attractive because the Greek and Vedic adjectives can both directly derive from the same PIE pre-form *ḱrÌ¥th1-ró-.11 In this way, we may account not only for the suffix âÏá½¹Ï (instead of âá½»Ï) but also for its extended form âεÏá½¹Ï: as we will see, *krÌ¥teróâ is reflected in three different dialect groups.
Although I still consider this etymology to be at least as attractive as the connection with Vedic krátuâ, I will not insist on its correctness here, as the present argument does not depend on it. Instead, since our main task is to establish that the only full grade of the root was PGr. *kretâ, let us turn to the arguments advanced by Benveniste (1969). Benveniste based his etymological analysis on the polysemy of various lexemes in Homer. Since he was unable to reconcile the different meanings of κÏαÏεÏá½¹Ï ~ καÏÏεÏá½¹Ï under one overarching concept, he concluded that this adjective is a conflation of two etyma, the one meaning âhardâ and related to Goth. hardus, the other meaning âsuperiority, prevalenceâ and related to Ved. krátuâ.12 According to Benveniste, the original difference between these groups is preserved faithfully in the semantics of most Greek formations. For instance, he claims that the abstract κÏá½±ÏÎ¿Ï means only âsuperiority, prevalenceâ (and that it is related to Vedic krátuâ), while the adjective κÏαÏÎ±Î¹á½¹Ï would mean only âhard, harsh, cruelâ (and is supposedly related to Goth. hardus). The two sets of meanings are supposed to coexist only in κÏαÏεÏá½¹Ï.13
Various objections must be made to this analysis (the details will be elaborated in the sections that follow). First of all, it remains unclear why the original semantic difference would have been preserved only in καÏÏεÏá½¹Ï ~ κÏαÏεÏá½¹Ï, and why, for instance, only one of the roots formed an abstract. Given the system of derivatives attested in Greek, the default assumption must be that we are dealing with just one etymon, and the burden of proof is on Benveniste. Moreover, several of the semantic claims made by Benveniste appear to be incorrect or incomplete once we consider them more closely:
-
κÏá½±ÏÎ¿Ï ~ κάÏÏÎ¿Ï does not only mean âpower, superiorityâ (as claimed by Benveniste 1969, II: 77), but it may also mean âfierceness, violenceâ or âenduranceâ (see section 5.2.6).
-
κÏá½±ÏÎ¿Ï ~ κάÏÏÎ¿Ï refers not only to human warriors (Benveniste, o.c. 78), but also to animals (Od. 3.370, the endurance of horses) and iron (Od. 9.393).
-
it is unlikely that κÏαÏÎ±Î¹á½¹Ï means âhard, cruelâ when occurring as an epithet of sons in the formula Ï á¼·Îµ κÏαÏαιώ âstrong sonsâ, or when Odysseus applies the term to himself (Od. 18.383).
Finally, it is difficult to account for the assumption that two originally different etyma independently made an adjective in âεÏá½¹Ï.
As we have seen, the proposed etymological connections with Indo-Iranian and Germanic both have their own morphological and semantic issues. There is also widespread disagreement on the meaning of the Homeric formations. For instance, κÏá½±ÏÎ¿Ï is translated as âstrength, forceâ by some scholars, but as âsuperiority, prevalenceâ by others;14 again others stress that κÏá½±ÏÎ¿Ï seems to refer to the hardness of iron in one Homeric passage.15 Indeed, given this wide range of meanings, Benvenisteâs proposal of a dual etymological origin may seem less surprising. We therefore have to ask how the various meanings of all derivatives can be derived from a single, more basic meaning. In sections 5.1.2â3, I will first deal with this question for the adjective καÏÏεÏá½¹Ï ~ κÏαÏεÏá½¹Ï in Homer. After that, it will be possible (in section 5.2) to combine semantics and derivational morphology in our analysis of the root allomorphy in other formations like κÏá½±ÏÎ¿Ï ~ κάÏÏοÏ.
5.1.2 The Semantics of καÏÏεÏá½¹Ï ~ κÏαÏεÏá½¹Ï in Epic Greek
Classical scholars have traditionally posited âstrongâ as the basic meaning: see e.g. LSJ (s.v. κÏαÏεÏá½¹Ï): âstrong, stout, mighty, in Hom. mostly of bodily strength (â¦)â.16 This rendering is not incorrect, but it needs further qualification. First of all, considering the Homeric evidence, there is every reason to think of more specific translations than âstrongâ. As a general qualification of warriors or monsters, κÏαÏεÏá½¹Ï means âfierce, violentâ, and when qualifying concrete motion or applied strength, it means âmighty, impetuous, overwhelmingâ (cf. also κÏαÏá½³Ï âto be rampantâ and á¼ÏικÏαÏá½³ÏÏ âimpetuouslyâ). While κÏαÏεÏá½¹Ï occasionally qualifies âbodily strengthâ, to use the words of LSJ, it more often denotes a propensity towards violence or an applied force. In addition, it means âsolid, firmâ when qualifying e.g. shields, bonds, or oaths.17 Thus, although the translation âstrongâ may serve as a common denominator, it is often inexact and not sufficiently specific.
In my view, two basic sets of meanings must be distinguished for Homeric κÏαÏεÏá½¹Ï ~ καÏÏεÏá½¹Ï:
-
âfierce, mighty, vehement, violentâ (e.g. of warriors, arrows, winds);
-
âsteadfast; enduring, firm, solidâ (e.g. of warriors, chains, oaths, shields).
First, these lexical meanings will be illustrated with examples.18 After that, I will argue that 1. âfierceâ is the oldest meaning of the root that we can reconstruct within Greek, and then show how 2. âsteadfast; firmâ secondarily developed. Those interested only in the last-mentioned point may flip over to section 5.1.3.
Sense 1. âfierce, mightyâ is frequently attested as a qualification of warriors (or violent mythical beings).19 I will limit myself here to a brief discussion of two telling instances. In the first instance, the Trojan Helenus refers to Diomedes, who is at the summit of his aristeia at this point, as:
Il. 6.97â101á¼Î³Ïιον αἰÏμηÏὴν κÏαÏεÏὸν μήÏÏÏÏα Ïόβοιο,á½Î½ δὴ á¼Î³á½¼ κάÏÏιÏÏον á¼Ïαιῶν Ïημι γενέÏθαι.οá½Î´â á¼Ïιλá¿á½± Ïοθâ ὧδέ γâ á¼Î´Îµá½·Î´Î¹Î¼ÎµÎ½ á½ÏÏαμον á¼Î½Î´Ïῶν,ὠν Ïá½³Ï ÏαÏι Î¸Îµá¾¶Ï á¼Î¾á½³Î¼Î¼ÎµÎ½Î±Î¹Î á¼Î»Î»â ὠδε λίηνμαίνεÏαι, οá½Î´á½³ Ïá½·Ï Î¿á¼± δύναÏαι Î¼á½³Î½Î¿Ï á¼°ÏοÏαÏίζειν
(that) savage spearman, a fierce causer of rout who, in my view, is definitely the mightiest of the Achaeans. Not even Achilles did we ever fear in such a way, that leader of men, who, they say, is born of a goddess; no, this man rages excessively, and no one is able to vie with him in might.
Diomedes is called κάÏÏιÏÏÎ¿Ï á¼Ïαιῶν even in comparison with Achilles because of the routs which he causes at this very moment. In the Iliad, only Hector (once) and Diomedes (twice) are called κÏαÏεÏὸν μήÏÏÏÏα Ïόβοιο âfierce deviser of routâ.20 These routs are ascribed, here as elsewhere, to a Î¼á½³Î½Î¿Ï âdriveâ which is so vehement (λίην μαίνεÏαι) that no one present is able to vie with Diomedes in this respect.21 He can no longer be contained by mere human effort.
Likewise, if Achilles is repeatedly called κÏαÏεÏá½¹Ï, this is primarily because of his fierce fighting spirit. When, during his own aristeia, he fights the river god Scamander, the latter calls his brother Simoeis for help:
Il. 21.314â315ἵνα ÏαύÏομεν á¼Î³Ïιον á¼Î½Î´Ïαá½Ï δὴ νῦν κÏαÏέει, μέμονεν δâ ὠγε ἶÏα θεοá¿Ïι
so that we may stop the wild man who is now rampant; this manâs rage is equal to that of the gods
Some hundred lines earlier, Scamander has directly addressed Achilles as follows:
Il. 21.214â215ὦ á¼Ïιλεῦ, ÏεÏὶ μὲν κÏαÏέειÏ, ÏεÏὶ δâ αἴÏÏ Î»Î± ῥέζειÏá¼Î½Î´Ïῶν
of all men you are the most violent, and you do the most abominable deeds
The violent deeds of Achilles are characterized in exactly the same terms as the berserk battle rage of Diomedes: both are rampant and know of no restraint.22 Note, in particular, the verbal correspondences á¼Î³Ïιον á¼Î½Î´Ïα ~ á¼Î³Ïιον αἰÏμηÏήν and μαίνεÏαι / Î¼á½³Î½Î¿Ï ~ μέμονεν. In the case of Achilles, this almost elemental force is explicitly related to his divine descent. Only the forces of nature can contain his fierce might, here described with the verb κÏαÏέει.23
The examples adduced here can easily be multiplied.24 They show that κÏαÏεÏá½¹Ï and related words are consistently applied to warriors that are fierce or mighty, either as a general characteristic, or at a specific moment. Moreover, κÏαÏεÏá½¹Ï is not only applied to human warriors, but also, more generally, to fierce or violent mythical beings. Polyphemus, whose κÏá½±ÏÎ¿Ï is said to be greatest among the Cyclopes (Od. 1.70), is characterized by his unrestrained use of violence and ferocity when he (or his force) is called κÏαÏεÏá½¹Ï.25 In Hesiod, κÏαÏεÏá½¹Ï qualifies the Giants (Th. 50), the Erinyes (Th. 185), Cerberus (Th. 312), the Chimaera (Th. 320, cf. 322), the Hundred-Arms (Th. 670), and the feet of Typhoeus (Th. 824). For all these monstrous creatures, the translation âfierce, violentâ is more pertinent than âstrongâ.
Not only fierce warriors are called κÏαÏεÏá½¹Ï. In Homer, the compound καÏÏεÏá½¹Î¸Ï Î¼Î¿Ï âwith impetuous spiritâ qualifies Diomedes, Achilles, and Heracles, but in Hesiod (Th. 378) it appears as an epithet of the three winds. The frequent phrase κÏαÏεÏὸν Î¼á½³Î½Î¿Ï qualifies not only the fierce battle spirit that Poseidon puts into the Aiantes (Il. 13.60), but also the destructive elemental force of the fire that consumes a corpse (ÏÏ Ïá½¸Ï ÎºÏαÏεÏὸν Î¼á½³Î½Î¿Ï Î±á¼°Î¸Î¿Î¼á½³Î½Î¿Î¹Î¿, Od. 11.220). Arrows, spears, and thrown stones receive the epithet κÏαÏεÏá½¹Ï on various occasions, e.g. κÏαÏεÏὸν Î²á½³Î»Î¿Ï âpowerful missileâ (Il. 5.104), βολάÏν Ïε κÏαÏεÏá½±Ïν (Th. 683, thrown by the Hundred-Arms and Titans). In such instances, κÏαÏεÏá½¹Ï denotes the momentum and destructive impact of the missiles.26 It is conceivable that this use of κÏαÏεÏá½¹Ï originated in its application to heavenly missiles, notably the thunderbolt, the weapon by means of which Zeus ensures his supremacy (κÏá½±ÏοÏ).27
In another set of instances, κÏαÏεÏá½¹Ï âvehement, violentâ qualifies a powerful movement, drive or emotion. Very frequent in formulae is κÏαÏεÏá½´ á½Ïμίνη, where the epithet refers to the ardor or fervor of battle. The same sense can be recognized in the phrases κÏαÏεÏá½¸Ï ÏÏá½¹Î¼Î¿Ï âvehement tremblingâ (Il. 6.137), κÏαÏεÏá½´ λύÏÏα âvehement rageâ (Il. 9.239, 21.543), κÏαÏεÏὸν Î´á½³Î¿Ï âvehement fearâ (Od. 14.88), κÏαÏεÏὸν Î¼á½³Î½Î¿Ï âfierce driveâ (passim) and κÏαÏεÏâ á¼Î»Î³ÎµÎ± âvehement sufferingâ (passim). Finally, when κÏαÏεÏá½¹Ï describes the physical qualities of a warrior or his body parts, it underlines the extreme nature of an applied force or violence,28 e.g. κÏαÏεÏá¿Ïι βίηÏι(ν) âwith mighty violenceâ (Il. 21.501, Od. 9.476, 12.210), cf. also κάÏÏÎµÏ ÏειÏῶν âvigor of the armsâ (Il. 8.226 and 11.9).29
In its second sense, κÏαÏεÏá½¹Ï ~ καÏÏεÏá½¹Ï means âsteadfast, enduring, toughâ or âfirm, solidâ. In the following examples it refers to the stamina or resilience of warriors: á¼ÏÏάμεναι κÏαÏεÏá¿¶Ï âto stand oneâs ground firmlyâ (Il. 11.410, 13.56, 15.666), á¼Î»Î»â á¼Ïεο κÏαÏεÏῶÏ, á½ÏÏÏ Î½Îµ δὲ λαὸν á¼ ÏανÏα âbut hold on firmly, and encourage all your menâ (Il. 16.501 and 17.559), á¼Î¼ÏίβαÏιν κÏαÏεÏήν âtough defenseâ (Il. 5.623), Ïá½±Î»Î±Î³Î³ÎµÏ ÎºÎ±ÏÏεÏαί âsteadfast phalanxesâ (Il. 5.591â592, 13.126â127), κÏαÏεÏá½°Ï â¦ Ïá½±Î»Î±Î³Î³Î±Ï (Il. 13.90) and κÏαÏεÏαὶ ÏÏá½·ÏÎµÏ á¼ÏÏιÏÏá½±Ïν (Il. 4.90, 201) âthe steadfast ranks of the shield-bearersâ. This meaning is shared by Homeric Greek and the Classical language.30 Benveniste and Lamberterie have argued that phalanxes were called κÏαÏεÏαί because they were âhard, massive, solidâ like a wooden log. However, I agree with Strunk (1975: 270â275) that κÏαÏεÏá½°Ï â¦ Ïá½±Î»Î±Î³Î³Î±Ï must refer to the firm spirit of the warriors that form a phalanx:
Das homerische á¼ÎºÎ±ÏÏύνανÏο Ïá½±Î»Î±Î³Î³Î±Ï meint ein festigen der Schlachtreihen auch oder gerade unter Wiederherstellung des Abwehrâ und Angriffsgeistes oder âwillens. Dieser Kampfgeist heisst á¼Î»Îºá½µ (â¦). Bezeichnenderweise wird in einer an Agamemnon gerichteten Schmahrede des Diomedes die á¼Î»Îºá½µ als âdas grösste κÏá½±ÏοÏâ bezeichnet. Die von Benveniste geleugnete semantische Brücke zwischen κÏá½±ÏÎ¿Ï und καÏÏύνεÏθαι is damit zumindest in der homerischen Sprache (â¦) greifbar: mit á¼Î»Îºá½µ, âkämpferischer Gesinnungâ, haben beide Wörter zu Tun.
Strunk 1975: 273â274
The sense âfirm, tenaciousâ can also be recognized in the following cases: κÏαÏεÏá¿· á¼Î½á½¶ δεÏμῷ âin a firm bondâ (Il. 5.386),31 ÏεÏÏὶ Ïίεζε νÏλεμέÏÏ ÎºÏαÏεÏá¿Ïι âhe tenaciously checked (his mouth) with his firm handsâ (Od. 4.287â288, Odysseus seals the mouth of Anticlus as they lie in ambush in the Trojan horse), κÏαÏεÏá½´ δâ á¼Ïεν á¼²Ï á½Î´Ï Ïá¿Î¿Ï (Il. 23.719â720, Odysseus puts a check on Menelaus during their wrestling match), and καÏÏεÏὸν á½ Ïκον âfirm oathâ. In several cases we are dealing with a force that continues to be applied, and this meaning may well be more original than âhard, solidâ as a qualification of materials.
In fact, examples where κÏαÏεÏá½¹Ï qualifies a material as solid or hard do occur in Homeric Greek, but they are rare: á¼Î½ÎµÎ³Î½á½±Î¼Ïθη δέ οἱ αἰÏμὴ / á¼ÏÏίδâ á¼Î½á½¶ κÏαÏεÏá¿ âthe point of his spear was bent away in the tough shieldâ (Il. 3.349 = 17.45) and θώÏÎ·ÎºÎµÏ ÎºÏαÏαιγύαλοι âharnesses with hard breastplatesâ (Il. 19.361).32 This sense remains current after Homer,33 but various alleged instances in early Greek epic are ambiguous at best. For example, ῥαιÏÏá¿Ïα κÏαÏεÏήν (Il. 18.477) is not a âhard hammerâ, but rather the âpowerful hammerâ of the smith Hephaestus (note that ῥαιÏÏá½µÏ is an old agent noun to á¿¥Î±á½·Ï âto smash to piecesâ). Similarly, in ÏιδήÏÎ¿Ï â¦ ÎºÏá½±ÏÎ¿Ï (Od. 9.393) and ÏίδηÏÎ¿Ï á½ ÏÎµÏ ÎºÏαÏεÏá½½ÏαÏá½¹Ï á¼ÏÏιν (Th. 864), it is not certain that the hardness of iron is intended, rather than its fierce or violent qualities in action. Finally, the traditional epithet κÏαÏεÏá¿¶Î½Ï Î¾ (of horses and mules) is normally understood as âwith hard hoofsâ,34 but âwith fierce hoofsâ, referring to the violent kick of a horseâs leg, would be at least as appropriate.35
A special case is the verse-end κÏαÏαίÏεδον οá½Î´Î±Ï (Od. 23.46). This is normally translated as âground with hard soilâ, assuming that the second member is Ïέδον âgroundâ, but that assumption is not obvious. In my view, it would be more attractive to assume that the second member reflects Ïέδη âshackle, hobbleâ, so that κÏαÏαίÏεδον means âwhich has firm shacklesâ, to be compared with the phrase κÏαÏεÏá¿· á¼Î½á½¶ δεÏμῷ. Moreover, Hades is regularly called the âfirm gatekeeperâ (ÏÏ Î»á½±ÏÏαο κÏαÏεÏοá¿Î¿) of the Netherworld, Achilles refers to the Earth as restraining even men who are κÏαÏεÏá½¹Ï (á¼¥ Ïε καÏá½° κÏαÏεÏόν ÏÎµÏ á¼Ïύκει, Il. 21.63); and Moira is said to âshackleâ warriors when they are killed (verse-final Îοá¿Ïα ÏέδηÏε, passim). In the episode containing κÏαÏαίÏεδον, the suitors have just been killed by Odysseus: âthey are lying all around him, the one over the other, occupying the κÏαÏαίÏεδον οá½Î´Î±Ïâ. It would make excellent sense if the poet referred to the earth as having fetters that will restrain the suitors forever.
5.1.3 Reconstruction of the Semantic Developments
In the view of Lamberterie (1990), âhard, solidâ was the basic meaning of κÏαÏá½»Ï before this form went out of use. One piece of evidence in support is the post-Homeric factitive verb κÏαÏá½»Î½Ï âto harden, solidify, confirmâ.36 He also claims that the original meaning of κÏαÏεÏá½¹Ï is âhard, firm, solidâ, and that this adjective is a replacement of κÏαÏá½»Ï. Finally, he shows that κÏαÏÎ±Î¹á½¹Ï is semantically equivalent to κÏαÏεÏá½¹Ï, and analyzes it as reflecting the old feminine of κÏαÏá½»Ï (1990: 337â343). This allows him to derive all meanings of κÏαÏεÏá½¹Ï ~ κÏαÏá½»Ï directly from the root meaning of PIE *kertâ âto cutâ: âcutting, sharpâ would have developed to âsevereâ and âviolentâ on the one hand, and to âhard, solidâ on the other.
Lamberterieâs assessment of the different meanings and their relations is intriguing, but as remarked above, the wrong vowel slot of *kertâ âto cutâ with respect to Aeol. κÏá½³ÏÎ¿Ï and Ion. κÏá½³ÏÏÏν remains an insurmountable objection to the etymology he defends. As a matter of fact, the meanings âenduring, steadfastâ (of animate beings) and âfirm, solidâ (of inanimate entities) may well be secondary with respect to âfierce, mightyâ, as the Homeric material allows us to retrace possible intermediate stages. For instance, the traditional phrase κÏαÏεÏὸν Î¼á½³Î½Î¿Ï can be translated as âfierce energyâ when it refers to the destructive ardor of fire that consumes a corpse (ÏÏ Ïá½¸Ï ÎºÏαÏεÏὸν Î¼á½³Î½Î¿Ï Î±á¼°Î¸Î¿Î¼á½³Î½Î¿Î¹Î¿, Od. 11.220). Fire is a prototypical example of a fervent energy that cannot be stopped once it has been released. On the other hand, κÏαÏεÏὸν Î¼á½³Î½Î¿Ï can also qualify an arduous stamina, as in the following passage. Menelaus and Meriones carry the corpse of Patroclus towards the ships as they are protected by the Aiantes from the assault of the Trojans. They are compared to a pair of mules that draw heavy wooden logs from the mountains:
Il. 17.742â746οἳ δâ á½¥Ï Î¸â ἡμίονοι κÏαÏεÏὸν Î¼á½³Î½Î¿Ï á¼Î¼ÏιβαλόνÏεÏá¼Î»ÎºÏÏâ á¼Î¾ á½ÏÎµÎ¿Ï ÎºÎ±Ïá½° ÏαιÏαλόεÏÏαν á¼ÏαÏÏὸνἢ δοκὸν á¼ á½² δόÏÏ Î¼á½³Î³Î± νήÏονΠá¼Î½ δέ Ïε Î¸Ï Î¼á½¸ÏÏείÏεθâ á½Î¼Î¿á¿¦ καμάÏῳ Ïε καὶ ἱδÏá¿· ÏÏÎµÏ Î´á½¹Î½ÏεÏÏινÎá½£Ï Î¿á¼µ γâ á¼Î¼Î¼ÎµÎ¼Î±á¿¶Ïε Î½á½³ÎºÏ Î½ Ïá½³Ïον.
Like mules that have put on enduring spirit drag forth from a mountain down a rugged path a beam or a large piece of ship-timber; as they struggle, their spirit is distressed by toil and sweat alike: in a similar way the two struggled to bear away the corpse.
In this passage, we are no longer dealing with the fierce energy of a warrior, but with the untiring pull of draught animals. In the same way, fierce attackers and stout defenders can be qualified as κÏαÏεÏá½¹Ï. Common to both uses is the arduous nature of the energy and its unrestrained operation. Examples of this meaning âenduringâ also occur with derived formations. Athena tells Nestor to provide his guest Telemachus with horses that are á¼Î»Î±ÏÏá½¹ÏαÏοι θείειν καὶ κάÏÏÎ¿Ï á¼ÏιÏÏοι (Od. 3.370), âswiftest in running and best in staminaâ.37 A beautiful testimony for κÏαÏÎ±Î¹á½¹Ï in the meaning âenduring, toughâ is Od. 18.383 (see section 5.2.11).
Concerning the possible pathways of semantic development, it is not difficult for âenduring, steadfastâ (as a qualification of animate beings) to develop into âfirm, tenaciousâ (of an applied corporeal force) and âdurable, lastingâ (of inanimate entities such as bonds). The latter meaning may then have developed to âsolid, hardâ (e.g. of an impenetrable shield).
A clear parallel for these semantic developments is found in Lat. dÅ«rus. There can be no question that the original meaning of the adjective PIE *duh2-róâ was âlong-lasting, enduringâ, as reflected in Gr. δηÏόν âfor a long timeâ, Arm. erkar âlong-lastingâ.38 In Latin, the denominative verb dÅ«rÅ, âÄre means âto last, persevere; endureâ (thus mostly in Plautus), which may well reflect the older root meaning.39 As for dÅ«rus, many of its lexical meanings are similar to those of κÏαÏεÏá½¹Ï. For instance, it may mean âtough, hardy, vigorousâ in a military sense,40 and âharsh, roughâ as a qualification of human behavior or words (cf. Hom. κÏαÏεÏá¿¶Ï á¼ÏοειÏεá¿Î½). And, of course, dÅ«rus also qualifies hard materials, but not yet in Plautus. Thus, this example shows that a âphysicalâ sense like âsolid, hardâ is not necessarily the more original meaning.
Concerning the two etymologies traditionally proposed for the group of κÏαÏεÏá½¹Ï, it appears that the lexical meaning shared with Goth. hardus âhardâ may have come into being by a series of secondary semantic developments. Given that the root of the Germanic word-group has a different full grade slot, the comparison must be rejected. A connection with Ved. krátuâ and Av. xratuâ âdetermination, resolveâ remains conceivable, but it requires that Indo-Iranian underwent the same development from âfierce, mightyâ to âenduring, steadfastâ as in καÏÏεÏá½¹Ï. I have proposed an alternative etymology deriving κÏαÏεÏá½¹Ï âfierce, mightyâ and Ved. Åithiráâ âlooseâ from the same pre-form PIE *ḱrÌ¥th1-róâ. This is less straightforward semantically, but it has the advantage that the suffix âεÏá½¹Ï (instead of âÏá½¹Ï) can be easily accounted for. In any case, even if the comparison with Ved. krátuâ is correct (which is not certain), it is unlikely that the complete system of âCalandâ derivations in Greek arose from the adjective κÏαÏá½»Ï: the suffix âεÏá½¹Ï is hard to account for in this scenario, and as we will see, both καÏÏεÏá½¹Ï and κÏαÏá½»Ï are old within Greek.
5.2 The Allomorphy of κÏαÏâ and καÏÏâ in Homer and Classical Greek
In dealing with the origins of the root allomorphs κÏαÏâ and καÏÏâ, we may now start from one single etymological origin: a Proto-Greek root *kretâ meaning âfierce, mightyâ. Before analyzing the derivational relations between different formations and the genre distinctions between the different allomorphs, let us consider the dialectal evidence in more detail.
5.2.1 Dialectal Reflexes and Proto-Greek Reconstruction
Evidence from Aeolic and Arcado-Cyprian shows without a doubt that the original root shape was *kretâ. The Lesbian poet Alcaeus uses the neuter κÏá½³ÏοÏ, a present á¼ÏικÏá½³Ïει, and an aorist κÏá½³ÏηÏαι. The same vocalism is preserved in onomastic material: Cypr. ti-mo-ke-re-te-se /TÄ«mokretÄs/ (ICS2 361, 5thâ4th c. BCE), Arc. ÎαικÏεÏÎ·Ï (Dubois 1988, I: 111â112), Lesb. ÎαμικÏεÏÎ·Ï (Hodot 1974: 116). The original full grade is also preserved in the Ionic-Attic comparative (Hom. κÏείÏÏÏν, Ion. κÏá½³ÏÏÏν, Att. κÏείÏÏÏν). Lesbian personal names such as ÎκεÏÏÎ·Ï have been adduced as evidence for an alternative full grade κεÏÏâ, but as Hodot (1974) has shown, these forms are not only of late attestation, but probably have nothing to do with âκÏεÏÎ·Ï (see section 5.2.9 below).
In Central Cretan, the positive is reflected as καÏÏεÏÎ¿Ï âhaving right of sayâ (over property, Lex Gortyn), âfirm, trustworthyâ (of a witness, IC IV 63.4, Gortyn, late 6th or early 5th c.). The comparative κάÏÏÏν âbetterâ in literary Doric reflects a pre-form *krÌ¥t-i̯osâ, with a zero grade root introduced from the positive.41 The same form is found as καÏÏÏν âmore trustworthyâ (as a witness) in the Lex Gortyn.42 In fact, as far as our evidence goes, Central Cretan has generalized the root shape καÏÏâ in all derived forms: καÏÏαιÏοδα âcattleâ, καÏÏÎ¿Ï âviolenceâ (Lex Gortyn), personal names in âκαÏÏηÏ, and a stative verb καÏÏεÏ.43 Since the comparative form (which is shared by other West Greek dialects) points in the same direction, it looks as if most of West Greek generalized the zero grade root of PGr. *krÌ¥teróâ in all derivations.44 Another dialectal reflex of PGr. *krÌ¥teróâ is attested as κοÏÏεÏάΠκÏαÏεÏá½±, á¼°ÏÏÏ Ïá½± (Hsch.).45 This form is judged by LSJ to be âprob. Aeolicâ, but since the regular vocalization of *rÌ¥ in the Aeolic dialects was âÏοâ (see section 3.3), it is more likely that κοÏÏεÏá½± was taken from Arcadian or even Cyprian; the Aeolic outcome of *krÌ¥teróâ is unattested.46
Thus, genuine dialectal reflexes of the adjective *krÌ¥teróâ are found for Ionic-Attic, West Greek, and probably Arcado-Cyprian. This strongly suggests that *krÌ¥teróâ was inherited from Proto-Greek. An additional argument for a high antiquity of *krÌ¥teróâ is the semantic divergence between Cretan καÏÏεÏÎ¿Ï and Ionic-Attic καÏÏεÏá½¹Ï: the former is used to qualify witnesses as âfirmâ or owners of property as âhaving right of sayâ. In Attic, the normal way to say the same thing is κύÏιοÏ, a lexical archaism. Moreover, in Cretan καÏÏÏν functions as the comparative of καÏÏεÏοÏ, which is not the case for Classical κÏείÏÏÏν (or even Homeric κÏείÏÏÏν, as we will see below). On the other hand, the neuter καÏÏÎ¿Ï means âviolenceâ in Cretan; this meaning is also attested for Ionic-Attic κÏá½±ÏοÏ, but diverges from the meaning of καÏÏεÏÎ¿Ï in Cretan. Thus, the Cretan word group cannot have been borrowed wholesale from Ionic-Attic, but the dual sense of âviolence, mightâ beside âfirm, authoritativeâ, attested for both Ionic-Attic and Cretan, may well have been inherited from Proto-Greek.
Three important conclusions can be drawn. First, the full grade form of the root was certainly *kretâ in Proto-Greek, and reflexes of this root are found in three Greek dialect groups. Secondly, Proto-Greek had an adjective *krÌ¥teróâ, the meaning of which probably at least included âsteadfast, firmâ and âfierce, violentâ. Since κÏαÏá½»Ï cannot have been secondarily created within Greek, Proto-Greek must have had two adjectives *krÌ¥túâ and *krÌ¥teróâ, with forms of comparison *kréti̯osâ and *krétistoâ.47 Thirdly, there is a dichotomy between dynamic meanings (e.g. âviolent, vehement, mightyâ) and state-oriented meanings (e.g. âpower, controlâ), which may well hark back to Proto-Greek.
5.2.2 Adjectives in âá½»Ï, âÏá½¹Ï and âεÏá½¹Ï
The root vocalism of the adjectives in âá½»Ï has been discussed in chapter 4. As for the adjectives in âÏá½¹Ï, they derive from a thematic formation whose root was normally in the zero grade, cf.:48
-
μακÏá½¹Ï âtall, longâ, Lat. macer âmeagreâ < *mh2ḱ-róâ beside μήκιÏÏÎ¿Ï âlongestâ;
-
á¼ÏÏ Î¸Ïá½¹Ï âredâ, Lat. ruber âid.â < *h1rudh-róâ beside á¼ÏÎµá½»Î¸Ï âto redden, make redâ.49
A zero grade root is also found in PGr. *krÌ¥teróâ âfierce; firmâ and its reflexes. However, why do we find two different adjectives, καÏÏεÏá½¹Ï ~ κÏαÏεÏá½¹Ï and κÏαÏá½»Ï? Clearly, κÏαÏá½»Ï is a relic form, as adjectives in âá½»Ï are unproductive in Greek generally. Another example where Greek may preserve two inherited adjectival formations to the same root is á¼Î»Î±ÏÏá½¹Ï ânimbleâ (cf. OHG lungar âfast, cheerfulâ) beside á¼Î»Î±Ïá½»Ï âsmallâ (cf. Ved. raghú- âfastâ). We might therefore leave it at this and assume that both καÏÏεÏá½¹Ï ~ κÏαÏεÏá½¹Ï and κÏαÏá½»Ï were inherited from PIE. However, this does not yet explain why the former has a suffix âεÏá½¹Ï rather than simply âÏá½¹Ï. It is therefore reasonable to ask how PGr. *krÌ¥teróâ (or its predecessor in PIE) may have been created.
A fair number of *ro-formations in Greek and Indo-Iranian are inherited from PIE, but in addition deverbal ro-adjectives were productive. In Indo-Iranian, many instances occur beside state-oriented verbs or verbs of motion (Rau 2009: 163, with tables on pp. 164â167). In Greek too, âthe vast majority of ro-stems are made to roots that make primary verbsâ (Rau 2009: 168), and again, they pair either with activity verbs or state-oriented verbs. In my view, this implies that the PIE âCaland suffixâ *âroâ was deverbal tout court.50 This may explain why different Indo-European languages sometimes used different adjectival suffixes for the same root: Hitt. tÄpu- âsmall, littleâ beside Ved. dabhrá- âid.â and Hitt. daÅ¡Å¡u- âstrong, etc.â beside Ved. dasrá- âartful; skilledâ, OAv. daÅra- âid.â. Since the verbal root is still attested in Indo-Iranian, it is attractive to assume that the Anatolian u-stem adjectives preserve something older and that the forms underlying Ved. dabhrá- and dasrá- are deverbal innovations of Nuclear PIE or Proto-Indo-Iranian.51 Likewise, the meaning of á¼Î»Î±ÏÏá½¹Ï ânimbleâ is relatively close to that of the verbal root *h1lengwhâ (which is not attested in Greek, but cf. Ved. ráá¹hate âto move without effortâ) while á¼Î»Î±Ïá½»Ï âsmall, littleâ is completely lexicalized.
All this is not meant to exclude that some adjectives in *âroâ are very old, but merely to sketch a possible scenario for the rise of PGr. *krÌ¥teróâ. If the etymological connection with Ved. Åithirá- âloose, relaxedâ and the verbal root Årathi âto loosenâ (proposed in Van Beek 2013) is correct, we may posit an adjective PIE *ḱrÌ¥th1-ró- âloose, unbridledâ, whence in Greek âfierce, mighty, overwhelmingâ. The root-final laryngeal would immediately account for the extended shape of the suffix âεÏá½¹Ï in καÏÏεÏá½¹Ï (PGr. *krÌ¥teróâ), which remains unexplained in other scenarios. There must have been a semantic split between *ḱrÌ¥th1-róâ and *ḱrÌ¥th1-úâ, but it is hard to recover the details as κÏαÏá½»Ï occurs only in a single formula. If the meaning of κÏαÏá½»Ï was âsolid, hard; firmâ (as proposed by Lamberterie 1990: 327â331), this would make sense within the above scenario: the adjective in âá½»Ï would be more lexicalized, while the meaning of καÏÏεÏá½¹Ï âenduring, perseveringâ would be closer to that of the verb κÏαÏá½³Ï.
5.2.3 Synchronic Description of the Classical Prose Forms
Before analyzing the variation κÏαÏâ ~ καÏÏâ in Epic Greek, let us first consider the situation in Classical times, for clear distributions between both root variants can be found there. Table 7 shows all word-forms with κÏαÏâ and καÏÏâ attested in Classical Ionic-Attic. Forms that are exclusively attested in poetry are marked as such, but forms exclusively attested in hexameter poetry are not included. Unless otherwise indicated, prose forms are attested both in Attic and in Ionic.
Table 7
κÏαÏâ versus καÏÏâ in Classical Ionic and Attic
|
Forms with κÏαÏâ |
Forms with καÏÏâ |
|---|---|
|
κÏαÏεÏá½¹Ï (poetic only) |
καÏÏεÏá½¹Ï |
|
κÏαÏÎ±Î¹á½¹Ï (poetic only) |
|
|
κÏαÏαιâ (poetic only)52 |
καÏÏαιâ (poetic only)53 |
|
ÎÏαÏιâ (in PNâ¯s only) |
ÎαÏÏιâ (in PNâ¯s only)54 |
|
κÏá½±ÏÎ¿Ï |
κάÏÏÎ¿Ï (poetic only)55 |
|
compounds in âκÏαÏá½µÏ |
|
|
κÏαÏá½³Ï |
|
|
κÏαÏá½»Î½Ï |
καÏÏá½»Î½Ï (poetic only)56 |
|
κÏá½³ÏÏÏν, κÏείÏÏÏν (âÏÏâ) |
|
|
κÏá½±ÏιÏÏÎ¿Ï |
κάÏÏιÏÏÎ¿Ï (poetic only)57 |
|
κάÏÏα |
The forms κάÏÏοÏ, καÏÏαιâ, καÏÏύνÏ, and κάÏÏιÏÏÎ¿Ï are attested only once or twice each, and always in poetry; they did not belong to the Ionic-Attic vernacular. Moreover, καÏÏεÏá½¹Ï is the only regular prose form; κÏαÏεÏá½¹Ï is only found occasionally, but always in poetry.58 Thus, Classical prose has the following forms:
-
adjective καÏÏεÏá½¹Ï âfierce; firm, solid; persevering, steadfastâ
-
comparative κÏείÏÏÏν âbetter, prevailing, strongerâ
-
superlative κÏá½±ÏιÏÏÎ¿Ï âmost powerful, supreme; bestâ
-
adverb κάÏÏα âvery, heavily; surelyâ
-
neuter κÏá½±ÏÎ¿Ï âpower, controlâ
-
compounded adjectives:
á¼Î³ÎºÏαÏá½µÏ âtenacious; in controlâ
á¼ÎºÏαÏá½µÏ âpowerlessâ
-
denominative verb κÏαÏá½³Ï âto prevail, conquerâ
-
factitive verb κÏαÏá½»Î½Ï âto make solid, fortify; be in chargeâ
The single most striking fact about the forms attested in prose is that there are no doublets with âαÏâ beside âÏαâ. Moreover, a number of forms have been lexicalized and are no longer derivationally associated with their etymological relatives. First of all, the comparative κÏείÏÏÏν and the superlative κÏá½±ÏιÏÏÎ¿Ï no longer belong with the adjective καÏÏεÏá½¹Ï. This appears not only from the respective lexical meanings, but also from the creation of new forms of comparison καÏÏεÏá½½ÏεÏοÏ, καÏÏεÏá½½ÏαÏοÏ. Instead, it is commonly believed that Attic κÏείÏÏÏν and κÏá½±ÏιÏÏÎ¿Ï are the comparative and superlative belonging with á¼Î³Î±Î¸á½¹Ï âgoodâ, in the specialized sense âstrongâ.59 This can be contrasted with the situation in Cretan (see 5.2.1) and in Homer (see below). Secondly, there is no synchronic derivational relation between καÏÏεÏá½¹Ï and the abstract κÏá½±ÏοÏ. Again, this is not only shown by the distinct lexical meanings of both groups and their different root shapes καÏÏâ versus κÏαÏâ,60 but also by the existence of a denominative verb καÏÏεÏá½³Ï âto persevereâ and a derived abstract καÏÏεÏία âperseveranceâ. Finally, the adverb and particle κάÏÏα and the factitive verb κÏαÏá½»Î½Ï (in the sense âto hardenâ attested in prose) are synchronically isolated.
We are now in a position to draw an important conclusion: the classical Ionic-Attic forms καÏÏεÏá½¹Ï and κάÏÏα must be the regular reflexes of their respective pre-forms with *krÌ¥tâ because their vowel slot is aberrant with respect to the full grade *kretâ. In all non-poetic forms containing κÏαÏâ (κÏá½±ÏÎ¿Ï and its derivatives âκÏαÏá½µÏ, κÏαÏá½³Ï; superlative κÏá½±ÏιÏÏοÏ; κÏαÏύνÏ), this root variant can be explained as replacing an older form with κÏεÏâ (cf. Lesbian á¼ÏικÏá½³Ïει, κÏá½³ÏηÏαι, and personal names in âκÏá½³ÏηÏ) or as having an analogically influenced vocalization.
The question remains where the root allomorph κÏαÏâ originated. The most natural form to have influenced κÏá½±ÏοÏ, κÏαÏύνÏ, and κÏá½±ÏιÏÏÎ¿Ï would be the positive of the adjective. This positive cannot have been κÏαÏεÏá½¹Ï, however, because the only form (and regular outcome) in Ionic-Attic prose was, as we have just seen, καÏÏεÏá½¹Ï. I see only one possible scenario: the allomorph κÏαÏâ originated in the adjective κÏαÏá½»Ï, and was thence secondarily introduced, at an early date, in other âCalandâ derivations. This introduction did not take place in καÏÏεÏá½¹Ï and κάÏÏα because these forms had already diverged semantically at the relevant time.
We have already encountered a potential argument for the antiquity of κÏαÏá½»Ï (Lamberterie 1990: 328â330): in its sense âto hardenâ, the verb κÏαÏá½»Î½Ï can be understood as a factitive derived from κÏαÏá½»Ï if the latter had the meaning âhard, solidâ.61 However, while the abstract κÏá½±ÏÎ¿Ï indeed never means âhardnessâ in classical Greek, it could be objected that κÏαÏá½»Î½Ï âto hardenâ may have been derived from κÏá½±ÏÎ¿Ï somewhere between Homer and the late 5th century, as the meaning âhardnessâ may in fact be attested for κÏá½±ÏÎ¿Ï in Homer62 and may have fallen into disuse only later. Having said this, it remains probable that κÏαÏá½»Ï was more frequent in Ionic until not too long before Homer, as this is the only way to account for the analogical spread of κÏαÏâ.
5.2.4 Synchronic Description of the Homeric Forms
Table 8 shows the forms with κÏαÏâ and καÏÏâ attested in Homeric Greek.
Table 8
κÏαÏâ versus καÏÏâ in Homeric Greek
|
Forms with κÏαÏâ, *κÏεÏâ |
Forms with καÏÏâ |
|---|---|
|
κÏαÏá½»Ï63 |
|
|
κÏαÏεÏá½¹Ï |
καÏÏεÏá½¹Ï |
|
κÏαÏεÏá½¹-ÏÏÏν, âá¿¶Î½Ï Î¾ |
καÏÏεÏá½¹-Î¸Ï Î¼Î¿Ï |
|
κÏαÏÎ±Î¹á½¹Ï |
|
|
κÏαÏαί-ÏεδοÏ, âÎ³á½»Î±Î»Î¿Ï |
|
|
κÏá½±ÏÎ¿Ï |
κάÏÏοÏ64 |
|
á¼ÏικÏαÏá½³ÏÏ |
|
|
(á¼Ïιâ)κÏαÏá½³Ï65 |
|
|
καÏÏá½»Î½Ï |
|
|
κÏείÏÏÏν |
|
|
κάÏÏιÏÏÎ¿Ï |
As has been noted in the previous section, a number of the specifically epic forms occur occasionally in other poetic genres. Among the Homeric formations there are only two doublets: κÏαÏεÏá½¹Ï ~ καÏÏεÏá½¹Ï and κÏá½±ÏÎ¿Ï ~ κάÏÏοÏ. Furthermore, in κάÏÏιÏÏÎ¿Ï and καÏÏύνÏ, Homeric Greek has καÏÏâ as opposed to a classical form with κÏαÏâ. How did the doublets come into being, and which variants are older?
In the following sections, I will argue that the forms restricted to epic are not the regular outcome of *krÌ¥tâ in some vernacular, but arose within the traditional language of hexameter poetry. It is well-known that epic poets could apply normal mechanisms of linguistic change (e.g. proportional analogy) in order to adapt word-forms to metrical circumstances.66 Indeed, it appears to be possible to indicate a concrete model and motive for the creation of most of the variant forms just listed. By accounting for the mechanisms by which these variant forms came into being, we may further corroborate the insight that καÏÏâ is the regular Ionic-Attic outcome of *krÌ¥tâ.
Another striking fact is that three different adjectives are attested in Homer: κÏαÏá½»Ï, κÏαÏεÏá½¹Ï (with variant καÏÏεÏá½¹Ï), and κÏαÏαιόÏ. In compounds, too, different stem-forms are used as a first member: κÏαÏεÏοâ, καÏÏεÏοâ, and κÏαÏαιâ. To what extent are these stems metrical variants of each other? Is καÏÏεÏá½¹Ï really lexically equivalent to κÏαÏεÏá½¹Ï, or are there perhaps subtle differences in usage? Does κÏαÏÎ±Î¹á½¹Ï have the same lexical meaning as the former two? In order to answer these questions, we will have to pay close attention to philological details. For purposes of reconstruction, it is also important to know more about the historical origin of the different formations. For instance, what does âÎ±Î¹á½¹Ï in κÏαÏÎ±Î¹á½¹Ï represent?
Let us start with the epic doublets κÏαÏεÏá½¹Ï ~ καÏÏεÏá½¹Ï and κÏá½±ÏÎ¿Ï ~ κάÏÏοÏ. Given that καÏÏεÏá½¹Ï and κÏá½±ÏÎ¿Ï are the regular prose forms, it is reassuring to find that κÏαÏεÏá½¹Ï and κάÏÏÎ¿Ï can be analyzed as analogical creations of the epic language.
5.2.5 καÏÏεÏá½¹Ï and κÏαÏεÏá½¹Ï in Homer
Attempts to establish a clear-cut semantic difference between καÏÏεÏá½¹Ï and κÏαÏεÏá½¹Ï within Epic Greek turn out to be futile. Both forms can be used to qualify the same noun, as in Ïá½±Î»Î±Î³Î³ÎµÏ ÎºÎ±ÏÏεÏαί (Il. 5.591â592, 13.126â127) which is mirrored by κÏαÏεÏá½°Ï â¦ Ïá½±Î»Î±Î³Î³Î±Ï (Il. 13.90, cf. also κÏαÏεÏαὶ ÏÏá½·ÏεÏ, 2â¯Ã Il.).67 Furthermore, the first members of κÏαÏεÏá½¹ÏÏÏν and καÏÏεÏá½¹Î¸Ï Î¼Î¿Ï are clearly metrical variants.
The Homeric numbers and distributions are as follows:
-
κÏαÏεÏá½¹Ï (162 times):68 normal in all case forms;
-
καÏÏεÏá½¹Ï (28 times): mainly âá½¹Ï, âόν (mostly after |B), rarely âοί, âαί, âá½±.
The fact that κÏαÏεÏá½¹Ï is almost six times as frequent as καÏÏεÏá½¹Ï is partly due to its occurrence in a number of frequent formulae and set phrases: κÏαÏεÏá½¸Ï ÎÎ¹Î¿Î¼á½µÎ´Î·Ï (or a metrically equivalent PN), κÏαÏεÏὸν μήÏÏÏÏα Ïόβοιο (3â¯Ã), κÏαÏεÏὸν Î¼á½³Î½Î¿Ï (6â¯Ã), κÏαÏεÏá¿Ïι βίηÏι(ν) (3â¯Ã), á¼Î½á½¶ κÏαÏεÏá¿ á½Ïμίνῠ(11â¯Ã), καÏá½° κÏαÏεÏὴν á½Ïμίνην (8â¯Ã), etc. The form was easy to use due to its metrical structure.69 On the other hand, καÏÏεÏá½¹Ï was awkward for epic poets because it could not be used in a hexameter line in a large number of case forms: the dat. and gen. pl. of all genders and most case forms of the feminine contain a cretic sequence. It was easy to handle only in the neuter plural (cf. καÏÏεÏá½° á¼Ïγα) and before a vowel-initial heavy syllable (cf. καÏÏεÏá½¸Ï á¼Î½á½µÏ)
In view of these facts, one might well ask: why did καÏÏεÏá½¹Ï exist at all? The answer is, as we have seen, that καÏÏεÏá½¹Ï was the form of the spoken language, while κÏαÏεÏá½¹Ï is all but restricted to Epic Greek. It became so frequent because it supplied for impracticable case forms of καÏÏεÏá½¹Ï, and perhaps also of moribund κÏαÏá½»Ï. Thus, the epic evidence confirms the conclusion reached on the basis of the classical forms: καÏÏεÏá½¹Ï regularly reflects Proto-Ionic *krÌ¥teróâ.
It remains to indicate how exactly κÏαÏεÏá½¹Ï came into being. One possibility would be that its root was based on that of κÏαÏá½»Ï before the latter lost its currency. I will further pursue this question in section 8.4.1. On the complicated relation between κÏαÏεÏá½¹Ï and κÏαÏá½»Ï in Epic Greek, see also section 5.2.10.
5.2.6 The Neuter Abstract Nouns (and Derivatives) in Homer
The oldest form of the neuter noun in Ionic-Attic is clearly κÏá½±ÏοÏ, with introduction of the root vowel of κÏαÏá½»Ï (cf. Aeol. κÏá½³ÏοÏ). As we have seen, this is also the only form attested in prose. But why and how was κάÏÏÎ¿Ï created?
Let us start from the assumption (which will have to be nuanced later on) that no semantic difference existed between κÏá½±ÏÎ¿Ï and κάÏÏοÏ, just as with κÏαÏεÏá½¹Ï and καÏÏεÏá½¹Ï. Table 9 contains the numbers for both Homeric variants of the doublet, in all attested case forms. For comparison, I have added the figures for ÏÎ¸á½³Î½Î¿Ï âforceâ, which is semantically close and metrically equivalent to κÏá½±ÏοÏ.70
Table 9
Forms of κÏá½±ÏοÏ, κάÏÏÎ¿Ï and ÏÎ¸á½³Î½Î¿Ï in Homeric Greek
|
κÏá½±ÏÎ¿Ï |
κάÏÏÎ¿Ï |
ÏÎ¸á½³Î½Î¿Ï |
|
|---|---|---|---|
|
nom.-acc. |
κÏá½±ÏÎ¿Ï (28â¯Ã) |
κάÏÏÎ¿Ï (6â¯Ã) |
ÏÎ¸á½³Î½Î¿Ï (21â¯Ã) |
|
26â¯Ã before |B |
no fixed position |
16â¯Ã before |B |
|
|
gen. |
â |
â |
ÏÎ¸á½³Î½ÎµÎ¿Ï (5â¯Ã) |
|
dat. |
κÏá½±ÏÎµÏ (2â¯Ã) |
κάÏÏÎµÏ (7â¯Ã) |
ÏÎ¸á½³Î½ÎµÏ (15â¯Ã) |
Forms of κÏá½±ÏÎ¿Ï ~ κάÏÏÎ¿Ï are only found in the nom.-acc. and dat. sg.71 In the nom.-acc. κÏá½±ÏÎ¿Ï is by far the most frequent form, but in the dat. sg., κάÏÏÎµÏ (5â¯Ã Il., 2â¯Ã Od.) is more frequent than κÏá½±ÏεÏ. This is a natural distribution given that κÏá½±ÏÎµÏ consists of three consecutive short syllables (though cf. á½ Ïε κÏá½±ÏÎµÏ ÏÏοβεβήκá¿, Il. 16.54).72 It is therefore plausible that the dative κάÏÏÎµÏ was introduced, on the model of the already-existing doublet καÏÏεÏá½¹Ï ~ κÏαÏεÏá½¹Ï, to supply for the ill-practicable form κÏá½±ÏεÏ. Subsequently, a new nominative-accusative κάÏÏÎ¿Ï was created.
This scenario is corroborated by the lexical semantics of κάÏÏÎ¿Ï and κÏá½±ÏοÏ. In Homer, κÏá½±ÏÎ¿Ï has two primary meanings.73 The first is âfierceness, overpowering force, preponderance, predominanceâ in concrete violent confrontations. The power in question is typically granted to warriors by Zeus or another divinity (cf. formulaic μέγα κÏá½±ÏÎ¿Ï á¼Î³Î³Ï Î±Î»á½·Î¾Ï and inflected forms).74 The second sense is âsupremacy, power, dominionâ, of a military leader or ruler over his subjects.
On the other hand, κάÏÏÎ¿Ï appears to be used only in the first set of meanings of κÏá½±ÏοÏ, âfierceness, preponderanceâ. In βίῠκαὶ κάÏÏÎµÏ Îµá¼´ÎºÏν âyielding to violence and brute forceâ (Od. 13.143 and 18.139, both about criminal or violent deeds) and κάÏÏÎµÏ ÏειÏῶν â(relying on) the might of their handsâ (Il. 8.226 = 11.9), κάÏÏÎ¿Ï denotes an unrestrained, brute force.75 This meaning is also attested in κÏá½±ÏÎ¿Ï â¦ ÎºÏείνειν âthe power to killâ (Il. 11.192â193 = 207â208), as well as in Od. 1.70 where the κÏá½±ÏÎ¿Ï (physical strength, brute force) of Polyphemus is said to be greatest among the Cyclopes.76 A second nuance is âstrengthâ in the sense of endurance or the ability to persevere. This is attested e.g. in Il. 16.524 for κÏá½±ÏÎ¿Ï (the wounded Lycian hero Glaucus asks Athena for strength), and in e.g. Il. 17.561â562 for the variant κάÏÏÎ¿Ï (Menelaus is oppressed by enemies and asks Athena for the strength to hold on).77 There is no place where κάÏÏÎ¿Ï clearly means âpolitical power, dominionâ.
Crucially, then, the abstract κάÏÏÎ¿Ï could only be used when its meaning corresponded to that of the base form, κÏαÏεÏá½¹Ï ~ καÏÏεÏá½¹Ï. In the meaning âpower, dominionâ, κÏá½±ÏÎ¿Ï was used: cf. also the denominative verbs κÏαÏá½³Ï âto rule; be rampantâ and á¼ÏικÏαÏá½³Ï âto hold swayâ, which have no variant with καÏÏâ.78 Thus, κάÏÏÎ¿Ï was derived from καÏÏεÏá½¹Ï âimpetuous, violent; enduringâ on the model of κÏá½±ÏÎ¿Ï beside κÏαÏεÏá½¹Ï. It is not a true doublet of κÏá½±ÏοÏ.79
Having explained the origin of κάÏÏοÏ, it remains to briefly discuss the other forms derived from κÏá½±ÏοÏ. Homer has only one compound in âκÏαÏá½µÏ, which occurs as an adverb: á¼ÏικÏαÏá½³ÏÏ âimpetuouslyâ.80 A by-form in âκαÏÏá½µÏ* does not exist, perhaps because there was no metrical incentive to create it.81 An adjective *á¼ÏικÏαÏá½µÏ is also presupposed by the denominative stative verb á¼ÏικÏαÏá½³Ï âto have the upper hand; to be master, rule (over)â (7â¯Ã Hom.). Frequent is κÏαÏá½³Ï (13â¯Ã), which in Homer means âto have κÏá½±ÏοÏâ in both senses of the noun: âto be violent/rampantâ or âto dominate, be in control, ruleâ. On three occasions, κÏαÏá½³Ï is modified by the adverb μέγα; it is therefore derivationally linked with κÏá½±ÏÎ¿Ï (cf. μέγα κÏá½±ÏÎ¿Ï 6â¯Ã Hom.).
The aorist of κÏαÏá½³Ï, κÏαÏá¿Ïαι âto obtain victoryâ, is unattested in Homer; this form, usual in Classical Greek, was probably avoided in Epic Greek for metrical reasons.82 In theory, the metrical problem could have been solved by creating an artificial form *καÏÏá¿Ïαι, based on a proportion like κÏá½±ÏÎ¿Ï : κÏαÏá¿Ïαι (both spoken Ionic) = κάÏÏÎ¿Ï : X (Epic Greek).83 One reason for the absence of *καÏÏá¿Ïαι may be that, in the meaning âbe victorious, prevailâ, it would have been metrically equivalent to its synonym νικá¿Ïαι, and therefore of little use. Furthermore, as we have just seen the variant κάÏÏÎ¿Ï did not mean âvictory, predominanceâ but âstrength, brute forceâ, so it may have diverged too much semantically to serve as a base form.
5.2.7 The Forms of Comparison in Homer
Homeric grammars and lexica generally state that κÏείÏÏÏν and κάÏÏιÏÏÎ¿Ï are the forms of comparison corresponding to the positive κÏαÏá½»Ï.84 This true from a historical perspective, but not synchronically. Apart from the fact that the precise meaning of κÏαÏá½»Ï cannot be determined, there are two problems, as we will presently see: in Homer κάÏÏιÏÏÎ¿Ï does not function as the superlative of κÏαÏá½»Ï, and κÏείÏÏÏν is paradigmatically isolated, i.e. it remains without a corresponding positive or superlative.
Let us start with the superlative. As a form, Homeric κάÏÏιÏÏÎ¿Ï is artificial and probably due to the avoidance of the normal Ionic form κÏá½±ÏιÏÏÎ¿Ï for metrical reasons.85 The root allomorphy that existed in the positive καÏÏεÏá½¹Ï ~ κÏαÏεÏá½¹Ï âfierceâ was extended to the superlative. This can be backed up by the semantics. While Classical κÏá½±ÏιÏÏÎ¿Ï means âmost powerful, supremeâ, Homeric κάÏÏιÏÏÎ¿Ï (11â¯Ã) usually means âfiercest, mightiestâ. In other words, κάÏÏιÏÏÎ¿Ï actually functions as the superlative of κÏαÏεÏá½¹Ï ~ καÏÏεÏá½¹Ï,86 as the following two examples illustrate:
Il. 6.98â99á¼Î³Ïιον αἰÏμηÏὴν κÏαÏεÏὸν μήÏÏÏÏα Ïόβοιο,á½Î½ δὴ á¼Î³á½¼ κάÏÏιÏÏον á¼Ïαιῶν Ïημι γενέÏθαι
[Diomedes] that savage spearman, a fierce causer of rout who, in my view, is definitely the mightiest of the Achaeans.87
Il. 6.185καÏÏá½·ÏÏην δὴ Ïήν γε μάÏην Ïá½±Ïο δύμεναι á¼Î½Î´Ïῶν
this battle of men, he said, was the fiercest that he ever took part in
Compare the phrases κÏαÏεÏá½´ á½Ïμίνη âfierce battleâ (frequent in Homer) and καÏÏεÏá½´ μάÏη âid.â (Hdt., Th.). In eight of the remaining nine attestations, the being qualified as κάÏÏιÏÏÎ¿Ï is the fiercest or strongest of its group or class. Warriors may be âfiercestâ in comparison with other men,88 and Zeus calls himself θεῶν κάÏÏιÏÏÎ¿Ï á¼ÏάνÏÏν (Il. 8.17) when he threatens the other gods that he will subdue them and throw them into murky Tartarus (cf. also Il. 20.243). The eagle is called ἠμα κάÏÏιÏÏá½¹Ï Ïε καὶ ὤκιÏÏÎ¿Ï ÏεÏεηνῶν (Il. 21.353) âthe strongest and also the swiftest of birdsâ.
There is only one place in Homer where κάÏÏιÏÏÎ¿Ï allegedly means âbestâ (which is also the superlativeâs only occurrence in the Odyssey). Circe instructs Odysseus how to evade the monstrous Scylla:
Od. 12.118â124ἡ δέ Ïοι οὠθνηÏá½µ, á¼Î»Î»â á¼Î¸á½±Î½Î±Ïον κακόν á¼ÏÏι,δεινόν Ïâ á¼Ïγαλέον Ïε καὶ á¼Î³Ïιον οá½Î´á½² μαÏηÏόνÎοá½Î´á½³ ÏÎ¹Ï á¼ÏÏâ á¼Î»Îºá½µÎ ÏÏ Î³á½³ÎµÎ¹Î½ κάÏÏιÏÏον á¼Ïâ αá½Ïá¿Ï.ἢν Î³á½°Ï Î´Î·Î¸á½»Î½á¿Ïθα κοÏÏ ÏÏá½¹Î¼ÎµÎ½Î¿Ï ÏαÏá½° Ïá½³ÏÏá¿,Î´Îµá½·Î´Ï Î¼á½µ Ïâ á¼Î¾Î±á¿¦ÏÎ¹Ï á¼ÏοÏμηθεá¿Ïα κίÏá¿ÏιÏá½¹ÏÏá¿Ïιν κεÏαλá¿Ïι, Ïá½¹ÏÎ¿Ï Ï Î´â á¼Îº ÏῶÏÎ±Ï á¼Î»Î·Ïαι.á¼Î»Î»á½° μάλα ÏÏοδÏá¿¶Ï á¼Î»á½±Î±Î½, â¦
She is not mortal, you know, but an immortal evil: terrible, difficult, wild and not to be fought with. There is no resistance: you must flee from her with all your might (κάÏÏιÏÏον). For if you tarry arming yourself by the cliff, I fear that she will attack again and reach you with as many heads [as before], and catch as many men. No, you should row with all your might (â¦).
Line 120 is commonly translated as: âthere is no defense: fleeing from her is [the] best [thing to do]â.89 Instead of this, I propose to interpret ÏÏ Î³á½³ÎµÎ¹Î½ as an infinitivus pro imperativo, and to take κάÏÏιÏÏον as an adverbially used accusative. This yields the interpretation âyou must flee from her with all your mightâ. This is attractive, because it allows us to view ÏÏ Î³á½³ÎµÎ¹Î½ κάÏÏιÏÏον as syntactically parallel to μάλα ÏÏοδÏá¿¶Ï á¼Î»á½±Î±Î½ in line 124.90
The second (and more complicated) issue is the paradigmatic status of the comparative κÏείÏÏÏν âstronger, superior; betterâ in Homer (17â¯Ã). In West Greek dialects, as we have seen, the root shape of the positive καÏÏεÏá½¹Ï has spread to all derivationally connected forms, including the comparative κάÏÏÏν (literary Doric), καÏÏÏν (Gortyn). In Ionic-Attic, however, the zero grade root was introduced in κÏá½±ÏιÏÏÎ¿Ï but not in κÏείÏÏÏν. This remarkable difference would be explained if κÏείÏÏÏν no longer functioned as the comparative corresponding to κÏá½±ÏιÏÏÎ¿Ï when the latter was reshaped. Can this conjecture be backed up by the semantics of these forms?
In Classical Greek, the core meaning of κÏείÏÏÏν (Ionic κÏá½³ÏÏÏν) is âbetter, strongerâ. In most Homeric attestations, κÏείÏÏÏν means âmore powerful, superiorâ, in a violent confrontation or a duel of main force.91 It occurs either with or without a genitive complement, depending on whether a concrete duel is thought of or whether someone is said to be superior in general. Although the genitive complement betrays its origin as a comparative, κÏείÏÏÏν may almost function as a plain adjective, not only in Classical Greek but already in Homer.92 There is an interesting restriction in the use of κÏείÏÏÏν: almost without exception, it refers to the result of a future conflict that can still be avoided.93 This is neatly illustrated in the first book of the Iliad. The seer Calchas is afraid of Agamemnonâs wrath and asks Achilles for protection before he interprets the dire situation of the Achaeans. His motivation for this request is as follows:
Il. 1.80â83κÏείÏÏÏν Î³á½°Ï Î²Î±ÏÎ¹Î»Îµá½»Ï á½ Ïε Ïá½½ÏεÏαι á¼Î½Î´Ïὶ Ïá½³ÏηÏεἴ ÏÎµÏ Î³á½±Ï Ïε Ïόλον γε καὶ αá½Ïá¿Î¼Î±Ï καÏαÏá½³Ïá¿,á¼Î»Î»á½± Ïε καὶ μεÏá½¹ÏιÏθεν á¼Ïει κόÏον, á½ÏÏα ÏελέÏÏá¿,á¼Î½ ÏÏήθεÏÏιν á¼Î¿á¿ÏιΠÏὺ δὲ ÏÏá½±Ïαι εἴ με ÏαώÏειÏ.
For a king prevails when he gets angry with a lower-ranked man. For even if he swallows down his wrath for that day, afterwards he will cherish resentment in his heart, until he will turn it into action. So tell me whether you will safeguard me.94
These lines illustrate well how Agamemnonâs political power (κÏá½±ÏοÏ) is based on the principle that the threat is stronger than its execution.
Furthermore, it is remarkable that κÏείÏÏÏν never means âmore violent, fiercerâ. Within Epic Greek, it is therefore paradigmatically unrelated to κÏαÏεÏá½¹Ï ~ καÏÏεÏá½¹Ï and κάÏÏιÏÏοÏ, which refer to the might or fierceness of a combatant in a concrete situation, or to being fierce as a characteristic or permanent property. If the Proto-Ionic precursor of Homeric κÏείÏÏÏν was already an isolated comparative formation, just like superior in English, this would explain why the e-grade of the root was preserved exclusively in this form.95
Finally, the paradigmatic isolation of κÏείÏÏÏν is also suggested by existence of a comparative form κÏαÏεÏá½½ÏεÏον ⦠á¼ÎµÎ¸Î»Î¿Î½ â[no] fiercer contestâ (Od. 11.624), which mirrors μάÏη καÏÏεÏá½µ (Hdt., Th.) and κÏαÏεÏá½´ á½Ïμίνη (Hom.). Morphologically, this κÏαÏεÏá½½ÏεÏÎ¿Ï is the regular comparative of κÏαÏεÏá½¹Ï in Homeric Greek.96 As we have just seen, the usual Homeric superlative of καÏÏεÏá½¹Ï ~ κÏαÏεÏá½¹Ï is κάÏÏιÏÏοÏ, but in Hesiod (Th. 864), the productive form κÏαÏεÏá½½ÏαÏÎ¿Ï is predicated of ÏίδηÏÎ¿Ï âironâ in the sense âthe hardest, fiercestâ.97
The conclusions of this discussion are summarized in Table 10.
Table 10
Homeric forms of comparison with the root *κÏεÏâ, κÏαÏâ, καÏÏâ
|
âSuperiorâ; âbetterâ |
âFierceâ |
|
|---|---|---|
|
Positive |
â |
κÏαÏεÏá½¹Ï ~ καÏÏεÏá½¹Ï |
|
Comparative |
κÏείÏÏÏν |
κÏαÏεÏá½½ÏεÏÎ¿Ï |
|
Superlative |
not attested |
κάÏÏιÏÏÎ¿Ï and κÏαÏεÏá½½ÏαÏÎ¿Ï |
5.2.8 κÏαÏá½»Ï and καÏÏá½»Î½Ï in Homer
In Homer, καÏÏá½»Î½Ï is attested only in the phrase á¼ÎºÎ±ÏÏύνανÏο Ïá½±Î»Î±Î³Î³Î±Ï âthey filled their ranks with battle spiritâ, in a repeated battle scene where a phalanx is formed.98 Its different vowel slot in comparison with κÏαÏá½»Ï and Classical κÏαÏá½»Î½Ï requires an explanation. It is unlikely that καÏÏá½»Î½Ï is a relic form whose vowel slot was not affected by that of the base form κÏαÏá½»Ï.
Fortunately, it is not difficult to find a motive for creating καÏÏύνÏ: like κÏá½±ÏιÏÏοÏ, the expected form κÏαÏá½»Î½Ï may have been avoided in Homeric Greek for metrical reasons.99 But what was the linguistic model? At first sight, it seems difficult to indicate an adequate proportional analogy. Whereas other analogical forms with καÏÏâ (κάÏÏοÏ, κάÏÏιÏÏοÏ) are derivationally connected with καÏÏεÏá½¹Ï ~ κÏαÏεÏá½¹Ï and semantically akin to it, καÏÏá½»Î½Ï cannot be derived from καÏÏεÏá½¹Ï for morphological reasons. It cannot be derived from κÏαÏá½»Ï either, because no by-form *καÏÏá½»Ï exists. The best solution is to assume that καÏÏá½»Î½Ï was derived from epic κάÏÏÎ¿Ï in its sense âendurance, strength to persevereâ. Apart from the pair κÏá½±ÏÎ¿Ï : κÏαÏá½»Î½Ï in the spoken language, another model may have been the semantically close θαÏÏá½»Î½Ï âto encourageâ beside θάÏÏÎ¿Ï âperseverance, courageâ. As argued in section 4.5, θαÏÏá½»Î½Ï could be reanalyzed as based on θάÏÏÎ¿Ï after the original form *θαÏÏá½»Ï had been ousted by θαÏÏαλέοÏ.100
Thus, καÏÏá½»Î½Ï âto strengthen, make firmâ is a by-form of the form κÏαÏá½»Î½Ï current in prose. It can be viewed as an inner-epic analogical creation meaning âto provide with κάÏÏοÏâ (in its sense âendurance in battleâ).101
5.2.9 κάÏÏα
The adverb and particle κάÏÏα âvery, much; vehemently; surely, indeedâ is relatively well-attested in Classical Greek: it is frequently used by Herodotus, in the Hippocratic corpus, and in the tragedians. It is not found, however, in inscriptions, Thucydides, Xenophon, or the orators, only a few times in Aristophanes, and it is almost entirely shunned in hexameter poetry.102 Its only occurrence in Plato (Tim. 25d) is in the story about Atlantis, which is said to be due to the elder Critias, who is supposed to have heard it from Solon (á½Ïὸ Ïοῦ Ïαλαιοῦ ÎÏιÏá½·Î¿Ï ÎºÎ±Ïâ á¼ÎºÎ¿á½´Î½ Ïὴν ΣόλÏνοÏ), who in turn was informed by an Egyptian priest. The word therefore probably belongs to a high register in Classical Attic. On the other hand, κάÏÏα remained alive in Eastern Ionic: this is shown by its occurrence not only in Herodotus, but also in Hipponax (fr. 32.2), the poet from Ephesus and Clazomenae reputed for his use of low register vocabulary. Thus, an adverb κάÏÏα existed in Proto-Ionic; it was replaced in Attic by other adverbs such as ÏÏόδÏα.
The semantic relation between κάÏÏα and κÏαÏεÏá½¹Ï ~ καÏÏεÏá½¹Ï in the sense âfierce, vehementâ is clear, and they are surely related.103 The adverbs in âα are a relic formation. The reconstruction of the suffix âα itself is unclear and debated,104 but the root vocalism is usually identical to that of the corresponding adjective, cf.:
-
λίγα âloudlyâ Î»Î¹Î³á½»Ï âsonorousâ
-
μάλα âveryâ μᾶλλον âmoreâ, μάλιÏÏα âmostâ
-
Ïύκα âclosely; frequentlyâ ÏÏ ÎºÎ¹Î½á½¹Ï, ÏÏ ÎºÎ½á½¹Ï âcloseâ
-
ὦκα âswiftlyâ á½ Îºá½»Ï âswiftâ
-
Ïá½±Ïα âquicklyâ ÏαÏá½»Ï âfastâ, comp. θάÏÏÏν, superl. Ïá½±ÏιÏÏοÏ
-
ἦκα âsoftly, lightlyâ á¼¥ÏÏÏν âworseâ, ἤκιÏÏÎ¿Ï âleastâ, adv. ἥκιÏÏα
Thus, forms like Ïá½±Ïα and λίγα may have been influenced by the adjectives ÏαÏá½»Ï and λιγύÏ. In the case of κάÏÏα, however, such influence was only possible as long as the root was still *krÌ¥tâ (contrast the different root shape of κÏαÏá½»Ï). Hence, κάÏÏα must be the regular reflex of a pre-form *krÌ¥ta.
This conclusion is at odds with the widely-shared assumption of a regular development *rÌ¥ > âÏαâ. Illustrative for the embarrassment of previous scholarship is the treatment of Ruijgh (1980: 563 n. 10):
Noter que *κÏá½±Ïα est le résultat phonétique de *krÌ¥tnÌ¥ (cf. á¼Î´Ïακον : δέÏκομαι). Il faut donc expliquer κάÏÏα par une métathèse due au modèle de *κέÏÏÎ¿Ï (attesté par lââ¯anthroponyme Ἰ-κέÏÏηÏ), doublet de κÏá½³ÏοÏ; cf. la substitution de δαÏÏá½¹Ï Ã Î´ÏαÏá½¹Ï dââ¯après δεÏâ. (â¦) La métathèse se retrouve dans καÏÏεÏá½¹Ï, κάÏÏιÏÏοÏ, κάÏÏοÏ, καÏÏá½»Î½Ï et lac. κάÏÏÏν < *κάÏÏÏν, doublets de κÏαÏεÏá½¹Ï etc.
Here, Ruijgh ascribes the variation between âαÏâ and âÏαâ to a similar hesitation concerning the full grade slot in the abstract noun κÏá½³ÏÎ¿Ï ~ *κέÏÏοÏ. This assumption is completely unfounded, as there is no evidence whatsoever for *κέÏÏÎ¿Ï anywhere in Greek. Hodot (1974) has shown that the name ÎκεÏÏηÏ, which occurs only in Lesbian, is unrelated to âκÏá½±ÏηÏ. The normal Lesbian counterpart of Ionic names in âκÏá½±ÏÎ·Ï is âκÏá½³ÏηÏ, which is attested from the 5th c. BCE onward; on the other hand, names in âκέÏÏÎ·Ï first appear in the 2nd c. BCE and must therefore be considered an innovation. Hodot convincingly suggests that this late Lesb. âκέÏÏÎ·Ï is the regular outcome of âκÏá½·ÏηÏ.105
Ruijgh also misses the mark when suggesting (1980: 562 n. 8) that κάÏÏα could be the result of a post-Homeric analogy μάλα : μάλιÏÏα = X : κάÏÏιÏÏα. The problem is, as we have seen, that κάÏÏιÏÏÎ¿Ï is an artificial Homeric form unattested in the Ionic vernacular, whereas κάÏÏα is, conversely, unattested in Homer. Moreover, an adverbial form κάÏÏιÏÏα is unattested in Homer.106
Since an convincing analogical explanation of κάÏÏα is hard to find, I conclude that it probably reflects *krÌ¥ta, thus representing a valuable piece of evidence for the regular development of *rÌ¥ in Proto-Ionic.
5.2.10 From Proto-Ionic to Attic and Ionic
On the basis of our comparison between the Homeric and Classical Ionic-Attic forms in the preceding sections, it is possible reconstruct the following situation for Proto-Ionic, directly after the vocalization *rÌ¥ > âαÏâ and the levelling in κÏαÏá½»Ï, but before âÏαâ was generalized to other full grade forms:
-
adj. kratús
forms of gradation *kréti̯Ån, *krétistos
â factitive verb kratÅ«ÌnÅ
-
n. *krétos
â stative verb *kretéÅ
â compounds in *âkretá¸s
-
adj. karterós
-
adv. kárta.
Homeric Greek and Classical Ionic-Attic agree in having the forms κÏá½±ÏοÏ, âκÏαÏá½µÏ, and κÏαÏá½³Ï. Moreover, the superlative κÏá½±ÏιÏÏÎ¿Ï is found in both Attic and Ionic, and was probably avoided in Homer for metrical reasons. This suggests that as early as Proto-Ionic, âÏαâ was introduced from the adjective in these forms, but not in *kreti̯Ån. This introduction led to the following situation:107
-
κÏαÏá½»Ï âpowerful, fierce; firmâ
κÏá½±ÏιÏÏÎ¿Ï âmost powerfulâ > âbestâ
â κÏαÏá½»Î½Ï âto make firm, hardenâ
-
*kreti̯Ån âsuperiorâ
-
κÏá½±ÏÎ¿Ï âfierceness; powerâ
â κÏαÏá½³Ï âto be rampant; be in controlâ
â âκÏαÏá½µÏ âhaving powerâ
-
καÏÏεÏá½¹Ï âfierce, strong; steadfast, enduring; firmâ
-
κάÏÏα âvehemently, firmlyâ (> âveryâ).
The root variant κÏαÏâ first arose in the adjective κÏαÏá½»Ï. Whereas this form is a relic already in Homer, it must have been alive in Proto-Ionic because the introduction of a-vocalism in κÏá½±ÏÎ¿Ï and κÏá½±ÏιÏÏÎ¿Ï started out from this form.
It is impossible to assume influence of καÏÏεÏá½¹Ï on the vocalism of *krétos, *krétistos. First of all, the forms καÏÏεÏá½¹Ï and κÏá½±ÏÎ¿Ï have different vowel slots. If the vocalized zero grade had been levelled, the result would have been κάÏÏοÏ, as in Gortynian Cretan. The variant κÏαÏεÏá½¹Ï cannot be reconstructed for Proto-Ionic; it originated within the epic tradition. Had the Homeric doublet κÏαÏεÏá½¹Ï ~ καÏÏεÏá½¹Ï also existed in the vernacular, it would be difficult to understand why καÏÏεÏá½¹Ï, with its deviant vowel slot, was not ousted by κÏαÏεÏá½¹Ï. Secondly, in Classical Greek the semantic divergence between καÏÏεÏá½¹Ï âsteadfast, perseveringâ and κÏá½±ÏÎ¿Ï âpowerâ is considerable. Both forms have their own system of derivations, and already in Homer the analogical form κάÏÏÎ¿Ï serves as an abstract to καÏÏεÏá½¹Ï, not as a pure doublet of κÏá½±ÏοÏ. Admittedly, this argument is not decisive because the semantic distance between καÏÏεÏá½¹Ï and κÏá½±ÏÎ¿Ï may originally have been more narrow.108 However, it would make sense if the meaning of κÏαÏá½»Ï was something like âpowerful, violentâ (closer to that of κÏá½±ÏοÏ), while that of καÏÏεÏá½¹Ï was already more like âsteadfast, enduring, persevering, firmâ.
5.2.11 The Reconstruction of κÏαÏÎ±Î¹á½¹Ï and κÏαÏαιâ
It is generally agreed that κÏαÏαιόÏ, κÏαÏαιâ and κÏαÏαιá¿Ï (PN ÎÏá½±ÏαιÏÏ) had pre-forms with a zero grade root *krÌ¥tâ, but their precise morphological reconstruction and the origin of âαιâ have been a topic of debate. I will first summarize and criticize previous accounts of κÏαÏÎ±Î¹á½¹Ï and κÏαÏαιâ, and then propose my own scenario accounting for both forms. Concerning κÏαÏαιá¿Ï (ÎÏá½±ÏαιÏÏ), this form is derived from κÏαÏαιόÏ, as I will argue in the next section.
The meaning of κÏαÏÎ±Î¹á½¹Ï is more or less identical to that of κÏαÏεÏá½¹Ï ~ καÏÏεÏá½¹Ï.109 A broadly shared assumption is that the masculine κÏαÏÎ±Î¹á½¹Ï is a back-formation from the feminine attested in the formula Îοá¿Ïα κÏαÏαιή.110 It is supposed that κÏαÏαιή somehow continues an archaic motional feminine *krÌ¥th2u̯ih2 of the u-stem adjective κÏαÏá½»Ï, where the second âαâ would be the vocalization product of *h2. This reconstruction is inspired by that of the toponym ΠλάÏαια, which is analyzed similarly as the direct outcome of the old feminine *plÌ¥th2u̯ih2 of ÏλαÏá½»Ï âwideâ.111 That ΠλάÏαια (pl. ΠλαÏαιαί) is indeed a relic u-stem form is corroborated by its pattern of accentuation, which recurs only in a few archaic motional feminines of Greek u-stem adjectives: λίγεια, θάλεια, λάÏεια and the pluralia tantum ÏαÏÏειαί, θαμειαί. The root-final laryngeal of PIE *pleth2- is corroborated by the voiceless aspirate of Indo-Iranian (Ved. prathi) and the Greek noun ÏλαÏαμών âflat surfaceâ.
Severe problems arise, however, when this explanation is extended to κÏαÏαιόÏ. First of all, there is no independent evidence that the root ended in *h2. Meissner (2006: 62), accepting the etymological connection with Ved. krátu- and Av. xratu- (which exclude root-final *âh2â), assumes that an extended form *âh2uâ of the suffix *âuâ somehow became productive in Proto-Greek. Lamberterie (1990: 352â353), while deriving κÏαÏá½»Ï from *kertâ âcutâ, is forced to assume a contamination of that root with *(s)kerHâ âcutâ (which would have given rise to *kerth2â) in order to account for κÏαÏαιâ and κÏαÏαιόÏ. Both ideas are designed specifically in order to explain κÏαÏαιόÏ, and neither is supported unambiguously by further evidence.112
A second problem concerns the inner-Greek developments assumed to lead to κÏαÏαιόÏ. The expected feminine of a u-stem adjective would be *κÏá½±Ïαιᾰ, but this form is unattested, and Homer already has the archaic formula Îοá¿Ïα κÏαÏαιή. Both the quantity of the final vowel of κÏαÏαιή and its oxytone accentuation are problematic, and there would have been no motive for replacing *κÏá½±Ïαιᾰ in Îοá¿Ïα κÏαÏαιή, as petrified formulae cannot be expected to undergo unmotivated analogical morphological changes. Finally, the creation of a secondary masculine κÏαÏÎ±Î¹á½¹Ï beside κÏαÏαιή (Lamberterie 1990: 339) is not without problems: most of the parallels adduced by Lamberterie to explain this process are post-Classical, and their similarities with κÏαÏÎ±Î¹á½¹Ï are weak.113
In view of these objections, the reconstruction of κÏαÏÎ±Î¹á½¹Ï remains problematic. Before presenting my new account, let us consider the reconstruction of the first member κÏαÏαιâ. A number of details remain puzzling: what is the origin of âαι-? Is there a derivational relation between κÏαÏαιâ and κÏαÏαιόÏ? And how do personal names with ÎαÏÏιâ, ÎÏαÏι- fit into the picture?
A brief digression about Calandâs rule for replacing adjectival suffixes with âiâ in first compound members (cf. section 4.1) will be in order here. This rule may have been productive until not too long before Homer. The most important examples of first compound members in âιâ are the forms shown in Table 11 (next page).114 Several of these forms co-occur with adjectives in âÏá½¹â or âνόâ, just as in the Indo-Iranian examples: á¼ÏÏ ÏÎ¹Ë : á¼ÏÏ Î¸Ïá½¹Ï, ÎºÏ Î´Î¹Ë : ÎºÏ Î´Ïá½¹Ï, Î»Î±Î¸Î¹Ë : λάθÏá¿, ÏÏ ÎºÎ¹Ë : ÏÏ ÎºÎ½á½¹Ï, perhaps á¼ÏÎ³Î¹Ë : á¼ÏÎ³á½¹Ï (if one accepts Wackernagelâs idea of a dissimilation from *argróâ), and cf. also δαá¿ÏÏÏν < *dnÌ¥s-iË beside Ved. dasráâ âcapable, skilledâ < *dnÌ¥s-róâ. In all cases where more than one ablaut variant exists, the first member in âιâ has a zero grade root: λαθιâ, ÏÏ ÎºÎ¹â, δαÏâ, á¼ÏÏ Ïιâ, just like the corresponding adjectives.
Table 11
Greek âCalandâ first members in âιâ compared with related forms
|
First compound member |
Related formations |
|---|---|
|
á¼Ïγι-κέÏÎ±Ï Î½Î¿Ï, âÏÎ¿Î´ÎµÏ (Hom.) |
ËαÏÎ³á½µÏ âbrightâ, á¼ÏÎ³á½¹Ï âswift; whiteâ |
|
δαá¿-ÏÏÏν âcleverâ (Hom.)115 |
δήνεα âwilesâ, á¼Î´Î·Î½á½µÏ, ÏÎ¿Î»Ï Î´á½µÎ½Î·Ï |
|
καλλι-γύναικα (Hom.) |
Îºá½±Î»Î»Î¿Ï âbeautyâ, ÏεÏÎ¹ÎºÎ±Î»Î»á½µÏ |
|
ÎºÏ Î´Î¹-άνειÏα (Hom.) |
ÎºÏ Î´Ïá½¹Ï âstatelyâ, Îºá¿¦Î´Î¿Ï âgloryâ, á¼ÏÎ¹ÎºÏ Î´á½µÏ |
|
λαθι-ÎºÎ·Î´á½µÏ (Hom.) |
λάθÏá¾³ adv. âsecretlyâ, Hom. λάθÏá¿ |
|
ÏÏ ÎºÎ¹-Î¼á½µÎ´Î·Ï (Hom.) |
ÏÏ ÎºÎ½á½¹Ï, ÏÏ ÎºÎ¹Î½á½¹Ï âclose; sharp-wittedâ |
|
ῥᾴ-Î¸Ï Î¼Î¿Ï âlight-heartedâ (Att.)116 |
ῥεá¿Î± (< *á¿¥á¿Î±) âeasilyâ |
|
á¼ÏÏ Ïá½·-ÏελαÏ117 (Hp.+) |
á¼ÏÏ Î¸Ïá½¹Ï âredâ |
|
θεÏÏι-εÏá½µÏ (B.)118 |
Aeol. θέÏÏοÏ; names in ËθέÏÏÎ·Ï (Hom.+) |
|
PNâ¯s ÎÏαÏιâ, ÎαÏÏιâ (post-Hom.)119 |
κÏαÏεÏá½¹Ï, καÏÏεÏá½¹Ï âviolent, firmâ |
Several examples suggest that the basis for deriving first compound members in âÎ¹Ë has been extended to include s-stems, both abstract nouns and second compound members. The relic first member θεÏÏÎ¹Ë (retaining the e-grade) seems to have been formed beside θέÏÏÎ¿Ï (the older form of θάÏÏοÏ) and ËθέÏÏηÏ.120 Likewise, ÎºÎ±Î»Î»Î¹Ë with its geminate mirrors not the simplex καλόÏ, but the noun Îºá½±Î»Î»Î¿Ï and compounds in ËκαλλήÏ, and we may also compare á¼ÏÎ³Î¹Ë : ËαÏÎ³á½µÏ and ÎºÏ Î´Î¹Ë : ËÎºÏ Î´á½µÏ.
Another indication that Calandâs replacement rule had ceased to be operative is the use of κÏαÏεÏÎ¿Ë as a first compound member, instead of the outcome of *krÌ¥tiË or, for that matter, unattested *κÏαÏÏ Ë.121 This brings us back to the origins of κÏαÏαιË. Interestingly, as remarked by Meissner (2006: 18), κÏαÏÎ±Î¹Ë beside κÏαÏεÏá½¹Ï is the only instance in Homer where adjectival âÏοâ is not replaced by âÎ¹Ë in a first compound member. Is it possible that κÏαÏÎ±Î¹Ë somehow reflects *krÌ¥tiË?
Let us first reconsider the existing opinions on the origin of âÎ±Î¹Ë in κÏαÏαιË. Meissner (2006: 62), citing γεÏÎ±Î¹á½¹Ï âoldâ beside the comparative γεÏαίÏεÏÎ¿Ï as a parallel, argued that κÏαÏÎ±Î¹Ë is a remodeling of κÏαÏαιοË.122 This is unlikely: it remains unclear why the thematic vowel would be dropped; the assumed influence of γεÏÎ±Î¹á½¹Ï on κÏαÏÎ±Î¹á½¹Ï lacks a clear motivation; and γεÏαίÏεÏÎ¿Ï is not a compound but a comparative. Reconstructing a pre-form *ḱrth2iË does not really help: between two consonants, PIE *âh2iâ is expected to yield âιâ rather than âαιâ (cf. Nagy 1999: 86â87 with n. 5).
Nagy therefore assumed that κÏαÏÎ±Î¹Ë arose within Greek as a crossover of the adverb *krÌ¥t-a (> κάÏÏα) and the inherited first member *krÌ¥tiË. He follows a suggestion by Nussbaum that adverbs in âα could also appear in place of a first member in âιË, as in the names á¼Î»Îºá½±Î¸Î¿Î¿Ï (Il. 12.93) and á¼Î»ÎºÎ±Î¼á½³Î½Î·Ï (Bechtel 1917: 35) beside á¼Î»Îºá½·ÏÏÏν. Indeed, a first member ÎÏαÏÎ¹Ë is also found in personal names, but one would have to assume that the adverb *krÌ¥ta could be used as the first member of a compound, and that it was then contaminated with *krÌ¥tiË. This is not impossible, but difficult to substantiate.
Although the existence of κάÏÏα lends some support to this scenario, it remains difficult to indicate a motive for creating *krÌ¥taiË. Why not simply retain *krÌ¥tiË if this already existed anyway? Turning around our perspective, if we were to assume that *krÌ¥taiË is old, a motivation for its retention is readily available. In Epic Greek, κÏαÏÎ±Î¹Ë functions as an allomorph of κÏαÏεÏοË, as is clearly illustrated by personal names with ÎÏαÏÎ±Î¹Ë corresponding to noun phrases with κÏαÏεÏá½¹Ï (cf. Lamberterie 1990: 337):
-
ÎÏαÏÎ±Î¹Î¼á½³Î½Î·Ï (Th., inscr.)123 ~ Hom. κÏαÏεÏὸν μένοÏ;
-
ÎÏαÏÎ±á½·Î²Î¹Î¿Ï (inscr. Delos)124 ~ Hom. κÏαÏεÏá¿Ïι βιá¿Ïι.
Another crucial form is καÏÏαίÏοδâ, attested both in Pindar (qualifying a bull in Ol. 13.81) and in Cretan inscriptions (in the meaning âcattleâ). Its variant κÏαÏαίÏÎ¿Î´ÎµÏ âwith strong feetâ appears as an epithet of ἡμίονοι âmulesâ in the Herodotean Life of Homer.125 This must surely be compared with the Homeric epithet κÏαÏεÏá¿¶Î½Ï Î¾, which (with one exception) qualifies horses or mules.126 In other words, κÏαÏαίÏÎ¿Î´ÎµÏ âwhose Ïá½¹Î´ÎµÏ are κÏαÏεÏοίâ is parallel to κÏαÏεÏá¿¶Î½Ï Î¾ âwhose á½Î½Ï ÏÎµÏ are κÏαÏεÏοίâ.
What determined the choice between κÏαÏÎ±Î¹Ë and κÏαÏεÏοË? It is true that κÏαÏÎ±Î¹Ë also occurs in non-dactylic meters, but I think the underlying principle is best sought within the epic language. The only two Homeric compounds with a first member κÏαÏÎ±Î¹Ë are the hapax eiremena κÏαÏαιγύαλοι and κÏαÏαίÏεδον, whose second member has a light first syllable starting with a single consonant. The same applies to the post-Homeric personal names ÎÏαÏÎ±Î¹Î¼á½³Î½Î·Ï and ÎÏαÏαίβιοÏ, which may well be of epic origin, and to post-Homeric κÏαÏαίÏοδεÏ. In this phonological context, κÏαÏεÏÎ¿Ë and καÏÏεÏÎ¿Ë were both excluded for metrical reasons in Epic Greek. We do find κÏαÏεÏÎ¿Ë before second members starting with two consonants or a vowel-initial heavy syllable (κÏαÏεÏá½¹ÏÏÏν, κÏαÏεÏá¿¶Î½Ï Î¾), while καÏÏεÏÎ¿Ë was used if the second member started with a heavy syllable with a single initial consonant (cf. καÏÏεÏόθῡμοÏ).127
If the alternation of *krÌ¥taiâ with *krÌ¥teroâ as first compound members represents something old and structural, then the same must hold for *krÌ¥taiâ itself. But how old is this form? Lamberterie (1990: 343) proposes the following explanation:
Dans les composés, le système de Caland fait attendre un premier membre καÏÏιâ, κÏαÏιâ, attesté effectivement dans lââ¯onomastique; une fois constitué lââ¯adjectif κÏαÏÎ±Î¹á½¹Ï (â¦), on conçoit quââ¯il ait pu fournir aux aèdes un modèle pour faire entrer dans lââ¯hexamètre des formes amétriques comme *κÏαÏá½·ÏÎµÎ´Î¿Ï ou *κÏαÏιγύαλοÏ.
This is attractive, but I would prefer to view formulate the details slightly differently. An inherited first compound member *krÌ¥tiâ is expected on comparative grounds as a counterpart of *krÌ¥teroâ. In Epic Greek, it was problematic to use *krÌ¥tiâ before a single consonant followed by a light syllable. This may have initially been resolved with a metrical lengthening, *krÌ¥tipedoâ >> *krÌ¥tÄ«pedoâ. At some point, *krÌ¥tÄ«â was for some reason (perhaps because it had become isolated) reshaped as *krÌ¥taiâ under the influence of *krÌ¥taiu̯óâ, the pre-form of κÏαÏαιόÏ.
In his immediately following remark, however, Lamberterie expresses his doubts about this explanation:
Il reste que lââ¯existence de καÏÏαá¿ÏÎ¿Ï en Crète interdit de voir dans les composés en κÏαÏαιâ une création littéraire artificielle; ils ont bel et bien une réalité linguistique.
The Cretan word καÏÏαιÏοδâ is indeed highly relevant, as it shows that compounds in *krÌ¥taiâ existed before the vocalization of *rÌ¥ in this dialect.128 I would hesitate, however, to view καÏÏαιÏοδâ as a compound of vernacular origin. On the face of it, it looks like a poetic form: an epithet which replaced or supplemented an older word for âcattleâ such as ÏεÏÏαÏοδâ. In fact, the Cretan form could show that a compound *krÌ¥tai-podâ (reflected in post-Homeric κÏαÏαίÏοδεÏ) was created in an early form of the Greek poetic tradition, probably epic, and that it was borrowed thence into an early form of Cretan. One is tempted to think of an âAchaeanâ relic form.
The question still remains how the compounds with κÏαÏαιâ relate to personal names with ÎαÏÏιâ and ÎÏαÏιâ, which at first sight may contain the expected outcome of *krÌ¥tiâ.129 Meissner (1998: 244â245, cf. also GEW s.v. κÏá½±ÏοÏ) objects to this idea that the attestations are not very early: one example possibly dates from the fifth century, and the rest is from the fourth century or younger. For this reason, he claims that these names could be innovations of the classical period, when first members in âιâ enjoyed a certain productivity. Concerning the Homeric evidence, he views the absence of καÏÏιâ ~ κÏαÏιâ in Homer as an argument against its antiquity.130 However, this absence is not necessarily remarkable given that κÏαÏαιâ occurs only twice. It is also relevant that ÎÏαÏεÏοâ ~ ÎαÏÏεÏοâ is unattested as a first member in Ionic and Attic personal names. It therefore seems plausible that *krÌ¥tiâ once existed, whether or not the personal names with ÎÏαÏιâ and ÎαÏÏιâ directly reflect this form.131
Let us now return to the reconstruction of κÏαÏαιόÏ. In view of its oxytone accentuation, a derivation with the unaccented all-purpose appurtenance suffix âιοâ can be excluded. Three other adjectives are of special interest as possible parallel formations: ÏÎ±Î»Î±Î¹á½¹Ï âof the pastâ, γεÏÎ±Î¹á½¹Ï âoldâ, and Î´Î·Î½Î±Î¹á½¹Ï âlong-livedâ.132 Among these, Î´Î·Î½Î±Î¹á½¹Ï occupies a special position because it probably derives from a compound *du̯Än-aiu̯-óâ â(one) having a long life-spanâ, with the adverb δήν as a first member.133 On the other hand, the adjective ÏÎ±Î»Î±Î¹á½¹Ï is already attested in Myc. pa-ra-jo, thus excluding a compound with *âaiu̯â.134 Clearly, ÏÎ±Î»Î±Î¹á½¹Ï is an adjectivization of the adverb Ïάλαι âin times before, for some time nowâ, which also occurs as the first member of compounds (e.g. ÏÎ±Î»Î±Î¹Î³ÎµÎ½á½µÏ âborn in the pastâ, ÏαλαίÏαÏÎ¿Ï âsaid/hit some time agoâ). Later, ÏÎ±Î»Î±Î¹á½¹Ï may have induced the creation of γεÏαιόÏ.135
Thus, κÏαÏÎ±Î¹á½¹Ï might either be a âgenitivalâ derivation like ÏαλαιόÏ, or a possessive compound like δηναιόÏ. Concerning the former option, as remarked by Nagy (1999: 353),136 κÏαÏαιâ (unlike Ïάλαι) is not found as a simplex. Moreover, even if we were to explain κÏαÏÎ±Î¹á½¹Ï from a proportion based on κÏαÏαιâ beside Ïαλαιâ in compounds (cf. GEW s.v. κÏá½±ÏοÏ), the form κÏαÏαιâ itself remains in need of an explanation. The best solution is, therefore, to assume that κÏαÏαιâ was reshaped from metrically lengthened *krÌ¥tiâ under the influence of an already existing κÏαÏαιόÏ, perhaps aided by Ïάλαι beside ÏαλαιόÏ.
Most problems can be resolved by deriving κÏαÏÎ±Î¹á½¹Ï from a compound PGr. *krÌ¥t-aiu̯óâ âhaving powerful vital forceâ, in which *âaiu̯óâ continues the PIE u-stem *h2ói-u, *h2éi-uâ (cf. Ved. ÄÌyu(s)â n. âlife, lifetime, vital forceâ) or even the thematic derivative *h2eiu̯o- reflected in Lat. aevum âlifetime, ageâ, Goth. aiws âid.â. Note that in Homer αἰών does not only mean âlife, lifetimeâ, but is also used as an equivalent of Î¼á½³Î½Î¿Ï in the sense âvital forceâ (cf. LfgrE s.v. αἰών), a sense which may also be attested for the Vedic cognate just cited. The first member of *krÌ¥t-aiu̯óâ could be the bare root or, alternatively, a prevocalic variant of the âCalandâ allomorph *krÌ¥tiâ (with elision of âiâ). This reconstruction receives support from the existence of traditional phrases κÏαÏεÏὸν Î¼á½³Î½Î¿Ï and κÏαÏεÏá½´ βίη and the corresponding personal names ÎÏαÏαιμένηÏ, ÎÏαÏαίβιοÏ.137
5.2.12 ÎÏá½±ÏαιÏÏ and κÏαÏαιá¿Ï
The enigmatic form κÏαÏαιá¿Ï (PN ÎÏá½±ÏαιÏÏ) occurs twice in Homer, both times in the Odyssey. The first occasion is in the description of the sinner Sisyphus in the Nekuia:
Od. 11.595â598ὠμὲν ÏκηÏιÏÏá½¹Î¼ÎµÎ½Î¿Ï ÏεÏÏίν Ïε ÏοÏίν Ïελᾶαν á¼Î½Ï ὤθεÏκε ÏοÏὶ λόÏονΠá¼Î»Î»â á½ Ïε μέλλοιá¼ÎºÏον á½ÏεÏβαλέειν, Ïá½¹Ïâ á¼ÏοÏÏÏá½³ÏαÏκε κÏαÏαιá¿ÏÎαá½ÏÎ¹Ï á¼ÏειÏα Ïέδονδε ÎºÏ Î»á½·Î½Î´ÎµÏο Î»á¾¶Î±Ï á¼Î½Î±Î¹Î´á½µÏ.
⦠he [Sisyphus] would brace himself with hands and feet, and thrust the stone up toward the crest of a hill, but as often as he was about to throw it over the top, [a] κÏαÏαιá¿Ï would turn it back, and then the ruthless stone would come rolling down to the plain again.
In this episode, many commentators translate κÏαÏαιá¿Ï as âoverwhelming weight, preponderanceâ (e.g. âdas Ãbergewicht, seine Wuchtâ, Ameis-Hentze ad loc.). Editors like von der Mühll and van Thiel print ÎÏαÏαιá¿Ï,138 but personal names normally do not have oxytone accentuation. Moreover, a personal name ÎÏá½±ÏαιÏÏ, with retracted accent, is ascertained by the second attestation, when Odysseus is warned by Circe about the monstrous Scylla:
Od. 12.120â126οá½Î´á½³ ÏÎ¹Ï á¼ÏÏâ á¼Î»Îºá½µÎ ÏÏ Î³á½³ÎµÎ¹Î½ κάÏÏιÏÏον á¼Ïâ αá½Ïá¿Ï.ἢν Î³á½°Ï Î´Î·Î¸á½»Î½á¿Ïθα κοÏÏ ÏÏá½¹Î¼ÎµÎ½Î¿Ï ÏαÏá½° Ïá½³ÏÏá¿,Î´Îµá½·Î´Ï Î¼á½µ Ïâ á¼Î¾Î±á¿¦ÏÎ¹Ï á¼ÏοÏμηθεá¿Ïα κίÏá¿ÏιÏá½¹ÏÏá¿Ïιν κεÏαλá¿Ïι, Ïá½¹ÏÎ¿Ï Ï Î´â á¼Îº ÏῶÏÎ±Ï á¼Î»Î·Ïαι.á¼Î»Î»á½° μάλα ÏÏοδÏá¿¶Ï á¼Î»á½±Î±Î½, βÏÏÏÏεá¿Î½ δὲ ÎÏá½±ÏαιÏν,μηÏá½³Ïα Ïá¿Ï ΣκύλληÏ, á¼¥ μιν Ïέκε Ïá¿Î¼Î± βÏοÏοá¿ÏινÎá¼¥ μιν á¼ÏειÏâ á¼ÏοÏαύÏει á¼Ï á½ÏÏεÏον á½Ïμηθá¿Î½Î±Î¹.
There is no resistance: you must flee from her with all your might.139 For if you tarry arming yourself by the cliff, I fear that she will attack again and reach you with as many heads [as before], and catch as many men. No, you should row with all might, and call upon Crataeis, the mother of that Scylla, who bore her to be a bane to mortals. She will then keep her from leaping forth again.
Aristarchus held the opinion that lines 124â126 in this passage were later additions. However, the fact that line 124 contains both the hapax βÏÏÏÏεá¿Î½ and the rare ÎÏá½±ÏαιÏν strongly advises against athetizing it. Indeed, the idea has been mostly abandoned in more recent scholarship (see Heubeck, Comm. Od. ad loc.), but Merkelbach (1951) still argued that lines 125â126 (not 124) are late additions. Given that the two lines seem to contain general explanatory statements, this is a definite possibility. In fact, in view of the preceding μάλα ÏÏοδÏá¿¶Ï á¼Î»á½±Î±Î½ âyou should row very quicklyâ, the two lines bring up a question: why would Odysseus and his team have to row so quickly if they can also call upon Scyllaâs mother to restrain her daughter?140 In any case, it seems best to retain at least line 124 as authentic.
The correct morphological analysis of κÏαÏαιá¿Ï and ÎÏá½±ÏαιÏÏ has also yielded problems from antiquity onwards; see Lamberterie (1990: 340â343) for a clear summary of the issues. The following proposals are insufficient:
-
Aristarchus analyzed κÏαÏαιá¿Ï as an adverb in âιÏ. This cannot be correct, as the transitive verb á¼ÏοÏÏÏá½³ÏαÏκε âpushed back (repeatedly)â (Od. 11.597) is in need of a subject, and the only candidate to fulfill this role is precisely κÏαÏαιá¿Ï.
-
Chantraine (DELG) analyzed κÏαÏαιá¿Ï as an adjective: an anomalous feminine of κÏαÏÎ±Î¹á½¹Ï which in his view qualifies Î»á¾¶Î±Ï âstoneâ in the next line. For the formation, he compared the Homeric feminine θοῦÏιÏ, belonging to θοῦÏÎ¿Ï âfierceâ. However, as Lamberterie (l.c.) remarks, Î»á¾¶Î±Ï is always masculine in Homer.
-
The idea of an irregular elision in an underlying κÏαÏαίâ á¼´Ï cannot be maintained either, in view of the long âη as well as the initial digamma of á¼´Ï âforceâ. Lamberterieâs proposal (l.c.) to reconstruct a noun phrase *κÏαÏαιÏá½° Ïá½·Ï, which would have developed to κÏαÏαιá¿Ï by haplology, remains pure speculation.
-
Nagy (1999: 349â350) analyzed κÏαÏαιá¿Ï as a possessive compound *κÏαÏαι-ÏÎ¹Ï âwhose force has κÏá½±ÏοÏâ, which would make sense from a semantic point of view.141 However, the phonological problems remain. If the second member was indeed á¼´Ï âforceâ, with a long vowel from PIE *u̯iHâ, this would be contradicted by the accentuation of the name ÎÏá½±ÏαιÏÏ, which presupposes a short final syllable.142 It would be unmotivated to assume a secondary shortening of the *Ä«.
As was already seen by Wackernagel (1914: 111), the most natural analysis of κÏαÏαιá¿Ï ~ ÎÏá½±ÏαιÏÏ is to view it as a feminine substantivization of the type Î½Ï ÎºÏεÏá½·Ï âbatâ < â(creature) of the nightâ.143 As we have seen, the context in which κÏαÏαιá¿Ï occurs suggests that it means something like âoverwhelming forceâ. Nagy objects to Wackernagelâs analysis that the adjectives corresponding to Î½Ï ÎºÏεÏá½·Ï and ἡμεÏá½·Ï âcultivated vineâ (νύκÏεÏÎ¿Ï ânightlyâ and ἥμεÏÎ¿Ï âtame, cultivatedâ) retain their feminine in âοÏ, whereas κÏαÏÎ±Î¹á½¹Ï has a feminine in âá½µ. However, I fail to see why the possibility to derive a substantivization in âίδâ would be affected by the presence or absence of explicit feminine marking in the adjective, as âίδâ is a derivational suffix making substantivizations, not a flexional suffix creating feminine forms of an adjective.144 We may therefore conclude that κÏαÏαιá¿Ï was derived directly from κÏαÏαιόÏ, and that both forms reflect a pre-form with *rÌ¥.
5.3 Conclusions on the Vocalization of *r̥
Although various proposals for the etymology of κÏαÏá½»Ï and καÏÏεÏá½¹Ï have been made, we are clearly dealing with a homogeneous group of words, pace Benveniste (1969). The lexical meanings of the root are âpowerful, overwhelmingâ and âsteadfast, firmâ, and we have seen possibilities to connect these by semantic developments. The original full grade was *kretâ (cf. Ion. κÏá½³ÏÏÏν), and there is no convincing evidence for positing a variant *kertâ. This means that καÏÏεÏá½¹Ï and κάÏÏα show a regular reflex of the zero grade, and that the reflex of κÏαÏá½»Ï must be analogical after the full grade *kretâ, as in other u-stem adjectives, by the scenario discussed in chapter 4.
Both adjectives *krÌ¥túâ and *krÌ¥teróâ must be reconstructed for Proto-Greek. A secondary creation of *krÌ¥túâ > κÏαÏá½»Ï would be difficult to justify in view of the unproductive status of this category in Greek. On the other hand, *krÌ¥teróâ must also be old because this form is reflected in three different dialect groups. Early on, possibly already in Proto-Greek, it underwent a semantic development to âpersevering, steadfastâ, and was thereby dissociated from forms like *krétos âmight; powerâ and *krÌ¥ta âvehementlyâ. To be sure, the original semantic differences remain difficult to reconstruct, but it is plausible that a nuance between *krÌ¥teróâ and *krÌ¥túâ existed early on.
Our analysis of the derivational history of forms with κÏαÏâ and καÏÏâ has corroborated the conclusions reached in previous chapters. Starting from κÏαÏá½»Ï, the analogical zero grade reflex was introduced in κÏá½±ÏιÏÏοÏ, κÏá½±ÏÎ¿Ï (whence âκÏαÏá½µÏ) and the stative verb κÏαÏá½³Ï. None of these forms can therefore be used as evidence for the regular vocalization of *rÌ¥. Itself, κÏαÏá½»Ï was retained only in the name-epithet formula |H κÏαÏá½ºÏ á¼ÏγεÏÏόνÏÎ·Ï and was apparently eliminated relatively early (though after the splitting up of Proto-Ionic). In Epic Greek, the role of κÏαÏá½»Ï was taken over by κÏαÏεÏá½¹Ï. It is possible to view κÏαÏεÏá½¹Ï as a cross-over between κÏαÏá½»Ï and the older form *krÌ¥teróâ in the epic tradition.
Once the doublet κÏαÏεÏá½¹Ï ~ καÏÏεÏá½¹Ï existed, analogies within Epic Greek could lead to the creation of further doublet forms: κάÏÏÎ¿Ï (beside κÏá½±ÏοÏ), κάÏÏιÏÏÎ¿Ï (beside κÏá½±ÏιÏÏοÏ); καÏÏá½»Î½Ï (beside κÏαÏύνÏ). These forms are artificial in the sense that they arose by inner-epic analogies. Classical prose did not develop such by-forms: it only has καÏÏεÏá½¹Ï (whence καÏÏεÏá½³Ï), κÏá½±ÏοÏ, κÏá½±ÏιÏÏοÏ, and κÏαÏύνÏ.
The epic forms κÏαÏαιâ, κÏαÏÎ±Î¹á½¹Ï and κÏαÏαιá¿Ï also reflect a pre-form with *rÌ¥. Since it is difficult to account for their root shape as analogical, they seem to contradict the conclusion reached on the basis of καÏÏεÏá½¹Ï and κάÏÏα. However, we must also take into account that κÏαÏαιâ, κÏαÏÎ±Î¹á½¹Ï and κÏαÏαιá¿Ï are found mainly in Epic Greek and occasionally in later poetry. Moreover, using these forms in hexameter verse entailed a tautosyllabic scansion of muta cum liquida, which is relatively rare in Homer. In the next chapter, I will deal with these issues in more detail, and propose that the reflex âÏαâ < *rÌ¥ in these and other forms evolved not in a vernacular dialect, but in the epic tradition. This means that καÏÏεÏá½¹Ï and κάÏÏα are the only formations belonging to this root to display the regular vocalization of *rÌ¥ in Ionic-Attic.
For instance, Lejeune (1972: 196) invokes a âmobilité généraleâ of liquids within the syllable; Strunk (1975: 286) remarks that âinlautendes âαÏâ < *ârÌ¥â vor Konsonant (â¦) auch sonst gelegentlich statt oder neben âÏαâ vorkommtâ.
As far as the Greek evidence is concerned, previous treatments include Trümpy (1950: 202â¯ff.), Frisk (GEW s.v. κÏá½±ÏοÏ), Benveniste (1969), Strunk (1975), Breuil (1989), Lamberterie (1990: 323â353). For an overview of the older literature, see Strunk (1975: 265â266).
The meaning of the Indo-Iranian word has been much discussed; the best rendering seems to be âwill-power, resolution, resolveâ (German Entschlossenheit), which is close to a basic meaning of Î²Î¿Ï Î»á½µ in Homeric Greek. The connection with κÏαÏá½»Ï is accepted by Mayrhofer (EWAia s.v. krátuâ), Risch (1974), and also (with some hesitation) by Frisk (GEW s.v. κÏá½±ÏοÏ).
Lamberterie (1990: 336) concludes that âles emplois de κÏαÏεÏá½¹Ï concordent avec ceux de κÏαÏύνÏ: lââ¯un comme lââ¯autre amènent à restituer pour κÏαÏá½»Ï, par reconstruction interne, le sens de «â¯dur, ferme, solideâ¯Â».â
Attested as a verbal root in Hitt. kartae-zi âto cut offâ, Ved. kart âto cut (off), split, breakâ (pres. krÌ¥ntáti, them. aor. krÌ¥táâ, both RV+), Lith. kirÌsti âto hew, hit, cutâ (pres. 1sg. kertù), and PSlav. *Äersti âto carve, slashâ (ORu. ÄÑrsti), and probably in Hom. á¼ÎºÎµÏÏε âcut (off)â.
Lamberterie convincingly derives Lith. kartùs âbitterâ from the same root, from earlier âsharp, bitingâ (comparing Lith. kirÌstas âsharpâ, of persons). The Germanic and Baltic forms agree in having o-vocalism of the root; but as Strunk (1976) remarks, o-vocalism of the root can be old in neuter nouns (cf. δόÏÏ , Î³á½¹Î½Ï ), but hardly in an adjective. According to Lamberterie (1990: 349) it was secondarily adopted from a related action noun of the type Ïá½¹ÏοÏ, attested in Lith. karÌtas, OCS kratÑ âonce, timeâ < *kórt-oâ (*âcuttingâ).
This problem is not discussed by Lamberterie.
For example Nussbaum (1998: 147), Widmer (2004: 123â¯ff.).
The same problem applies to the proposal of Benveniste (1969): see below. It is true that κÏá½±ÏÎ¿Ï is often granted by a god in Homer, but that does not make it a magical force.
See e.g. Nussbaum (1998: 154 n. 189), who refers to κÏαÏá½»Ï as an âitem (â¦) that is demonstrably an internal derivative of an acrostatic substantiveâ.
The outcome Åithiráâ is the result of a regular dissimilation of *ÅrÌ¥thiráâ (see Lubotsky 1994: 96, with reference to Narten). I accept the view that aspiration of a preceding stop in Indic could be caused not only by *h2, but also by *h1 (see Lubotsky 2011: 115). The most important piece of evidence is Ved. sphÄyate âto become fatâ < PIE *sph1-oi-e/oâ, which belongs with Hitt. iÅ¡pÄi (3sg. pres.) âeats to satiationâ < *sph1-oi-ei; the color of the laryngeal is proven by OCS spÄti âto succeedâ, Ru. spetâ âto ripenâ, Lith. spÄÌti âto be in timeâ, OE spÅwan âto prosperâ as well as by Lat. spÄs âhopeâ (see Kloekhorst, EDHIL s.v. iÅ¡pÄi-i). The root *speh1- also formed an adjective *sph1-róâ, reflected in Ved. sphiráâ âfatâ and Lat. prosperus âprosperousâ.
Lallot has summarized Benvenisteâs chapter about κÏá½±ÏÎ¿Ï as follows (Benveniste 1969: II, 71): âKrátos ne signifie ni «â¯force physiqueâ¯Â» (iskhús, sthénos) ni «â¯force dââ¯Ã¢meâ¯Â» (alká¸), mais «â¯supériorité, prévalenceâ¯Â», soit au combat, soit à lââ¯assemblée. Ce sens, constant pour krátos, est confirmée par une partie des emplois du dérivé kraterós qui signifie alors «â¯sans égalâ¯Â», notamment au combat. Mais, dans dââ¯autres emplois, kraterós se rapproche, pour le sens, de krataiós «â¯dur, cruelâ¯Â», kratús «â¯durâ¯Â». Lââ¯Ã©tymologie rend compte de cette situation singulière: krátos est à rapprocher de lââ¯i.-ir. kratuâ qui désigne la «â¯vertu (magique) du guerrierâ¯Â»; kratús se rattache à un groupe tout différent, celui de got. hardus qui signifie exclusivement «â¯durâ¯Â».â
âDans les emplois de kraterós coexistent, sans se confondre, les deux notions que les autres termes en kratâ permettent de distinguer: dââ¯une part, la notion abstraite de «â¯prévalence, dominionâ¯Â», de lââ¯autre, la qualité physique de «â¯durâ¯Â».â (1969: II, 81).
The translation âstrength, forceâ is traditionally accepted, see e.g. LSJ (s.v. κÏá½±ÏοÏ: âstrength, might, in Hom. esp. of bodily strengthâ) and LfgrE (s.v. κÏá½±ÏoÏ: âüberlegene Kraft (â¦) beruht im wesentlichen auf Körperkraft, Stärkeâ). For âsuperiority, prevalenceâ, see e.g. Benveniste (1969).
Trümpy (1950: 202â206), Lamberterie (1990: 329).
The lemma runs, in slightly condensed form: âκÏαÏεÏá½¹Ï, epic variant of καÏÏεÏá½¹Ï, 1. strong, stout, mighty, in Homer mostly of bodily strength; with collational notion of stern, harsh, of Hades; 2. of things, conditions, etc. mighty, fierce, hard; 3. of passions strong, vehement; of acts and words, harsh, rough. II. Adv. strongly, stoutly, dashed roughly, refused sternly.â The interpretation of κÏαÏεÏá½¹Ï as á¼°ÏÏÏ Ïá½¹Ï âstrongâ is reflected in the Ancient lexicographical tradition.
Breuil (1989) forcefully translates κÏαÏεÏá½¹Ï as âprévalentâ. In doing so, he gravely oversimplifies the Homeric situation, for instance when speaking of the âdents prévalentsâ of a lion (1989: 34), or translating κÏαÏεÏá½´ á½Ïμίνη as âlutte prévalenteâ (o.c. 35).
When appropriate or necessary, I will adduce examples of other derivatives such as κÏá½±ÏοÏ, but the focus is on determining the basic meaning of the adjective.
Cf. Nordheider, LfgrE s.v. (shortened and slightly modified): âstark, kraftvoll, von Kriegern (gelegentlich Göttern, Tieren), Kräften, Sachen: überlegen, überwältigend, unwiderstehlich, unbändig, oder defensiv unnachgiebig, unbeugsam, hart, fest, gelegentlich mit Konnotation âüberschiessend, hemmungslosâ: zu stark, oder mächtig.â
Similarly, Phobos (personified Rout) is called ἠμα κÏαÏεÏá½¸Ï ÎºÎ±á½¶ á¼ÏαÏÎ²á½µÏ âboth κÏαÏεÏá½¹Ï and fearlessâ (Il. 13.299).
For the Î¼á½³Î½Î¿Ï of κÏαÏεÏá½¸Ï ÎιομήδηÏ, cf. Il. 5.135â143 and 239â256.
Later in the same book, during the episode relating his encounter with Agenor, Achilles and his rage receive the qualification κÏαÏεÏá½¹Ï on three occasions: λύÏÏα ⦠κÏαÏεÏá½µ (Il. 21.542â543), á½Ïὸ κÏαÏεÏοῦ á¼Ïιλá¿Î¿Ï (21.553), λίην Î³á½°Ï ÎºÏαÏεÏá½¸Ï ÏεÏὶ ÏάνÏÏν á¼ÏÏâ á¼Î½Î¸Ïá½½ÏÏν (21.566).
When applied to a champion, the sense of κÏαÏá½³Ï is âto be rampantâ, cf. also Il. 5.175 (Diomedes) and Il. 16.124 (Patroclus). Earlier in book 21, Achilles remarks about his Trojan opponent Lycaon that he did not expect him to appear in battle anymore, because he had taken him captive earlier on and sold him overseas to Lemnos. In Achillesâ words, the sea was apparently not able to keep Lycaon in check (οá½Î´á½³ μιν á¼ÏÏε, 21.58), and he adds: âLet us see whether the grain-growing earth will hold him back, which restrains even the fierce [warrior]â (á¼¥ Ïε καÏá½° κÏαÏεÏόν ÏÎµÏ á¼Ïύκει, 21.63). Once again, only the elemental forces of water and earth are considered capable of restraining a mighty warrior.
Two other telling passages are the following. In Il. 17.206â213, Zeus takes pity on Hector and decides to grant him μέγα κÏá½±ÏοÏ. As a consequence, a violent battle spirit (personified as Ares) enters Hector, and his limbs are filled with fighting spirit and might, á¼Î»Îºá¿Ï καὶ ÏθένεοÏ. Very close to this is Il. 13.59â61, when the two Aiantes are filled with fierce battle rage by Poseidon (Ïλá¿Ïεν Î¼á½³Î½ÎµÎ¿Ï ÎºÏαÏεÏοá¿Î¿). The effect is that they get âlight hands and feetâ. In two short character speeches (Il. 13.73â80), both warriors express this effect in almost identical terms: they are full of eagerness to fight (their Î¸Ï Î¼á½¹Ï or Î¼á½³Î½Î¿Ï is aroused) and their limbs are eager (μαιμώÏÏι). These two passages (and several others) show us that κÏá½±ÏÎ¿Ï is a combination of physical might and mental prowess.
On four occasions: Od. 9.407 and 446 of Polyphemus, 9.476 and 12.210 of his βίη. Cf. also OâSullivan (1990: 14â15).
Nordheider (LfgrE s.v. κÏαÏεÏá½¹Ï) recognizes this use when he speaks of â⦠Sachen ⦠die kraftvoll, wuchtig sind und kraftvoll zupacken, schlagen, treffenâ. More similar examples can be found in Pindar and the tragedians, e.g. καÏÏεÏá½½ÏαÏον Î²á½³Î»Î¿Ï (Pi. Ol. 1.112), ÏεÏÎ¼á½±Î´Î±Ï ÎºÏαÏÎ±Î¹Î²á½¹Î»Î¿Ï Ï (E. Ba. 1096), á¼Îº ÏεÏá½¸Ï Î¼ÎµÎ¸á½³Î½Ïα καÏÏεÏὸν λίθον (E. fr. 1044 Nauck).
The κÏá½±ÏÎ¿Ï which Zeus possesses and wields (οὠÏε κÏá½±ÏÎ¿Ï á¼ÏÏὶ μέγιÏÏον, Od. 5.4) is repeatedly referred to in connection with the destructive physical powers of lightning (e.g. Il. 2.118, 9.25, where his power to destroy the âcrowns of many citiesâ is mentioned), and he is called κάÏÏιÏÏÎ¿Ï in comparison with the other gods (Il. 8.17). His victory over the Titans, which yielded him lasting dominion (κÏá½±ÏοÏ), was ensured by his possession of the thunderbolt. Homer does not qualify the lightning bolt as κÏαÏεÏá½¹Ï, but the post-Homeric evidence suggests that this may be an old collocation: cf. Hes. fr. 343, Pi. Isthm. 8.34, fr. 70b.15 and 155.1, A. PV 922â923, S. OT 201.
Cf. Trümpy (1950: 162): âNeben den Substantiven μένοÏ, δεÏÎ¼á½¹Ï und á¼Î½á½±Î³ÎºÎ· bedeutet κÏαÏεÏá½¹Ï zweifellos âgewalttätigâ. Ebenso sicher aber drückt das Adjektiv neben anderen Wörtern einfach eine Intensitätssteigerung aus und ist mit âgewaltigâ oder âwuchtigâ zu übersetzen.â
From post-Homeric poetry, cf. κÏαÏεÏὸν ÏÎ¸á½³Î½Î¿Ï (B. Dith. 4.40), κÏαÏεÏá¾· ⦠ἰÏÏá½»Ï (B. Epin. 5.21, of Zeusâ eagle), ÏειÏῶν á½Ïο κÏαÏεÏᾶν (Pi. Pyth. 11.18, of the hands of Clytemnestra).
It is reflected in LSJâs translation stout (as in the archaizing English expressions stout resistance and stout supporter); in present-day English it could be rendered as âtoughâ.
Cf. also καÏÏεÏá½° δεÏμά (h. Herm. 409), κÏαÏεÏαá¿Ï á¼Î½ Î³Ï Î¹Î¿Ïá½³Î´Î±Î¹Ï ([A.] PV 168). It is possible that κÏαÏεÏá½µ has a similar meaning when qualifying á¼Î½á½±Î³ÎºÎ· âcoercion, constraintâ (κÏαÏεÏá¿Ï á½Ïâ á¼Î½á½±Î³ÎºÎ·Ï Hes. Th.; Cypr.; Parm.).
As we will see, κÏαÏαιâ functions as an allomorph of κÏαÏεÏοâ (see section 5.2.11).
After Homer: Hermes leads the cows he has stolen ÏῶÏον á¼Î½á½° κÏαÏεÏόν âover solid groundâ so that they will leave no footprints (h. Herm. 354), κÏαÏαίλεÏÏ âconsisting of solid rockâ (A., E., containing Î»á¾¶Î±Ï âstoneâ as its second member), κÏαÏαίÏÎ¹Î½Î¿Ï âwith a hard shellâ, of a turtle (oracle in Hdt. 1.47), καÏÏεÏὸν Ïεá¿ÏÎ¿Ï âsolid defense wallâ (Class. prose).
Cf. âstrong-hoofed, solid-hoofedâ (LSJ) and also Delebecque 1951: 148â150.
Cf. Nordheider (LfgrE s.v. κÏαÏαίÏÎ¿Ï Ï) for different suggestions: âkann auf Trittsicherheit, Ausdauer oder harte Hufe gehenâ (emphasis in original).
Lamberterie (1990: 328â331) stresses that κÏαÏá½»Î½Ï belongs to the technical vocabulary of medicine, metallurgy and warfare, where archaic meanings may be preserved.
Thus also Lamberterie (1990: 346): âqualité physique de force et dââ¯enduranceâ.
The meaning of Ved. dÅ«ráâ âfarâ is easily understood as secondary, e.g. as a qualification of journeys.
Pace de Vaan, who apodictically states that âThe meaning of dÅ«rÄre must have evolved from âbe hardâ to âendure, lastâ.â (EDL s.v. dÅ«rus).
E.g. fortes et duri Spartiatae, Cic. Tusc. 1.43; Ligures, durum in armis genus Liv. 27.48.
E.g. Alcm. fr. 105 Page, Epich. fr. 163, Sophr. fr. 59; for further attestations see LSJ (s.v. κάÏÏÏν) and Forssman (1980: 194 n. 77).
It is commonly accepted that âÏâ was analogically restored in Cretan καÏÏÏν from the stem of the positive καÏÏεÏÎ¿Ï and the neuter καÏÏÎ¿Ï (Bile 1988: 181, following DELG s.v. κÏá½±ÏοÏ; Forssman 1980: 194â195 n. 83, following Lejeune 1972: 111). However, since âονâ is not a comparative suffix, we must ask whether âÏâ does not rather represent the outcome of *âti̯â after a consonant (thus already Seiler 1950: 54). For the outcome of intervocalic *âti̯â in Cretan, cf. the overview in Bile (1988: 145â146); in Gortyn, it is regularly spelled âÏÏâ in the 5th c. BCE, as against âθθâ in the 4th c. and later. It could be objected to this that *âti̯â may not have had an affricate outcome after ârâ if Forssmanâs derivation (1980) of á¼ÏÏÏ from *u̯ert-i̯Š(PIE root *u̯ertâ âto turnâ) is correct. We could assume, however, that the regular outcome *karrÅn of *krÌ¥t-i̯onâ was remade into *kart-i̯onâ in early Cretan on the basis of the positive καÏÏεÏοÏ, and that it was this form which subsequently developed into the attested form καÏÏονâ.
The last-mentioned verb is attested in SEG 35.991 (Lyttos, early 5th c.): pres. καÏÏει (line 3), aor. κα[ÏÏηÏ]αι (line 4â5).
On evidence for the development of *r̥ in West Greek dialects, see section 3.2.
This gloss corroborates the meaning found in Classical prose, where both καÏÏεÏá½¹Ï and á¼°ÏÏÏ Ïá½¹Ï âstrongâ are used to qualify walls and fortresses.
The form κάÏÏεÏον (Alc. fr. 302c.8, cf. also καÏÏε.[ Alc. fr. 119.19) may be an epicism or a borrowing from Ionic (see section 3.3.5). In both instances, the meaning of the context is unclear.
For the reconstruction of the root vocalism of the comparative and superlative, see section 4.1.2.
There are only few possible exceptions, the most notable one being δηÏόν âlong-lastingâ, Arm. erkar âid.â, which is often reconstructed as *du̯eh2-róâ (cf. e.g. Nussbaum 1976: 13). However, in view of obviously cognate formations like Ved. dÅ«rá- âfarâ and Lat. dÅ«rus âhard; harsh; enduringâ, one wonders whether the Greek and Armenian forms can after all be explained from *duh2-róâ, as claimed e.g. by Olsen (2009). See also the elaborate discussion of full grade ro-formations by Vine (2002), who concludes that e-grade ro-formations are âessentially substantivalâ.
The root of μακÏá½¹Ï is sometimes reconstructed as *maḱâ, but this is unsatisfactory because the Greek forms show ablaut. Avestan mas- âbigâ, on the basis of which a reconstruction PIE *mh2ḱâ is sometimes excluded, has been analyzed as a crossover between this root and that of Skt. máhi, Gr. μέγα < *meǵ-h2 (see NIL 478â481), but Kümmel (2018: 165) has convincingly argued that masâ is a devoiced allomorph of mazâ conditioned by the following laryngeal. For the development of word-initial *RHCâ, see Beekes (1988b).
This is argued in more detail in Van Beek 2021a.
It has been suggested (e.g. Nussbaum 1976: 14) that there was a productive rule to use *âroâ instead of *âuâ when the root already contained *âuâ (as in á¼ÏÏ Î¸Ïá½¹Ï, Î»Ï Î³Ïá½¹Ï, á½Î³Ïá½¹Ï). However, this does not account for archaisms in Greek such as εá½Î¸á½»Ï âstraightâ and PIE *sueh2d-uâ âagreeable, sweetâ, nor for the occurrence of PIE *âroâ in roots not containing *âuâ (e.g. *mh2ḱ-róâ).
The occurrence of ÎÏαÏαιâ in epigraphically attested onomastic material may be due to epic influence.
καÏÏαί-Ïοδâ is only attested once in Pindar (Ol. 13.81) in the meaning âbullâ, and in Cretan in the meaning âcattleâ.
Meissner (1998: 244â245) thinks that the names in ÎαÏÏι- and ÎÏαÏιâ are secondary creations on the model of names in Îαλλιâ, but this is difficult to prove, and in my view unlikely. Interestingly, names with a first member κÏαÏεÏοâ are not found, except in Thessaly (Bechtel 1917: 260â261).
Outside of Epic Greek, κάÏÏÎ¿Ï is securely attested only in Simon. fr. 15.1.2 and B. Epin. 5.114, authors where the form may be an epicism. In Hdt. 8.2, κάÏÏÎ¿Ï is only a v.l. (codd. AB) for κÏá½±ÏÎ¿Ï (all other mss.). Trümpy (1950: 202) claimed that κάÏÏÎ¿Ï is the Ionic form, as against Attic κÏá½±ÏοÏ, referring to Bechtel (1921â1924, III: 86) and Smyth (1894: 132). In reality, Smyth and Bechtel merely remark that the variation in the adjective καÏÏεÏá½¹Ï ~ κÏαÏεÏá½¹Ï is also occasionally encountered in personal names, where we sporadically find âκαÏÏÎ·Ï instead of âκÏαÏηÏ. Bechtel mentions only three instances, but in the first two his reading differs from that given by other editors: ÎνεÏικαÏÏÎµÏ (Styra [Euboea], where the edition IG XII,9 56.283 reads ÎνεÏικαÏίεÏ); ÎαÏÏÎ¹ÎµÏ (ibid. 393), and ÎÏ Î¸Ï ÎºÎ±ÏÏÎ¹Î´Î·Ï â¦ hο ÎαξÏÎ¹Î¿Ï (Delos, SGDI 5419), which also occurs on a stone found on Naxos (IG XII Supp. 192.28). For the latter name, a derivation from κείÏÏ âto cutâ cannot be excluded: the literal meaning of the name would be âwho cuts straight incisionsâ. The name ΣÏκαÏÏÎ·Ï which occurs twice on an inscription from Miletus (SEG 13.498) belongs to new citizens that emigrated from Crete, where the regular root shape was καÏÏâ. Finally, a patronymic genitive Î Î¿Î»Ï ÎºÎ±ÏÏÎµÎ¿Ï is encountered once in a 5th c. inscription from Lycia (TAM II, 1184), but it is unknown whence the bearer of this name came. In view of the abundant evidence for names in âκÏαÏηÏ, no conclusions can be based on these examples. I conclude that κÏá½±ÏÎ¿Ï was the only form of the neuter noun in Ionic, as in Attic.
The only instance of καÏÏá½»Î½Ï in Classical poetry is Pi. Ol. 13.95, which may reflect a generalization of the license to substitute Î±Ï for Ïα: cf. the use of θÏá½±ÏÎ¿Ï and θάÏÏÎ¿Ï as variants in Pindar, and the compound καÏÏαίÏοδâ beside the epic first member κÏαÏαιâ.
Only in an epigram ascribed to Simonides (AP 7.344.1).
Cf. LSJ (s.v. κÏαÏεÏá½¹Ï): âEpic form of καÏÏεÏá½¹Ïâ. Note that κÏαÏεÏá½¹Ï is rare after Homer, and that most attestations are found in meters with dactylic rhythm. It is attested in Pindar (3â¯Ã), Bacchylides (5â¯Ã), Corinna (1â¯Ã), and among the tragedians only in Aeschylus (2â¯Ã, both times in lyrical parts). In Herodotus, κÏαÏεÏá½¹Ï only occurs twice, both times in oracles (1.67 and 8.77); Xenophon (Mem. 3.2.2, cf. also Symp. 4.6) quotes the form from Homer. The only attestation in Classical Attic prose seems to be Pl. Tim. 75b, which speaks of a ÏαÏκώδη ⦠καὶ Î½ÎµÏ Ïώδη κÏαÏεÏάν Ïε κεÏαλήν, a âhead ⦠fleshy and sinewy and toughâ. Since classical καÏÏεÏá½¹Ï never refers to physical toughness, this may again be an epicism. The same goes for the compound κÏαÏεÏαύÏην (Pl. Phdr. 253e).
See e.g. Bornemann and Risch (1978: 55): â⦠á¼ÏιÏÏÎ¿Ï (â¦) heisst oft âtüchtigsterâ, βελÏâ bezeichnet besonders den âsittlich besserenâ, κÏειÏ-/κÏαÏâ (vgl. κÏá½±ÏοÏ) eigentlich den âstärkerenâ.â See also Kühner-Gerth I1, 565.
In order to illustrate the claim that the lexical meanings of καÏÏεÏá½¹Ï and κÏá½±ÏÎ¿Ï are different in classical times, I have considered all attestations in Herodotus (Ionic prose) and Thucydides (Attic prose). It appears that in Thucydides, καÏÏεÏá½¹Ï is used in a marginal way: it either refers to places as âstrongâ, referring to their security in a military sense, or it qualifies a battle or military action as âmighty, violentâ. On the other hand, κÏá½±ÏÎ¿Ï either means âpower, rule, supremacyâ, or appears in the phrase καÏá½° κÏá½±ÏÎ¿Ï âwith all oneâs might, in a powerful wayâ or âby forceâ (often qualifying verbs referring to military action). In Herodotus, καÏÏεÏá½¹Ï has a slightly wider usage: apart from the two uses found in Thucydides, we find the phrase καÏá½° Ïὸ καÏÏεÏόν, equivalent to καÏá½° κÏá½±ÏÎ¿Ï in Thucydides; moreover, the claws of the crocodile are referred as âfierceâ, á½Î½Ï ÏÎ±Ï ÎºÎ±ÏÏεÏÎ¿á½»Ï (2.68). On the other hand, κÏá½±ÏÎ¿Ï only appears in the socio-political meaning âpower, rule, commandâ. Thus, the meanings âfierce, mightyâ and âsolid, enduringâ of epic κÏαÏεÏá½¹Ï ~ καÏÏεÏá½¹Ï are only marginally attested for καÏÏεÏá½¹Ï in Classical prose, where all these meanings have in fact been taken over by á¼°ÏÏÏ Ïá½¹Ï (which is post-Homeric). It thus seems that καÏÏεÏá½¹Ï is sometimes used as an archaizing stylistic variant of á¼°ÏÏÏ Ïá½¹Ï; Herodotus, for instance, uses καÏá½° Ïὸ á¼°ÏÏÏ Ïόν âby forceâ and καÏá½° Ïὸ καÏÏεÏόν without any apparent lexical difference.
The derivation of κÏαÏá½»Î½Ï is complicated by the fact that this verb has two meanings. The normal usage in prose is factitive, as historically expected for a verb in âύνÏ: âto harden, make solidâ (e.g. bones, metal), âto strengthen, fortifyâ (a place, position, or dominion). Besides, the tragedians attest an intransitive meaning âto rule, gain controlâ (+ gen. âoverâ). In the opinion of Lamberterie (1990: 328â330), it is difficult to derive κÏαÏá½»Î½Ï âto hardenâ from κÏá½±ÏοÏ, because the neuter does not mean âhardness, solidityâ synchronically in Classical Greek. In the meaning âto ruleâ, however, κÏαÏá½»Î½Ï was certainly derived from κÏá½±ÏÎ¿Ï (cf. Lamberterie 1990: 328 n. 4 with literature).
In the phrase Ïὸ Î³á½°Ï Î±á½Ïε ÏιδήÏÎ¿Ï Î³Îµ κÏá½±ÏÎ¿Ï á¼ÏÏίν Od. 9.393 (on which see below).
Only in κÏαÏá½ºÏ á¼ÏγεÏÏόνÏηÏ, name-epithet formula of Hermes.
Apart from early Greek epic, only in Tyrt. fr. 4.9 and twice in Classical poetry (see the previous section); in Hellenistic poetry, κάÏÏÎ¿Ï is used a few times by Call. and Theoc.
Present stem only; aor. κÏαÏá¿Ïαι may have been avoided for metrical reasons (section 5.2.6).
Cf. among other works Meister 1921, and the illustrative examples in Hackstein 2010.
Furthermore, κÏαÏεÏá½´ á½Ïμίνη âfierce battleâ is mirrored in Classical prose by καÏÏεÏá½´ μάÏη âid.â; Achilles and Diomedes are both qualified as καÏÏεÏá½¹Ï and κÏαÏεÏá½¹Ï.
Including the adverb κÏαÏεÏá¿¶Ï and the comparative κÏαÏεÏá½½ÏεÏοÏ.
Its forms could be used in any foot of the verse, without any noteworthy restrictions on the preceding or following word.
A lexical difference between κÏá½±ÏÎ¿Ï and ÏÎ¸á½³Î½Î¿Ï seems to be implied by the following syntactic difference. While ÏÎ¸á½³Î½Î¿Ï is often accompanied by ὦÏÏεν âevokedâ, á½ÏÎ½Ï Ïαι âarisesâ, κÏá½±ÏÎ¿Ï is always complement to a verb like δῶκεν, á¼Î³Î³Ï άλιξεν âgaveâ. It appears that ÏÎ¸á½³Î½Î¿Ï denotes a powerful, yet human characteristic, whereas κÏá½±ÏÎ¿Ï tends to be more extreme and of heavenly origin. Another difference is that |T (μέγα) ÏθένοÏ, followed by the genitive of a PN, occurs in formulae as a honorific title, just like |P ἱεÏὸν Î¼á½³Î½Î¿Ï (+ PN in genitive); there is no such construction with κÏá½±ÏοÏ. Finally, the meaning of κÏá½±ÏÎ¿Ï is broader: ÏÎ¸á½³Î½Î¿Ï does not mean âpowerâ but merely âforceâ.
Hesiod does attest the genitive κÏá½±ÏÎµÎ¿Ï (Th. 647, where Westâs reading κάÏÏÎµÏ Ï is doubtful). There is no support either for Westâs emendation of κάÏÏÎ¿Ï â¦ á¼ÏγÏν (Th. 710, all codd.) to κάÏÏÎµÏ Ï â¦ á¼Ïγον.
The form κÏá½±ÏÎµÏ could have been used (with epic correption or elision of âi) before vowel-initial, metrically long syllables, but this use is not attested. An irregular scansion is tolerated in οὠÏι κÏá½±Ïεῠγε (Il. 7.142); the same license is found with the dat. sg. of ÏÎ¸á½³Î½Î¿Ï in the first hemistich κάÏÏεῠÏε ÏθένεῠÏε |T (2â¯Ã). The form ÏÎ¸á½³Î½ÎµÏ occurs in the formula |H ÏÎ¸á½³Î½ÎµÏ Î²Î»ÎµÎ¼ÎµÎ±Î¹Î½â (6â¯Ã Il.; Î²Î»ÎµÎ¼ÎµÎ±á½·Î½Ï attested only in this formula), and is also used without metrical irregularity in the phrases ÏÎ¸á½³Î½ÎµÏ Î¼ÎµÎ³á½±Î»á¿³ (2â¯Ã Il.), κάÏÏÎµÏ ÎºÎ±á½¶ ÏÎ¸á½³Î½ÎµÏ ÏÏεÏá½³Ïῳ (Il. 17.322), and δίκá¿Ïá½· Ïε καὶ ÏÎ¸á½³Î½ÎµÏ á¾§ (< *hu̯Åi, Il. 16.542).
With regard to the etymological connection with Germanic âhardâ (see section 5.1.1), it has been stressed that κÏá½±ÏÎ¿Ï has the meaning âhardnessâ in Od. 9.393. However, the context (a simile, Od. 9.391â394) is not unambiguous: á½¡Ï Î´â á½ Ïâ á¼Î½á½´Ï ÏÎ±Î»ÎºÎµá½ºÏ Ïá½³Î»ÎµÎºÏ Î½ μέγαν á¼ á½² ÏκέÏαÏνον / εἰν á½Î´Î±Ïι ÏÏ ÏÏá¿· βάÏÏῠμεγάλα á¼°á½±ÏονÏα / ÏαÏμάÏÏÏνΠÏὸ Î³á½°Ï Î±á½Ïε ÏιδήÏÎ¿Ï Î³Îµ κÏá½±ÏÎ¿Ï á¼ÏÏίνΠ/ á½£Ï Ïοῦ Ïίζâ á½ÏÎ¸Î±Î»Î¼á½¸Ï á¼Î»Î±Ïνέῳ ÏεÏὶ μοÏλῷ. âAs when a bronze smith dips a large axe or an adze into cold water to temper it, and it hisses loudly: for of iron that is the κÏá½±ÏοÏ; likewise did his [the Cyclopsâ] eye hiss around the stake of olive wood.â Here, the poet could just as well refer to the hissing sound of the water vapor as characteristic for, or indicative of, the violent qualities of iron.
Several scholars (e.g. Benveniste 1969, followed by Strunk 1975; Breuil 1989) have stressed that κÏá½±ÏÎ¿Ï is often of a volatile and temporary character: it changes sides between Achaeans and Trojans according to the will of Zeus. They go too far, however, when denying that it may also mean âforce, mightâ: for this meaning, see Lamberterie (1990: 345â346) and OâSullivan (1990: 14â15). Both meanings, âforceâ and âsupremacyâ, must be admitted for Homeric Greek. Besides, the word may denote the lasting authority which allows a leader to control and direct a body of subjects. An appropriate German equivalent is Gewalt, which denotes both an applied physical force and the authority of a ruler or an institution.
For κάÏÏÎµÏ ÏειÏῶν âmight of the hands/armsâ, we may compare certain Homeric instances of θάÏÏÎ¿Ï with a close (but not identical) meaning, and especially the formula θÏαÏειάÏν á¼Ïὸ ÏειÏῶν.
OâSullivan (1990: 14) rightly criticizes Kirkâs view that κÏá½±ÏÎ¿Ï would here denote socio-political power. The sense âbrute forceâ is preserved in Classical Ionic-Attic in the prepositional expression καÏá½° κÏá½±ÏÎ¿Ï âwith all oneâs mightâ (e.g. in Thucydides).
Î´á½¸Ï Î´á½² κÏá½±ÏοÏ, á½ÏÏâ á¼Ïá½±ÏοιÏι ÎºÎµÎºÎ»á½¹Î¼ÎµÎ½Î¿Ï ÎÏ Îºá½·Î¿Î¹Ïιν á¼ÏοÏÏá½»Î½Ï Ïολεμίζειν, αá½Ïá½¹Ï Ïâ á¼Î¼Ïὶ Î½á½³ÎºÏ Î¹ καÏαÏεθνηῶÏι μάÏÏμαι âGive me strength, so that I may call my Lycian fellows and encourage them to do battle, and that I myself may fight over the dead body [of Sarpedon]â (Il. 16.524â526); εἰ Î³á½°Ï á¼Î¸á½µÎ½Î· δοίη κάÏÏÎ¿Ï á¼Î¼Î¿á½·, βελέÏν δâ á¼ÏεÏύκοι á¼ÏÏήν âMay Athena give me strength [to continue fighting], and may she ward off the storm of arrowsâ (Il. 17.561â562). In both cases, the power to persevere is granted by Athena. Cf. further the phrase κάÏÏÎ¿Ï Ïε βίη Ïε at Od. 4.415 (about the tenacious force which Menelaus has to apply in order to control the shape-shifting Old Man of the Sea), as well as Od. 3.370, where κάÏÏÎ¿Ï denotes the stamina of horses.
In Homer: Il. 16.54 (Agamemnon over the Achaeans), Od. 1.359 and 21.353 (Telemachus over his household), 5.4 (Zeus), and 11.353 (Alcinous over the Phaeacian δá¿Î¼Î¿Ï). I take the verse Ïοῦ δâ á¼Îº ΦαιήκÏν á¼ÏεÏαι κάÏÏÎ¿Ï Ïε βίη Ïε (Od. 6.197) to mean âon him (Alcinous) the force of the Phaeacians dependsâ, reading á¼Îº ⦠á¼ÏεÏαι with tmesis and noting that the phrase κάÏÏÎ¿Ï Ïε βίη Ïε has the same meaning âviolence and forceâ also at Od. 4.415. This distribution also holds good for Hesiod, provided that one follows the evidence of the mss. for the genitive κÏá½±ÏÎµÎ¿Ï in Th. 647 (the lasting dominion of Zeus), rather than emendating to κάÏÏÎµÏ Ï (pace Westâs edition, who bases this reading on the reading κα[ in one papyrus).
The situation is slightly more complex. There is no trace of κÏαÏεÏá½¹Ï ~ καÏÏεÏá½¹Ï in the meaning âpowerful, in controlâ in the Iliad, but there are two possible instances in the Odyssey (14.116 and 15.534). It seems that καÏÏεÏá½¹Ï has encroached in Ionic upon κύÏιοÏ, the normal word in this sense in Classical Attic and the lexeme inherited from Proto-Greek.
Further attested in Hes. Op. 206, Scut. 321, 419, 461, Stes. fr. 40.24 Page, Ibyc. fr. S199.6 Page.
The only evidence for âκαÏÏÎ·Ï comes from epigraphically attested personal names, most of which are of Cretan or Theran origin. See section 5.2.3 above.
The vocalism of Aeol. á¼ÏικÏá½³Ïει and κÏá½³ÏηÏαι may suggest that the pre-form of κÏαÏá¿Ïαι never contained a syllabic liquid. Therefore, using κÏαÏá¿Ïαι would require the application of the muta cum liquida license, which in Homer is not yet very frequent with original plosive plus liquid onsets (see chapter 6).
As in Hom. καÏÏá½»Î½Ï for κÏαÏύνÏ, which could be reanalyzed as derived from κάÏÏÎ¿Ï ~ κÏá½±ÏÎ¿Ï after the adjective κÏαÏá½»Ï had become obsolete. See below.
For instance, LfgrE has one single entry âκÏαÏá½»Ï (κÏείÏÏÏν, κάÏÏιÏÏοÏ)â; cf. also Chantraine (1958: 255â256).
See section 4.1.2 for the reconstruction of a full grade root *kret-istoâ, and chapter 6 for the avoidance of McL scansion before original full vowels.
Cf. Cunliffe 1924 (s.v. κάÏÏιÏÏοÏ).
For a discussion of this passage, see section 5.1.2 above.
In Il. 1.266â268, the Lapiths and the Centaurs both receive the same epithet κάÏÏιÏÏοι in a description of their war: The Lapiths were the fiercest mythical human warriors, the Centaurs were the fiercest non-human mortal creatures. For κάÏÏιÏÏοι ⦠á¼Î½Î´Ïῶν applied to the Lapiths (cf. also κάÏÏιÏÏον ⦠á¼Î½Î´Ïα Il. 7.155; κάÏÏιÏÏÎ¿Ï â¦ á¼Î½Î´Ïῶν Il. 9.558), cf. the phrase καÏÏεÏá½¸Ï á¼Î½á½µÏ (Od. 4.242, 4.271, 20.393).
Cf. âthe best course is to flee from herâ (tr. Dimock 1995); similarly LfgrE s.v. κÏαÏá½»Ï.
This is the only occurrence in Homer of the prose adjective ÏÏοδÏá½¹Ï âvehement, impetuous, fierce, energeticâ. It is hard to tell why μάλα ÏÏοδÏá¿¶Ï was preferred here to μάλα κÏαÏεÏῶÏ, which would fit the meter and occurs several times in Homer; there may have been a subtle semantic or stylistic difference.
Cf. Cunliffe 1924 (s.v. κÏείÏÏÏν). Only the adverbially used neuter κÏεá¿ÏÏον, attested just once (Od. 6.182), has the bleached meaning âbetterâ that is also found in Classical Greek.
In one passage in the Odyssey, κÏείÏÏÏν has the meaning âhaving right of say (over)â (+ gen.): μá¿ÏÎµÏ á¼Î¼á½µ, Ïόξον μὲν á¼Ïαιῶν οὠÏÎ¹Ï á¼Î¼Îµá¿Î¿ κÏείÏÏÏν, ᾧ κâ á¼Î¸á½³Î»Ï δόμεναί Ïε καὶ á¼ÏνήÏαÏθαι (â¦) Ïῶν οὠÏá½·Ï Î¼â á¼á½³ÎºÎ¿Î½Ïα βιήÏεÏαι, αἴ κâ á¼Î¸á½³Î»Ïμι καὶ καθάÏαξ ξείνῳ δόμεναι Ïάδε Ïόξα Ïá½³ÏεÏθαι âMother of mine, as for the bow, no one of the Achaeans has right of say over me (οὠÏÎ¹Ï á¼Î¼Îµá¿Î¿ κÏείÏÏÏν), that I may give or deny it to whoever I wish to: (â¦). No one of them shall force me against my will, even if I should wish to give this bow once and for all to the stranger to carry it away with him.â (Od. 21.344â345 and 348â349). Here, κÏείÏÏÏν seems to function as the positive of an adjective corresponding to the abstract κÏá½±ÏÎ¿Ï in the sense âpower, authorityâ (G. Gewalt) and κÏαÏá½³Ï âto be in chargeâ. In Class. Attic, κÏείÏÏÏν may also function as a positive, e.g. κÏείÏÏÏν αá½Ïοῦ âmaster over oneselfâ.
LfgrE gives the following translations: âaktuell sich im Zweikampf als der stärkere erweisen ~ siegen (â¦); dauernd stärker (â¦); mächtiger (â¦); mit Angabe des Bereichs überlegen in/an (â¦); besser (â¦).â On the basis of Il. 3.71 νικήÏῠκÏείÏÏÏν Ïε γένηÏαι, Trümpy asserted that victory is a prerequisite for being κÏείÏÏÏν: â⦠für κÏείÏÏÏν ist ein Sieg Voraussetzungâ (1950: 205â206). However, this formulation fails to take into account that κÏείÏÏÏν never qualifies actual victors in Homer (these are referred to with the ptc. νικήÏαÏ, and their victory with νίκη). I would therefore modify Trümpyâs words as follows: âfür κÏείÏÏÏν ist ein gedachter Sieg Voraussetzungâ.
Another illustrative case is Il. 19.216â219, where Odysseus addresses Achilles: ὦ á¼Ïιλεῦ Πηλá¿Î¿Ï Ï á¼±á½² μέγα Ïá½³ÏÏαÏâ á¼Ïαιῶν, κÏείÏÏÏν Îµá¼°Ï á¼Î¼á½³Î¸ÎµÎ½ καὶ Ïá½³ÏÏεÏÎ¿Ï Î¿á½Îº á½Î»á½·Î³Î¿Î½ ÏÎµÏ á¼Î³Ïει, á¼Î³á½¼ δέ κε Ïεá¿Î¿ νοήμαÏá½· γε ÏÏοβαλοίμην Ïολλόν, á¼Ïεὶ ÏÏá½¹ÏεÏÎ¿Ï Î³ÎµÎ½á½¹Î¼Î·Î½ καὶ Ïλείονα οἶδα, âAchilles, son of Peleus, by far the best of the Achaeans, stronger are you than I am and better not a little with the spear. But I would beat you by far in counsel, because I was born earlier and know more.â As Breuil (1989: 44) notes, â⦠la prévalence dââ¯Achille sur Ulysse ne sââ¯actualise que de manière indirecteâ. For the same typical use of κÏείÏÏÏν, cf. also Il. 20.334 and Il. 23.578. When the Î½á½¹Î¿Ï âmindâ of Zeus is qualified as κÏείÏÏÏν (Il. 16.688, 17.176), the idea is that his will (the ÎÎ¹á½¸Ï Î²Î¿Ï Î»á½µ) will prevail eventually, no matter what another god or a human being may devise.
A different question is to what extent κÏείÏÏÏν and κÏá½±ÏιÏÏÎ¿Ï are part of the same paradigm in Classical Attic (it is normally thought that both forms belong to the positive á¼Î³Î±Î¸á½¹Ï).
Cf. also κÏαÏεÏá½½ÏεÏον á¼Î»Î»Î¿ κεÏÎ±Ï Î½Î¿á¿¦ â[no] fiercer [weapon] than lightningâ (Hes. fr. 343.8).
ÏίδηÏοÏ, á½ ÏÎµÏ ÎºÏαÏεÏá½½ÏαÏá½¹Ï á¼ÏÏιν (Hes. Th. 864).
After Homer, á¼ÎºÎ±ÏÏύνανÏο is attested once in Hesiod (Th. 676) and in two epic fragments (Antimachus fr. 42.1 Wyss, also fr. 64.4 Lloyd-Jones & Parsons). Furthermore, the active present καÏÏύνειν is attested once in Pindar (Ol. 13.95). It is difficult to pinpoint the meaning of καÏÏá½»Î½Ï with absolute certainty on the basis of one single formula. Strunk (1975) points out that the line preceding á¼ÎºÎ±ÏÏύνανÏο ÏάλαγγαÏ, in all three Homeric attestations, depicts an army leader arousing the fighting spirit (á¼Î»Îºá½µ) of his men: the consequences of the leaderâs call âbestehen nicht nur im blossen zusammenrücken der ÏάλαγγεÏ, sondern auch darin, dass diese neuen Anlass zum Kampf sehen und frischen Mut fassen.â (1975: 273). The idea that Ïá½±Î»Î±Î³Î³ÎµÏ ÎºÎ±ÏÏεÏαί conveys the image of a solid and massive wooden log (Benveniste 1969, II: 80, Lamberterie 1990: 332) is in my view less likely.
The Ionic vernacular form κÏαÏá½»Î½Ï may have been derived from the u-stem adjective κÏαÏá½»Ï after the vocalization of the syllabic liquids and the subsequent spread of the allomorph κÏαÏâ (see section 4.3.3). Lamberterie (1990: 328) stresses that the meaning of κÏαÏá½»Î½Ï in later prose is âto harden, consolidate, reinforceâ in a material sense, but it also had the sense âto strengthen oneâs defense / fortificationsâ (cf. LSJ s.v.), which is close to the Homeric meaning.
Strunk (1975: 296) gives the same derivation, but does not account for the peculiar variations κÏαÏâ ~ καÏÏâ and θÏαÏâ ~ θαÏÏâ. Remember that θÏαÏá½»Ï ârecklessâ has not only the wrong vowel slot, but also a different lexical meaning as compared to θαÏÏύνÏ.
In the words of Strunk (1975: 273â274), âDie von Benveniste geleugnete semantische Brücke zwischen κÏá½±ÏÎ¿Ï und καÏÏύνεÏθαι (κÏαÏá½»Ï) ist damit zumindest in der homerischen Sprache (â¦) greifbar: mit á¼Î»Îºá½µ, »kämpferischer Gesinnung«, haben beide Wörter zu tun.â
In pre-Classical poetry only in Aristeas Epicus fr. 5.2, Empedocles fr. 4.4; Protagoras fr. 9, Epich. fr. 113.1.13.
The similarities between κάÏÏα and Lat. certus âcertainâ (adv. certe âcertainly, surelyâ) are superficial: the Latin word derives from *kritóâ âsifted, distinguishedâ (cf. EDL s.v. certus).
Nussbaum (1976: 122 n. 37) proposes to reconstruct *ânÌ¥t, the neuter form of a Caland nt-adjective. For a discussion with further literature, see Meissner (2006: 63â64), who prefers *âh2. For the Homeric instances, see Risch 1974: 363.
According to Meissner (2006: 68â69), names in âκÏεÏá½µÏ / âκÏá½±ÏÎ·Ï are recent creations because no examples are found in Homer or Mycenaean.
Alternatively, one could think of a proportion μάλα : μαλεÏá½¹Ï = X : καÏÏεÏá½¹Ï, as I did in Van Beek 2013: 140. However, this would require that μαλεÏá½¹Ï was current in spoken Ionic-Attic in the meaning âvehementâ when the analogy was applied. This is not evident, as μαλεÏá½¹Ï is exclusively poetic and its exact meaning is debatable. Cf. GEW s.v. μαλεÏá½¹Ï, âwegen der unbestimmten Bedeutung ohne überzeugende Etymologieâ. Blanc (DELG, Supp. s.v. μαλεÏá½¹Ï) has argued that μαλεÏá½¹Ï originally referred to the brilliant splendor of light, and that the form arose by dissimilation from *mar-eroâ, with the root of μαÏμαίÏÏ âto glitterâ. This seems unlikely to me, but if it is correct, it would imply that μαλεÏá½¹Ï cannot play a role in accounting for κάÏÏα.
In this overview, I leave aside the precursors of κÏαÏÎ±Î¹á½¹Ï and κÏαÏαιâ because these forms were limited to Epic poetry.
The same holds for the argument that Class. κÏαÏá½»Î½Ï in the sense âto make firm, hardenâ cannot have been derived from κÏá½±ÏοÏ.
See Lamberterie (1990: 337); to his list of examples proving this equivalence, I would add Od. 18.383, where κÏαÏÎ±Î¹á½¹Ï occurs in a speech by Odysseus. Still in disguise as a beggar, he addresses the suitor Eurymachus and warns him that Odysseus would beat him in any contest of endurance, be it in mowing the grass from morning till evening, in ploughing a field all day long, or in full war. Thus, in the verse καὶ Ïού ÏÎ¹Ï Î´Î¿Îºá½³ÎµÎ¹Ï Î¼á½³Î³Î±Ï á¼Î¼Î¼ÎµÎ½Î±Î¹ ἠδὲ κÏαÏÎ±Î¹á½¹Ï (Od. 18.383) âyou think you are some big and tough guyâ, κÏαÏÎ±Î¹á½¹Ï refers to the physical condition (fitness) which the suitor Eurymachus is lacking, according to Odysseus. Thus, apart from the sense âimpetuous, fierceâ (which is likely in the other Homeric instances), κÏαÏÎ±Î¹á½¹Ï also means âsteadfast, toughâ on at least one occasion. At least the poet of this line considered κÏαÏÎ±Î¹á½¹Ï and κÏαÏεÏá½¹Ï ~ καÏÏεÏá½¹Ï to be semantically interchangeable.
Risch (1974: 74), Nagy (1999: 85â89 and 349â354), Lamberterie (1990: 337â343), Meissner (2006: 62â63).
This toponym may have originally been an epithet denoting a wide or flat area (e.g. *ÏλάÏαια Ïá½½Ïα). Of course, the normal feminine ÏλαÏεá¿Î± has analogical âεá¿Î±.
In Van Beek 2013, section 5.3 I have argued that κÏαÏεÏá½¹Ï ~ καÏÏεÏá½¹Ï corresponds directly to Ved. Åithirá- âlooseâ. If this is correct, the root did have a final laryngeal, but its form was *ḱreth1- in view of the âεâ in Ion.-Att. καÏÏεÏá½¹Ï. It is not possible to start from *ḱreth2â and assume a reshaping of *καÏÏαÏá½¹Ï to καÏÏεÏá½¹Ï (for instance after ἱεÏá½¹Ï, which was close in meaning), as Cretan has καÏÏεÏÎ¿Ï beside ιαÏοÏ.
The only clearly Homeric case mentioned by Lamberterie is á¼Ïαá¿ÏÎ¿Ï âcompanionâ, which he views as a backformation from á¼ÏαίÏη. Here, however, metrical pressure may have been at work, since the older form of the masculine seems to be á¼ÏαÏοÏ. Moreover, it remains unclear whether an older athematic feminine ever existed.
See Risch (1974: 219) and the table in Meissner (2006: 19) for a few more uncertain items.
From *dnÌ¥s-iâ.
From *u̯rÄhiâ beside adv. *u̯rÄha, see Wackernagel (1897); ῥηá¿Î´Î¹Î¿Ï (already Homeric) may also be an original compound with first member *u̯rÄhiâ, as I argue in Van Beek 2020.
Name of the skin disease erysipelas; cf. perhaps also á¼ÏÏ Ïίβη ârust, red blightâ. Note the unproductive assibilation of /th/ (for which cf. Myc. ko-ri-si-jo /Korinsios/ âfrom Corinthâ).
The use in appellatives is post-Homeric, but the same first member is already attested in PNâ¯s such as ÎεÏÏίλοÏÎ¿Ï (Hom.+) and perhaps also in ÎεÏÏá½·ÏÎ·Ï (see section 2.3.1).
Meissner (1998: 244â246) claims that names with ÎÏαÏιâ, ÎαÏÏιâ are late, analogical formations.
Contrast Meissnerâs view (2006: 22) that θεÏÏιâ might be an archaism to be compared with Av. darÅ¡iâ.
Note that adjectives in âá½»Ï do not change their shape when appearing as a first compound member: cf. e.g. βαÏύκÏÏ ÏοÏ, Î²Î±Î¸Ï Î´á½·Î½Î·Ï, θÏαÏÏ Îºá½±ÏδιοÏ.
See also Meissner (1998: 244â246).
The oldest attestation is the name of an Achaean victor in Olympia (SEG 22.345, appr. 600â¯BCE). Further attested (mostly late) in ÎÏαÏÎ±Î¹Î¼á½³Î½Î¿Ï SEG 19.108 I.117 (Attica, cf. SEG 23.124.2), ÎÏαÏÎ±Î¹Î¼á½³Î½Î·Ï IG V,1 127.4 and 211 II.34 (Laconian), also IG V,2 419.8 (Arcadian, 2nd c.), Îá½Î´]Î·Î¼Î¿Ï ÎÏαÏÎ±Î¹Î¼á½³Î½Î¿Ï á¼ÏεÏÏÎ¹Îµá½»Ï IG XII,9 91.4 (Euboea). Cf. also ÎαÏÏαιμ̣έ̣νη[Ï], IG XII (Supp.) 312 III.31 (Tenos, Ionic Cyclades, 2nd c.).
ÎÏαÏÎ±á½·Î²Î¹Î¿Ï IG XI,2 287 A.146 passim (Delos). The form ÎαÏÏÎ±á½·Î²Î¹Î¿Ï (with âαÏâ) is also attested as the name of a Cretan in Miletus (Bechtel 1917: 256).
Vita Herodotea 14.9.
The exception is λύκοι κÏαÏεÏá½½Î½Ï ÏÎµÏ á¼ Î´á½² λέονÏÎµÏ (Od. 10.218), where it means âwith fierce clawsâ.
In κÏαÏαιÏίνοιο âhard-shelledâ (oracle in Hdt. 1.47, hexameter), the allomorph κÏαÏαιâ is used before a heavy syllable starting with a single consonant, but note that the prosodic behavior of initial á¿¥â varies. The compound κÏαÏαίλεÏÏ âconsisting of hard rockâ (trag.), containing Î»á¾¶Î±Ï âstoneâ as a second member, is a relatively recent formation (note the application of quantitative metathesis).
For a critique of the assumption of liquid metathesis in Cretan, see chapter 3.
The evidence consists of ÎÏαÏ-εÏμοÏ, ÎÏαÏ-ιÏÏιδαÏ, ÎÏαÏι-δημοÏ; ÎαÏÏι-δαμαÏ, ÎαÏÏι-νικοÏ, ÎαÏÏι-ÏÎ¸ÎµÎ½Î·Ï (cf. Bechtel 1917: 256). Of these names, ÎαÏÏιÏÎ¸ÎµÎ½Î·Ï is attested in Cyrene, while ÎαÏÏÎ¹Î½Î¹ÎºÎ¿Ï and ÎαÏÏÎ¹Î´Î±Î¼Î±Ï are from Thera, and it is exactly in these two dialects that Î±Ï is the regular reflex of *rÌ¥. ÎÏαÏÎ¹Î´Î·Î¼Î¿Ï is attested in Ionian territory (Erythrae). Obviously, ÎÏαÏιâ may have the root shape of κÏá½±ÏοÏ, or alternatively it may be an epicism.
âdas Fehlen von κÏαÏιâ bei Homer [ist] nicht auf metrische Gründe zurückführbar. (â¦) Ein καÏÏιâ bzw. κÏαÏιâ wäre metrisch vielseitig verwendbar. Sein vollständiges Fehlen ist also auffällig.â (Meissner 1998: 245).
In Epic Greek, the replacement of *krÌ¥tiâ by *krÌ¥teroâ must have occurred before the vocalization of the syllabic liquids, which would have altered the metrical structure of *krÌ¥tiâ. There is no trace of καÏÏιâ in Greek poetry.
Other words which contain final âαιόâ are: á¼ÏÎ±Î¹á½¹Ï âthin, slenderâ (no etymology), Î²Î±Î¹á½¹Ï âsmall, slightâ (no etymology), γηÏÎ±Î¹á½¹Ï (probably a younger variant of γεÏαιόÏ), á¼Î»Î±Î¹á½¹Ï (Hsch., a by-form of á¼ Î»Îµá½¹Ï âcrazedâ), Î»Î±Î¹á½¹Ï âleftâ = Lat. laevus etc. (an old formation < PIE *leh2i-u̯oâ, cf. Î´ÎµÎ¾Î¹á½¹Ï âon the rightâ), ÏÎºÎ±Î¹á½¹Ï âleft, westernâ = Lat. scaevus (likewise old, < PIE *skeh2i-u̯oâ). One may also perhaps compare the ethnic á¼Ïαιοί, on which see Nagy (1999: 349â354). The common classical adjective á¼ÏÏαá¿Î¿Ï âancientâ does not occur in early epic, except in Hes. fr. 322. As the accent shows, this form was derived productively from the noun á¼ÏÏá½µ âbeginningâ; its suffix goes back to PGr. *âi(i̯)oâ.
As recognized by DELG (s.v. δήν), there is no reason to doubt this reconstruction of δηναιόÏ, because its meaning is consistently âlong-livedâ in both Homer and Aeschylus. Frisk (GEW s.v. δήν) has issues with this etymology, presumably because of the use of an adverb as a first member. He thinks that Î´Î·Î½Î±Î¹á½¹Ï could be an artificial formation influenced by ÏÎ±Î»Î±Î¹á½¹Ï and á¼ÏÏαá¿Î¿Ï (âvielleicht sogar nach ihrem Vorbild direkt aus δήν erweitert seinâ), but á¼ÏÏαá¿Î¿Ï cannot be compared in view of its different accentuation. Note that Homer uses δήν predicatively in nominal sentences (e.g. οá½Î´á½² Î³á½°Ï â¦ Î´á½´Î½ ἦν âfor he [Lycurgus] did not live longâ, Il. 6.130â131).
As Chadwick (1976) has shown, ÏÎ±Î»Î±Î¹á½¹Ï originally referred to the recent past: âthe length of a period does not normally extend beyond a lifetime, and may be much shorterâ. In the Mycenaean tablets, pa-ra-jo qualifies wine and is used in opposition to ne-wo âyoungâ.
Nothing excludes that γεÏÎ±Î¹á½¹Ï (30â¯Ã Hom.) is also an old formation. An adverb *gerai is not attested, but it would not be unthinkable that the precursor of γεÏÎ±Î¹á½¹Ï was somehow reshaped under the influence of ÏαλαιόÏ, in view of their close meanings. Alternatively, one might consider a compound PGr. *ger-aiu̯-óâ âbelonging to an old lifetime / generationâ.
A similar scenario had already been proposed in DELG (s.v. κÏá½±ÏοÏ). I disagree with Nagyâs proposal to reconstruct κÏαÏÎ±Î¹á½¹Ï as the feminine of a compound *krÌ¥tai-u̯i(H)âi(e)h2 âhaving strong forceâ (Gr. á¼´Ï, Lat. vis). The pre-form would first have lost the laryngeal in its second member, and then removed its suffixal ablaut to yield *kratai-u̯i̯Ä. This would, finally, have given rise to a secondary masculine form. Apart from the fact that assuming laryngeal loss in a compound is slightly ad hoc, the objections to the other two points are the same as for Rischâs derivation from the feminine of a u-stem adjective discussed above.
An objection to reconstructing a compound is that κÏαÏÎ±Î¹á½¹Ï has a motional feminine κÏαÏαιή, but there are other archaic-looking compounds in Homer with a motional feminine (cf. e.g. νὺξ á¼Î²Ïá½¹Ïη). Furthermore, it cannot be excluded that κÏαÏÎ±Î¹á½¹Ï was reanalyzed as a simplex early on, as all other adjectives in âÎ±Î¹á½¹Ï (ÏαλαιόÏ, γεÏαιόÏ, λαιόÏ, ÏκαιόÏ) were simplicia.
Cf. also the comment by Heubeck in Comm. Od., ad loc.: â[W]e would expect to find a personal subject for á¼ÏοÏÏÏá½³ÏαÏκε, e.g. a mythical figure ÎÏαÏαιá¿Ï, homonymous with the mother of Scylla.â
For this translation, see the discussion of the superlative κάÏÏιÏÏον above (section 5.2.7).
Given the new interpretation proposed for ÏÏ Î³á½³ÎµÎ¹Î½ κάÏÏιÏÏον âflee with all your mightâ (section 5.2.7), it would be attractive to view ÎÏá½±ÏαιÏÏ as a personified force which grants impetus to the boat, just like κÏαÏαιá¿Ï in Od. 11.597 is a force which accelerates a stone. Therefore, the idea that ÎÏá½±ÏαιÏÏ was Scyllaâs mother (lines 125â126) could conceivably be due to a post-Homeric reinterpretation of the passage; the lines may have been added after the meaning of the word (and name) had been forgotten.
In a number of Homeric instances, á¼´Ï refers to the impetus of natural forces (wind, river). In my view, a translation âwhose force is κÏαÏεÏá½¹Ïâ would be preferable: as we have seen, κÏαÏαιâ functions as a relic allomorph of κÏαÏεÏá½¹Ï.
In fact, all analyses of the form as a compound (or noun phrase) with á¼´Ï âforceâ suffer from the same problem.
In Van Beek 2013, I presented this solution as originating with Nagy (1999: 349), without knowing that Wackernagel had made the same proposal a century before. Nagy does not refer to Wackernagel either.
Moreover, if the analysis of κÏαÏÎ±Î¹á½¹Ï as an original compound is correct (see the previous section), this problem disappears.