Chapter 4. Anger’s History: Κότος and Etymology

{88|89} With Calchas as my guide, I have shown kótos to be an anger, that, in the Iliad, relates to the fall of Troy, while in the Odyssey it calls to mind Odysseus’s vengeance against the suitors. The method, thus far, has been to focus on the diction of Homeric poetry, its phraseology and narrative contexts. In the remaining two chapters of Part I, I turn to two methods that can illuminate this area in a manner external to the text. In this chapter, the comparative method will be used to help throw light from cognate languages on this species of Homeric anger, and in the next chapter, I apply to these problems the notion of the feud as developed in the social sciences. I turn here to the comparative method in order to produce evidence for the semantic force of this important word that, when placed beside our findings from the Iliad and the Odyssey, indicates how deeply rooted is Calchas’s definition of kótos.
The evidence from historical analysis points to a meaning consistent with Calchas’s explanation of kótos, but before the analysis presented here, there was little systematic support from Greek to the effect that kótos means anything other than “wrath” (or “grudge”) as opposed to “mere anger.” Now it is possible to view the cognates from five Indo-European branches (Celtic, Germanic, Anatolian, Indo-Iranian, and Slavic) as providing corroboration for the profound implications of Calchas’s definition, namely, that kótos is a term with a specific denotation within the terminology of conflict displayed in the Homeric poems and descended from the earliest strata of the tradition that produced those poems. My own investigation of kótos, we will find, strengthens the case for the connections between kótos and the cognates suggested in the literature.
Both Frisk and Chantraine propose connections for kótos with certain Celtic and Germanic forms referring to conflict, specifically to “battle, combat, etc.,” [1] found, for example, in Catu-rīges (Gaulish name), Hadu-brand (Old High {89|90} German name); [2] cf. Old Norse “battle.” [3] The first element of these compounds presents evidence for a Celtic and Germanic root *katu- “fight, quarrel.” In Catu-rīges, the second element of this name (-rīg-) has cognates in other Celtic languages, in Latin, and in Sanskrit (cf. Old Irish , Lat. rēx, Skt. rāj-, all “king,” or “ruler” [IEW 855]) which connection reminds us that for Calchas, kótos has a special connection to the basileús or ánax. [4]
Sanskrit ṡát-ru- “enemy” (with -ru- suffix) is connected by Mayrhofer to the Celtic and Germanic forms just cited, and he also confidently compares Old Church Slavic kotora (translating mákhē). [5] However, Mayrhofer finds Greek kótos questionable as a comparandum, presumably on the semantic grounds that kótos does not mean “battle.” Mayrhofer here is in harmony with Frisk [6] among others who find Greek kótos difficult to connect with a word for “battle.” There is indeed room to question—however attractive the formal connections—how a word for “battle” is related to a word for “anger,” so that the semantic connection can be accounted for in a systematic fashion. In this regard, the Anatolian branch of the Indo-European language family adds important evidence, specifically Hittite kattawatar and Cuneiform Luvian kattawatnalli. Both forms will be important as we delve further into two problems related to the cognates of kótos.
First, the morphological variation between the Greek, Germanic, Celtic, Slavic, Anatolian, and Indo-Iranian terms can be reconciled, if we bring into play the notion of a Caland-system root with a series of suffixes including stative verbal forms with -e- (cf. kot-ḗ-eis, ke-kot-ē-óti), and nominal forms with -u- (e.g., Old Irish cath), -ru- (e.g., Skt. ṡát-ru-), and a neuter -es- stem noun. [7] In Greek, the verbal forms survive and are still productive through the conservative platform of Homeric epic, but the noun is reanalyzed, so that an inherited neuter noun in -es- was reinterpreted as thematic, perhaps in the epic tradition itself, thereby producing a thematic back-formation based on the o-grade of the stative perfect. [8]
Thus, it may not be best to see the verbal forms related to kótos as recent denominative formations, especially since at least some of the verbal forms point to old inherited classes consistent with the basic meaning of such a verb: for example, the present stem looks to continue an iterative category inherited from Indo-European, with expected o vocalism of the root, and stem in -eip-; [9] in addition to the present forms, the perfect forms also belong to a class that has vital claims to antiquity; note that the term itself ceases to be productive after the archaic period in favor of other terms for anger (including khólos, but also orgḗ, and thumós). [10]
Second, let us turn to the Anatolian forms, since they have given some problem to those seeking cognates for kótos.
(1) Consider here Hittite kattawatar in the next example from Mursilis’ prayer to the sun goddess (Lebrun 1980, 162-63 [A 49-53]):
ki-nu-na a-ra-ah-zé-na-an-te-eš [ud-ni-an-t]e-eš
hu-u-ma-an-te-eš KUR URUKÙ.BABBAR-ti {90|91}
[w(a-al-h)] a-an-ni-eš-ki-u-wa-an da-a-ir.
na-at ANA dUTU URUA-ri-in-na kat-ta-wa-a-tar nam-ma
And now the neighbors of foreign lands all of them took to attacking Hatti. Thus may it be once more a kattawatar for the Sun-goddess of Arinna. [11]
Now the Hittite word kattawatar refers to the anger of the goddess on seeing the neighboring cities attacking Hatti, and it occurs in a prayer of great antiquity [12] performed in a context of social crisis. [13] Lebrun’s translation of kattawatar as “un motif de vengeance” suggests a value similar to the one that we have found for kótos in the Greek contexts. Further this passages points to a long-standing relationship (here between countries, “neighbors of foreign lands”) of the kind that we will see (in the next chapter) is required for a feuding relationship. Moreover, Mursilis continues his prayer through the notions of anger that he thinks come into play with respect to the goddess’s kattawatar (Lebrun 1980, 163 [A 54-56]):
nu ANA DINGIR.MEŠ ku-iš kar-pí-iš kar-tim-mi-ya-az
ku-iš DINGIR.MEŠ-na-aš U-UL na-ah-ha-an-za na-aš-ta le-e
a-aš-ša-u-e-eš i-da-a-la-u-wa-aš an-da har-kán-zi
This wrath, violent anger for the gods, it is not respectful of the gods; let not the good and evil perish together. [14]
After the mention of the goddess’s vengeance, this passage shows two other words for anger in Hittite, both representing a more generalized form of wrath, karpís and kartimiyaz. The order is iconic, with the word for the divine revenge heading a list of anger-terms. Recall here the way that kótos could be used in clusters with other Homeric terms for anger: ei mḗ tis theós esti kotessámenos Trṓessin / hirōn mēnísas; khalepè dè theoû épi mênis, “Unless it is a certain god that has kótos with the Trojans, being enraged over sacrifices; difficult is the rage of a god” (Il. 5.174-78); mḕ pṓs toi metópisthe kotessámenos khalepḗnēi / hoútō nûn apópempe, Diòs d’ epopízeo mênin “lest with kótos in the future, he (Zeus) take it ill, give now (to Odysseus) passage home, and take heed of the rage of Zeus” (Od . 5.146-47). [15] In both Homeric cases, kótos seems to be the more grave form of anger, and in both cases, a god’s anger is the focus. In Mursilis’ prayer, the sun goddess’s kattawatar, a marked term, is the central request, with other (unmarked) more conventional terms for anger following in its wake.
(2) Now the form kattawatar is an abstract with the suffix -atar, attached to the root katu-, like idāluātar (“evil” based on idālu-) “bad,” [16] and we have, with different suffixal material, the same root in an important Cuneiform Luvian word, kattawatnalli- “vindicatif.” [17] In Cuneiform Luvian the word kattawaatnallinza, for example, is in the following phrase (KUB XXXV.45.ii.18): {91|92} KÚR MEŠ -in-zi kat-ta-ua-at-na-al-li-in-za ú-ud-na-as-si-in-za “the enemies, the kattawatnallinza, the lords of the land” (Davies). [18] Since kattawatnalli is an attribute of the word for enemy, [19] and since Hittite kattawatar designates an anger of the divinity that can be called upon to exact vengeance on an enemy, I suggest that the stem *katu- [20] presents us with a significant comparison point for Greek anger and in particular for kótos. Indeed, the parallels to Homeric narrative can be multiplied, as when Mursilis asks the sun goddess that she inflict plague and pestilence on the enemy, just as does Chryses of Apollo in Book 1 of the Iliad, at which point the Hittite king suggests, in the context of his prayer, that the goddess distinguish between good (a-aš-ša-a-u-e-eš) and bad (i-da-a-la- u-wa-aš). [21] [22] In 1965, Laroche argued for a connection between the Anatolian forms and Greek kótos, but more recently, Melchert is reluctant to accept Laroche’s suggestion that the Anatolian and Greek words are connected. [23] Although the morphological variation between the Greek thematic noun and the Germanic, Indo-Iranian, Celtic, and Anatolian u-stems can be referred here to Caland morphology, the semantics remain a problem. But that problem can be resolved by a closer look at the evidence from early Greek narrative. For example, such forms as Gaulish Catu-rīges, where the derivative of the root that yields Greek kótos is coupled with the element rī-, meaning king, make a striking parallel to the connection of kótos with the basileús in Homeric Greek. Moreover, if there is an anger term that invokes both military aggression (as in the Luvian passage) as well as military encounters (“battles,” etc., as in Germanic, Celtic, and Slavic), that term may also refer us to the clear and well-defined sense of anger displayed, as shown above, by Greek kótos. The Homeric evidence makes a striking argument in favor of reaffirming the connection of kótos to its most likely congeners. [24]
Moreover, the Anatolian forms come to be of special interest, in light of the definition of Calchas. For the Achaean prophet, kótos is presented as primarily social in that it is based on violations of the social code. The Anatolian forms encompass socially volatile concepts such as vengeance and reciprocal violence, just as does Greek kótos. A reasonable conclusion is that the Anatolian passages show forms based on katu- in a way that establishes an undeniable semantic fit between terms for this kind of anger in Greek and Anatolian. [25] For example, Luvian kattawatnallinza (discussed earlier) is used as an adjective modifying the word for enemy, with a clear comparison here to Sanskrit ṡátru- “enemy,” but not just any kind of enemy: the magical sequence that provides us with a Luvian cognate of kótos includes the formula KUR MEŠ -in-zi kat-ta-ua-at-na-al-li-in-za ú-ud-na-as-si-in-za, [26] whose meaning encompasses vengeance, violence and society, a meaning forcefully consistent with the kind of feuding enmity firing the Trojan War.
A semantic process links these terms together. An inherited word referring to some aspect of the feud as an institution in the pastoral or seminomadic stage {92|93} of development of the Indo-European world is retained in some of the daughter languages, often in ways that refer to the old institution of the feud. In Greek the cognate refers to an accompanying “emotion,” namely kótos; in Germanic, Celtic and Slavic it is the violent event that marks moments in the course of an ongoing feud (for example, a “battle”); the Anatolian languages earmark the general institution itself with words that refer to “vengeance-seeking,” “vendetta,” or an aggrieved party (“plaintiff’). [27]
My focus on the meaning of kótos as a term that reflects social ruptures serious enough to precipitate the Trojan War resolves the semantic problems surrounding the relation of the Greek term to its cognates. By metonymy, kótos and its congeners come to indicate various aspects of feuding conflict. The basic sense of this term involves political hostility of a severe nature, severe enough to involve divinities, plagues, and disputes between distinct political entities. As we have seen, it is in the Anatolian evidence with its social context that we find a meaning that comes very close to the notion of feud. Succinctly, Piero Meriggi glosses Hittite kattawatar as “vendetta.”27
In conclusion, the Homeric material provides a good reason to see the Hittite term kattawatar as resonant with the characteristics of kótos in Homer, especially with regard to the violation of Laomedon’s contract, the destruction of Troy, and the restoration of Odysseus’s kingship. The notion of vendetta certainly is attractive in its association with Greek kótos, given what we know of its duration over time and that it is related to kingship. In fact, the Greek case provides a secure link between the notion of a legal violation and a battle, both meanings evident in the Indo-European cognates of kótos, the latter meaning manifest in Celtic and Germanic and the former in Anatolian. The semantic difficulties seen as blocking a connection between the Hittite and Cuneiform Luvian forms and kótos in Greek disappear when the anger of the king and the relationship of kótos to dikē (see again Il. 16.384-92) is set alongside the startling conjunction of kótos with the origins of the Trojan war and with the denouement of Odysseus’s nóstos. I hope that such a conclusion leads to further research into the Indo-European vocabulary for aggression and the institutions of violence that surround the proto-society’s warrior figures. But now I turn to the ethnographic literature on feuding to elucidate the kind of anger that is kótos. {93|94}


[ back ] 1. Frisk 1960-1970 s.v., Chantraine 1968-1980 s.v.
[ back ] 2. Indeed katu- shows up in many proper names in Germanic and Celtic: an early form is Runic HaỏulaikaR [Kjølevig c. 500, Krause Nr. 61, cited in de Vries 1961 s.v. hoö]; cf. the warrior’s name Hoöbroddr also cited in de Vries; cf. Old English Heaöulāc. De Vries proposes an etymology different from the one elaborated below. For Celtic names, see Vendryes 1959- s.v. cath.
[ back ] 3. De Vries, s.v.: cf. Old English heaâu-, Old Saxon hathu- with Old High German hathu-, hadu-. De Vries dismisses a connection with Old Church Slavic kotora.
[ back ] 4. In this regard, look also to other Celtic cognates, where katu- is identical to the ancestor of Old Irish cath (u-stem, masc., gen. catha, Lewis-Pedersen 1961, 170; DIL s.v. cath “battle,” cf. Welsh cad “battle,” and the Gaulish onomastic element catu-. See IEW 534, with Vendryes 1959- s.v. cath).
[ back ] 5. The variation in initial consonant is accounted for through “Gutturalwechsel” (* k’ ~ k), Mayrhofer 1986- s.v. ṡátru-.
[ back ] 6. See Frisk (note 1) with his comment, “keine überzeugende Etymologie.”
[ back ] 7. For the connection with -u- and -r- stems, see DELG s.v. kótos; cf. Frisk s.v. kótos; for a clarification of the morphology see Tucker 1990, 40-41. I am grateful for a valuable suggestion made to me by Calvert Watkins regarding the applicability of Caland’s Law to these forms (interview, 1988).
[ back ] 8. On the possible origin of kótos as an old neuter in s, see W. Schulze 1892, 404.2, with Fraenkel 1909-1910. See now Tucker 1990, 35, “The thematic noun ho kóto- (Hom.) “anger” has probably replaced an inherited -s- stem *tò kótos” with her note to Fraenkel’s suggestion that we should “assume s-stem with the expected vocalism, namely, *tò kétos.” Such a transformation would present no metrical problems in the only nominal case attested in Homeric epic, the accusative case (with the exception of the doubtful reading kótōi in Il. 14.111).
[ back ] 9. With o vocalism, cf. pothéō, póthesan, with thésssasthai; phobéō, ephóbēsa “faire fuir” with phébomai; phoréō, ephórēsa, phérō; potéomai, pepótēmai, as listed in Chantraine 1973a, 348-49, where he also tells us that “les itératifs à vocalisme o se confondent dans beaucoup de cas avec les dénominatifs en -έω tirés de noms thématiques,” a confusion that might well apply to the verbal forms of kótos. Note that DELG (s.v. kótos) leaves the matter open (“verbe probablement dénominatif”). In my opinion, kótos belongs with pothéō, etc. as a true deverbative.
[ back ] 10. After Pindar and Aeschylus we have no instances in drama, apart from a passage in Euripides’ Rhesus (828) and some late prose (DELG s.v. kótos).
[ back ] 11. My rendering of the A text in Lebrun follows his translation (“Mais maintenant l’ensemble des [pay]s voisins a commencé à attaquer le pays hittite; que cette situation devienne une fois de plus un motif de vengeance pour la déesse Soleil d’ Arinna” [Lebrun 1980, 170]). I thank Carol Justus for drawing my attention to this passage.
[ back ] 12. Lebrun 1980, 155.
[ back ] 13. Une prière adressée par le roi Mursili II à la déesse Soleil d’ Arinna à la suite de l’ épidémie de peste dévastatrice et de difficultés extérieures” (Lebrun 1980, 155).
[ back ] 14. After Lebrun’s “Tel (est un objet) d’ irritation, de violente colère pour les dieux, [55-56] tel n’est pas respectueux des dieux; aussi, que les bons ne périssent pas avec les mauvais” (Lebrun, 1980).
[ back ] 15. And note erízetai on l. 172.
[ back ] 16. Cf. Friedrich 1960, §83; the comparison to idālu- is made by Laroche (1965, 51).
[ back ] 17. DLL s.v.; H. Craig Melchert (1987, 188-90); see also Davies (1986, esp. 142), who carefully explores the meaning of the word (Hieroglyphic Luvian) KATUNI— including a possible connection with kótos—on the Karkamiš inscriptions (A6 and A7), eighth century; and Weitenberg 1984, 116-17.
[ back ] 18. KUB XXXV 45ii18, as cited in Davies 1986, 142, from a ritual text, Starke 1980, 152. Davies elaborates upon the placement of kattawatnallinza next to LÚKÚRMEŚ- in-zi “enemies”; cf. Rosenkranz 1952, 22; cf. 70-72 (“Die rachesuchenden Feinde des Landes”).
[ back ] 19. Cf. Weitenberg 1984, 117.
[ back ] 20. Davies 1986.
[ back ] 21. See Lebrun 1980, 178-79, with the text on p. 163 (A 56).
[ back ] 22. Melchert 1987, 190 n. 20; Laroche 1965, 51; reluctant also is Mayrhofer 1986;
[ back ] 23. See also the comments by O. Szemerényi (1992, 1592), where he says “At least as important as the data of Celtic, Germanic, and Slavic, are the Hittite forms recognized as cognate by Laroche (1965, 51): kattawatar ‘vengeance, rancune’, and the adjective kattawannalli—based on an unattested *katu- ‘haineux, rancunier’ immediately comparable to κότος.”
[ back ] 24. Rosenkranz 1952, 22.
[ back ] 25. With Davies’ translation (above), rendered by Rosenkranz (1952, 71) “die rachesuchenden Feinde des Landes.”
[ back ] 26. Melchert 1994, 232; see also 1979, 268-71, where he carefully reviews the glosses translating Hittite kattawatar, including “revenge, retribution, reparation; object of revenge, grounds for revenge,” and the juridical connotations of “(legal) grounds for a quarrel, compliant, plaintiff.” Cf. Laroche’s gloss of kattawanalli- as “vindicatif” DLL 55, with Rosenkranz’s translation of the same term as “Rachesuchender” (1952, 71). The institution of the feud involves a wide range of categories, from the quasi-legalistic issues surrounding the resolution of conflict, to the emotions generated by that same conflict; especially salient for kótos are the “legal connotations” of kattawatar, given skholiàs … thémistas “crooked judgments” (Il. 16.387) and ek … díkēn elásōsi “they drove out justice” (388) in the context of Zeùs … kotessámenos (386).
[ back ] 27. Meriggi 1980, 304 ff.