THE LUKKA PROBLEM-AND A POSSIBLE SOLUTION

T. R. BRYCE, University of Queensland

LIKE many other problems associated with the geography of Anatolia in the
second millennium B.C., the question of Lukka's location (or locations) has been widely
disputed ever since the decipherment of the Hittite texts.1 F. J. Tritsch has summarized
some of the major difficulties involved in this question:

Who were these Luqqd or Lukki? Hittite documents mention Luqqd countries in different
places at different times, or so it seems. In the 14th century texts we are led to believe that
there is one (hostile) near Arzawa, another (allied) near Kizwatna and the Hurri territory, and
the Lukki are also in Cyprus, interfering with the Egyptians. In the 13th century texts, the
Luqqd seem to be neighbours of the Azzi and the Ka'ka countries and, together with these,
enemy territory. Whereas in the 14th century they were on the coast, in the 13th century we
find them high up in the mountains north of Syria and near the later Urartu. They also come
to join in the battle of Kadesh, but whereas on this occasion they are the allies of the Hittite
king Muwatallis, their country is called 50 years later, under Tudhalia' IV, enemy territory.
Then, another 10-20 years later we find them among the peoples of the sea landing in Egypt.
Small wonder that one despairs of being able to nail them down to a certain spot before they
settle in Lycia.2

Without doubt the picture thus presented is one of bewildering complexity. However,
it does seem important to recognize that while some texts referring to the Lukka people
have a definite bearing on Lukka's location, others probably do not. And by making such
a distinction, we should at least be able, through a process of elimination, to reduce the
overall problem to less formidable proportions.
Several Egyptian texts appear among the sources which contain references to the
Lukka people. These include the following:

1. A hieroglyphic inscription on an obelisk in the "Temple aux obelisques" at

Byblos, dating to ca. 2000 B.C., and honoring Kwkwn, son of Rwqq. The inscription, which
was transcribed and translated by M. Dunand,3 is discussed by Albright,4 who points out
that Rwqq can be equated with Lukk-5 and Kwkwn with Kukunnis.6 But while this
1 There are no
justifiable grounds for the common 3 Fouilles de Byblos, vol. 2, (Paris, 1958), p. 878
assumption that Lukka = Lycia, e.g., W. F. Albright, No. 16980.
AJA 54 (1950): 168, J. Garstang and 0. R. Gurney,
4 BASOR No. 155 (1959): 33-34.
The Geography of the Hittite Empire (1959), p. 81
(hereafter cited as G.G.), G. L. Huxley, Achaeans and 57Cf. M. Astour, Revue d'Assyriologie 53 (1959):
Hittites (1960), p. 33, F. H. Stubbings, CAH 2 new 72-73.
ed., fasc. 26 (1964), p. 22. Contrast A. Goetze,
6This name, in the form Kuk(k)un(n)i, has been
Kleinasien, Kulturgeschichte des Alten Orient, 2d ed. identified as an "authentically Lycian name" (see
(Munich, 1957), p. 181 (hereafter cited as Kleinasien2) Albright's references to the observations of Friedrich,
and JCS 14 (1960): 48, J. MacQueen, "Geography and Sundwall, and Laroche p. 34 n. 10). One should also
History in Western Asia Minor in the Second Mil- note, as Albright points out, that the name Kukunnis
lennium B.C.," Anatolian Studies 18 (1968): 174-75 is not exclusive to Lukka and Lycia. It also appears
(hereafter cited as "Geography"). in the so-called "Alaksandus Treaty" (G.G., p. 102,
2 Archiv Orientdlni
18 (1950): 499. sec. 5. The text, with German translation, appears in
J. Friedrich, Staatsvertrdge des Hatti-Reiches in
[JNES 33 1974] Hethitischer Sprache, vol. 2, p. 51 ff.) as the name of
? 1974 by The University of Chicago. All rights reserved. Alaksandus's predecessor in Wilusa.

395

Journal of Near Eastern Studies 1974.33:395-404.

Downloaded from www.journals.uchicago.edu by Uppsala University on 05/16/20. For personal use only.
396 JOURNAL OF NEAR EASTERN STUDIES

inscription is highly significant in that it seems to indicate the presence of the Lukka
people in the Near East as early as 2000 B.C., it throws little light on the actual where-
abouts of this people at the time in question.7

2. An Amarna letter, written by the king of Alasiya to Akhenaton and referring to a

raid on the Egyptian coast by a group of people called Lukki8:
This document is commonly cited as evidence for the activities and possible location
of the Lukka people during the first half of the fourteenth century B.C.9 It seems clear
that the raid by the Lukki on Egypt was an extension of their seasonal raids on Alasiya,
and, in view of the regularity of these raids,10 that they had a permanent or semi-
permanent base in the area. If we accept the almost certain identification of Alasiya with
Cyprus,11 then it would seem likely that the Lukki raiders were operating from a base
somewhere on the southern coast of Asia Minor.12 Unfortunately, we cannot be any more
precise than this, and it is indeed possible that they were operating from some other base
in the eastern Mediterranean.13
But even if we accept, as a working hypothesis, (a) that the Lukki raiders were based
on the Asia Minor coast, and (b) that they were linked with the Lukka people of the
Hittite texts,14 it is still very likely that they were an isolated group, with little or no
connection with the area of their origin. We must admit, then, that the Amarna docu-
ment is at best a very dubious source of information on the location and activities of the
Lukka people during the fourteenth century B.C.

3. Records of the Battle of Kadesh, fought in the fifth year of Ramesses II's
reign. 5 Lukka men figure amongst the allies of the Hittite king Muwatallis in this
conflict.16 Most of these allies can be identified with peoples whose names are known to us
from the Hittite texts,17 and many of them came from states which lay relatively close
to Kadesh.18 But other contingents came from much further afield, e.g., from Masa and
Karkisa, which have been plausibly located by Goetze on the Aegean coast.19 The Lukka
contingent may well have belonged to this second category, especially in view of the
frequent association of Masa, Karkisa and Lukka in the Hittite texts.20 Consequently,
7 Cf. Ph. H. J. Houwink ten Cate, The Luwian Catling still does not regard this identification as
Population Groups of Lycia and Cilicia Aspera during fully proven (ibid. p. 62).
the Hellenistic Period (Leiden, 1961), p. 196 and n. 7. 12 One may speculate, on this basis, that they were
8 J. A. Knudtzon, Die El-Amarna Tafeln (1915) the forerunners of the Cilician pirates of later times.
38.10, S. A. B. Mercer, The Tell-El-Amarna Tablets 13 It is perhaps conceivable that they were based
(1939), vol. 1, p. 203, H. Winckler, The Tell-El- in Cyprus itself in an area outside the control of the
Amnarna Letters (1896), p. 86, no. 28 (B") 10-13. The Alasiyan king, if the Alasiyan kingdom did not cover
letter is a reply to Akhenaton's complaint that the the whole island (cf. Catling, CAH 2 new ed., fasc.
people of Alasiya are joint participators in this raid 43, pp. 58-59).
on Egypt. 14 The proposed connection rests essentially on the
9 E.g. Houwink ten Cate, Luwian Population linguistic similarity.
Groups, p. 195, P. Kretschmer, Glotta 21 (1933): 240, 15 I.e., ca. 1299 B.C., according to Astour, AJA 69
J. MacQueen, "Geography," p. 178 n. 4, E. Meyer, (1965): 253. For the source material, see J. H.
Geschichte des Altertums (1913) 12 p. 704, D. L. Page, Breasted, Ancient Records of Egypt III, pp. 125 ff.
History and the Homeric Iliad (1959), p. 24, F. J. (hereafter cited as AR), and for a selection of the
Tritsch, Archiv Orientdlni 18, pp. 497 and 499, extensive literature on this topic, see CAH 2, new ed.,
G. A. Wainwright, JEA 25 (1939): 149. fasc. 37 (1965), p. 62 sec. 1.
10 The letter indicates that the Lukkihad for a num- 16 See in particular, AR III secs. 309 and 312.
ber of years been raiding the Alasiyan city of Sihru. 17 See R. D. Barnett, CAH 2, new ed., fasc. 68
11 The identification with
Cyprus is now accepted (1969), p. 4 and A. Goetze, CAH 2, new ed., fasc. 37
as a matter of course, and substantial arguments have (1965), p. 41.
been put forward in its favor. For a summary of 18 E.g., Kizzuwadna, Carchemish,
Ugarit, Halpa.
these, see H. W. Catling, CAH 2 new ed., fasc. 43 19 See Kleinasien2, map.
(1966), pp. 58-64. One should note, however, that 20 See below.

Journal of Near Eastern Studies 1974.33:395-404.

Downloaded from www.journals.uchicago.edu by Uppsala University on 05/16/20. For personal use only.
 THE LUKKA PROBLEM--AND A POSSIBLE SOLUTION 397

it is probably misleading to regard the presence of Lukka men at Kadesh as having any
direct bearing on the location of their homeland.

4. An account of the "Sea Peoples"' onslaught on Egypt during the reign of Mernep-
tah.21 The names of the sea peoples involved are the 'Ikw', Tr', Lk, Srdn, Skr' (vocalized
as Akawasha/Akaiwasha/Ekwesh, Teresh/Tursha, Lukku/Lukka, Sherden/Shardana
Sheklesh/Shakalsha).22 It seems, from the inclusion of the Lukka people in this account,
that Lukka was a sea power of some significance in the latter half of the thirteenth cen-
tury B.C., although this particular episode was little more than a prelude to the mass
movements which took place the following century (and in which, apparently, Lukka
played no part). However, beyond the fact that Lukka men must have occupied territory
on either the Mediterranean or the Aegean coast (they may even have had settlements on
both) there is little else that can be gleaned from the document as to Lukka's location.

It is obvious, then, that the Egyptian sources throw little light on the question of where
the Lukka lands were actually situated. However, when we turn to the Hittite docu-
ments, the prospects for locating these lands appear, initially at least, somewhat more
promising. These documents suggest two main possibilities:
A. On the basis of KUB XXI 6a (which refers to an unknown enemy who overran
Lukka territory during the reign of Urhi-Teshub or Hattusilis III), F. Cornelius argues
for locating the Lukka Lands in the region of Lycaonia.23 Admittedly, Cornelius's
interpretation is open to some doubt, especially in view of the fragmentary state of the
text in question. His proposal seems to rest initially on the assumption that the countries
listed in 11. 5-11 are in fact the Lukka Lands of 1. 4. One cannot be certain about this;
but if the assumption is correct, a Lycaonia location for the Lukka Lands is quite
plausible, and indeed likely, in view of references to Zallara, to the Lower Land, and to
Harziuna as territorial boundaries (11. 12-14).24 Moreover, there are several other
references to Lukka in the Hittite texts which are at least compatible with this loca-
tion, notably (a) KUB XXVI 12 ii 14 ff.25 where the Lukka Lands are grouped with
Azzi and Kaska as potential enemy territory lying near the frontiers of Hatti,26 (b)
KUB XXIV 3 ii 38 ff.,27 where Lukka is mentioned between Kalasma and Pitassa,28 and

21 AR
III, pp. 238ff. See also Barnett, CAH 2, CAH 2 new ed., fasc. 37 (1965), pp. 25 ff. See also
new ed., fasc. 68, p. 10 and Albright, AJA 54 (1950): Cornelius, "Geographie," p. 239. If Lukka = Ly-
166-67. caonia, then the territories of Azzi, Kaska, and Lukka
22 Barnett, CAH 2, new ed., fasc. 68,
p. 10. would lie directly to the northeast, north, and south-
23 "Geographie des Hethiterreiches," Orientalia west (respectively) of the Hatti lands.
n.s. 27 (1958): 381 (hereafter cited as "Geographie"). 27 Goetze, JCS 14: 46, and Gurney, LAAA 27
24 On Zallara, see G.G.,
p. 64, and on the Lower (1940): 29-31.
Land, G.G., p. 95 and map 1. (Note that Goetze 28 The sequence in which Lukka appears is Ara-
Kleinasien2, map, locates the Lower Land slightly wanna, Kalasma, Lukka, Pitassa. On Kalasma and
further to the northeast.) The most convincing pro- Arawanna, see G.G. pp. 43-46. As far as Pitassa is
posal for Harziuna is, in my view, that put forward concerned, scholars are generally agreed that it lay
by Goetze, e.g., in JCS 14 (1960): 48. Cornelius's view to the west or northwest of the Salt Lake, and north
coincides to some extent with the view once expressed of the plain of Konya (e.g. G.G., p. 74 and map 1,
by Forrer, who located Lukka in southwestern Goetze, Kleinasien2, map, and JCS 14: 48, MacQueen,
Lycaonia, Pisidia, Pamphylia, and Lycia (see P. "Geography," p. 176) although Cornelius places it
Kretschmer, Klein. Forsch. [ 1927], p. 1 f.). Cf. also Hou- further west than most and equates it with Pisidia
wink ten Cate, Luwian Population Groups, p. 195. ("Geographie," pp. 393 and 396). If one can assume
25 The text is edited by E. von Schuler, Archiv a geographical basis for the sequence of names ex-
fiur Orientforschung, Beiheft 10 (1957): 24-25. Cf. also tending from Arawanna in the southeast to Pitassa in
G.G., p. 37. the northwest, a Lycaonia location for Lukka would
26 On the Kaska and Azzi people, see Goetze, be quite consistent with this pattern.

Journal of Near Eastern Studies 1974.33:395-404.

Downloaded from www.journals.uchicago.edu by Uppsala University on 05/16/20. For personal use only.
398 JOURNAL OF NEAR EASTERN STUDIES

possibly (c) KBo XI 40 vi 17 ff.,29 where Lukka appears between Masa and Kaska.30
Cornelius tentatively identifies Nahita, which occurs in the text of KUB XXI 6a,
with Nagidos on the coast of Cilicia Aspera, and claims that Cilicia Aspera was in fact
part of Lukka territory.31 Perhaps, however, Nahita is better identified with the modern
Nigde.32 It is certainly questionable whether the Lukka lands extended from Lycaonia
as far south as the Mediterranean coast, especially in view of the mountain barrier which
almost completely isolates the coast from the interior.
But there can be little doubt that Lukka territory did reach the coast at some point
or other. This conclusion, which is suggested by the Egyptian documents dealt with above
(nos. 2 and 4), is confirmed by the Ugaritic text PRU 5 no. 61 =RS 18, 147,33 in which
Ammurapi, the king of Ugarit, informs the Alasiyan king that he has sent his entire
fleet to the waters off the coast of Lukka. Obviously, important strategic considerations
underlay the decision to send the Ugaritic fleet to this area at a time when the "Sea
Peoples" were beginning to advance south. The coastland in question may be the Cilician
coast, but Astour could well be right in his claim that the intention was "to defend the
passage from the Aegean to the Mediterranean."34 If so, one would need to consider the
possibility of a second location for Lukka.
B. The second proposal depends largely on references to a group of places which appear
to have been situated in or near Lukka territory. To begin with, on the basis of a reference
in the "Tawagalawas letter," 35 the city of Attarimma was almost certainly inhabited by
Lukka people. Moreover, Attarimma was associated with Suruta and Huwarsanassa in a
rebellion against Hittite rule early in the reign of Mursilis II,36 and a further instance of
their association occurs in the Maduwattas document37 where they appear together in
the list of Madduwattas's conquests.38 Possibly, then, all three lay in Lukka
territory.
The above mentioned list includes five other places-Zumanti, Wallarimma, Iyalanda,
Zumarri (?),39 Mutamutassa. Iyalanda also figures in the Tawagalawas document,40 and
its reappearance here in association with Attarimma etc., suggests that it too
belonged to
Lukka territory. One can infer, then, that Zumanti, Wallarimma, Zumarri and Mutamu-
tassa were also Lukka settlements, on the basis of their grouping with
Iyalanda and
Attarimma.
It is possible that this group of eight settlements formed two
geographical clusters-
what we might call an "Attarimma cluster" and an "Iyalanda cluster." The
juxtaposition
of Attarimma, Suruta, and Huwarsanassa in the Madduwattas document
exactly reflects
the association of these places in the Annals of Mursilis. On the other hand, the association
of Iyalanda, Zumarri, and Wallarimma later in the Madduwattas document41 marks
them out also as a distinct group. Zumanti and Mutamutassa do not
appear in Madd.,
rev. ?29. However, if one can assume that the places referred to in rev.
?24 have been

29 Otten, JCS 15 (1961): 112. 36 See Goetze, Die Annalen des

Mursilis, MVAG 38
30 The reference to Lukka in this document is very (1933), pp. 39, 41, 53, 59.
problematical. See below, n. 72. 37 KUB XIV 1. Goetze, Madduwattas, MVAG 32
31 "Geographie," p. 381.
32 See Houwink ten Cate, Luwian (1927) (hereafter cited as Madd.).
Population
Groups, p. 193, n. 7. 38 Madd., rev., sec. 24.
33 Discussed by Astour, AJA 69 (1965): 255.
34 ibid. 39 A conjectural restoration on the basis of Madd.,
rev., sec. 29.57.
35 KUB XIV 3 i 1. See G.G., p. 111, and F. Som-
mer, Die Ahhijavd- Urkunden (Munich, 1932) I, I, 40 AU I, I ff.
1 ff. (hereafter cited as A U). 41 Madd., rev., sec. 29.

Journal of Near Eastern Studies 1974.33:395-404.

Downloaded from www.journals.uchicago.edu by Uppsala University on 05/16/20. For personal use only.
 THE LUKKA PROBLEM-AND A POSSIBLE SOLUTION 399
listed systematically, in accordance with the order in which they were captured, then
one may argue for including Zumanti in the Iyalanda cluster. It is more difficult to place
Mutamutassa because of its intermediate position in the text between the two clusters.42
Further additions to the list of Lukka settlements are more problematical. In the
fragmentary text KUB XXIII 83, Iyalanda is mentioned in association with Dalawa
and Kuwalapassa. According to the first reading proposed by G.G. for this text,43 the
men of Dalawa and Kuwalapassa agreed to join forces with the Hittites in an attack on
Iyalanda. G.G. infer from this document that Dalawa and Kuwalapassa lay close to
Iyalanda, and were in fact Lukka towns. The latter inference does not necessarily follow,
but there may be some justification for the former. It has sometimes been suggested that
Dalawa is to be identified with the classical Tlos (modern Diiver), which lay in the Xan-
thos valley, about 25 miles north of Xanthos.44 Linguistically speaking, the identification
is possible, on the grounds that in the Lycian inscriptions, Tlos is called Tlava.45 We do
not yet know how old the city of Tlos is, although in Roman times at least, it was regarded
as a city of considerable antiquity.46 But Goetze sounds the warning that Dalawa = Tlos
is only a linguistic and not a geographical equation," and also refers to the fact that there
was a Pisidian as well as a Lycian Tlos.48 If Dalawa was in fact the forerunner of the Pi-
sidian Tlos, then there may be some grounds for identifying Kuwalapassa with Olbasa.49
However, the unreliability of such linguistic identifications,50 together with the probable
frequency with which place names were reduplicated, makes any investigation based on
such linguistic grounds extremely precarious.
Nevertheless, the possibility that Dalawa and Kuwalapassa lay in the vicinity of Lukka
territory suggests that Hinduwa may also have lain in the same area. Hinduwa is closely
associated with Dalawa in the Madduwattas document51 which implies that a type of
mutual defence pact existed between the two states. If Dalawa has been correctly identi-
fied with the Lycian Tlos, then G.G.'s identification of Hinduwa with Candyba52 would
also seem a reasonable possibility,53 and certainly more plausible than the identification
with KtVS6w7which, according to Strabo,54 lay north of Halicarnassus.55 At all events, it
seems likely that Dalawa, Kuwalapassa and Hinduwa formed a third cluster of towns
which lay close to, but was apparently distinct from the Iyalanda cluster.
The overall situation can most effectively be illustrated in diagrammatic form (see
table 1).
The asterisks in the table indicate the degree of probability with which the various
places listed can be assigned to Lukka territory. Attarimma is allotted four asterisks.

42 A further
possible addition to the Iyalanda hero Tlos is the brother of Xanthos, Pinaros, and
cluster is the Land of Arinna which is placed directly Kragos, and the son of Tremiles and Praxidice-a
before Wallarimma in the list of Tudhaliyas IV's tradition which again suggests that Tlos itself was
conquests (KUB XXIII. 11 and 12. See G.G., p. 121, a very old foundation.
and also p. 79). One can at least assume that the Arinna 47 Mladd., pp. 153-4.
in question lay somewhere in the vicinity of Lukka 48 Cf. W. M. Ramsay, The Historical Geography
territory. of Asia Minor (1890), p. 414. See also Steph. Byz. s.v.
43 G.G., p. 80. TACs (Meineke, p. 627).
44 Ibid.
49 Cf. G.G., p. 82.
50
5 TAM I, 21.3, 44b.30, 45.2-referred to by Cf. Forrer, Forschungen 1/1, pp. 2-3.
E. Laroche, Bulletin de la Societe de Linguistique de 51 Mladd., obv. secs. 66 ff.
Paris 53 (1957-58): 172. 52 G.G., p. 80. Candyba is the modern Gendova,
46 Quintus Smyrnaeus refers to it in his account of in the mountains west of Kash/Antiphellus.
the Trojan War (X, 147-66), but on the other hand, 53 This is the view favoured by G.G., p. 80.
Homer nowhere makes any mention of it. One might 54 Strabo, 14. 658.
also note that in Panyassis, Heraclea, the eponymous 55 See ]Madd., p. 154.

Journal of Near Eastern Studies 1974.33:395-404.

Downloaded from www.journals.uchicago.edu by Uppsala University on 05/16/20. For personal use only.
 TABLE 1

Madduwattas Madduwattas
Annals of document, document, Tawagalawas KUB XXIII K
Mursilis, 3d year rev. ?30 rev. ?57 document 11 and 12
Attarimma Attarimma Attarimma
Attarimma Suruta Suruta
Cluster ***
Huwarsanassa Huwarsanassa

Mutamutassa
**

Zumanti
**

Iyalanda Iyalanda Iyalanda I

Iyalanda Wallarimma Wallarimma Wallarimma
Cluster ** ** **
(Zumarri) Zumarri
Arinna

D
Dalawa K
Cluster

Journal of Near Eastern Studies 1974.33:395-404.

Downloaded from www.journals.uchicago.edu by Uppsala University on 05/16/20. For personal use only.
 THE LUKKAPROBLEM-AND A POSSIBLESOLUTION 401
The reference to it in the Tawagalawas document puts it almost certainly in the Lukka
lands. Suruta and Huwarsanassa are allotted three asterisks because of their association
with Attarimma. Iyalanda is also allotted three asterisks. The reference to it in the
Tawagalawas document, together with its association with Attarimma, etc., in the Mad-
duwattas document, indicates that it very probably was a Lukka settlement, but less
certainly so than Attarimma. The remaining settlements of the Iyalanda cluster are
allotted two asterisks (with the exception of Arinna) because of their association with
Iyalanda. Arinna, Dalawa, Kuwalapassa, and Hinduwa are each alloted one asterisk.
Although they are associated with possible Lukka settlements, the documentary evidence
does not allow us to assign them to the Lukka lands with any great degree of confidence.
Whether or not all or most of these places lay in Lukka territory, they can obviously
be linked together in one general network. For example, Hinduwa, through its association
with Dalawa, must have been situated relatively close to the Iyalanda cluster. Of course,
the cluster theory does not in any way imply that the states making up each cluster
had any overall political unity. More likely, temporary alliances were established be-
tween various states whenever the need arose.
The problem now is to determine where these Lukka settlements lay within the overall
geographical pattern of Anatolian states. In the first place it is certain that Iyalanda lay
close to Millawanda/Milawata, for Atriya lay in the district of Iyalanda,56 but is treated
in KUB XIX 55,57 as a possession of Millawanda. Thus it is clear that the location of
Millawanda has an important bearing on the location of Iyalanda and consequently on
that of the Lukka lands. Opinion on Millawanda's location is largely divided between two
possibilities-the Milyas district which according to Herodotus was the original name for
Lycia,58 and the Carian Miletus.59 It is also generally inferred, in the absence of any
explicit statement, that Millawanda lay on the coast. This inference is almost certainly
correct, on the basis of A U I, I, 62, in spite of Mellaart's preference for an inland loca-
tion.60 The case for equating Millawanda with Milyas on linguistic grounds has been
supported by Forrer, and later by Cornelius, who is inclined to support Forrer's conclu-
sions while leaving the question of Millawanda's location relatively open.61 The matter
was also debated by Sommer who ultimately rejected the linguistic identification of Mill-
awanda with Milyas.62 Laroche, however, takes issue with Sommer on this point,63 and
Huxley argues in favor of the Miletus identification.64
This divergence of opinion merely serves to indicate the uncertainties associated with
a linguistic approach to the question. Even if it were possible to reach a general agree-
ment on this matter, one may still doubt whether it would settle the problem of the

56
Sommer, A U I, I, 35-39. Hermos, and probably the Maeander and Cayster
57 The so-called "Milawata letter" (A U pp. 198- valleys, and he argues that places like the Lukka lands
240), rev. 47 ff. See also G.G., p. 115. lay north rather than south of Arzawa. Apart from
58 Herodotus, 1. 173. other considerations, it would be very difficult to
59 A third possibility is proposed by MacQueen explain the reference to the Lukka lands in the Ugari-
who places Millawanda on the southern shores of the tic text referred to above if in fact Lukka was situated
Propontis in the vicinity of the later town Miletopolis in the Troad area.
("Geography," p. 175). This, as MacQueen points out, 60 Mellaart, AJA 62 (1958): 26.
was not a colony of the Carian Miletus, but its name 61 Cornelius, "Geographie," 395. Note, however,
was derived from an indigenous tribe of Milatoi. that Cornelius in RHA 62 (1958): 11 states a clear
MacQueen's proposal ties in with his theory that the preference for the Milyas identification.
Lukka lands lay in the region of the Troad (p. 175 62 Sommer, AU
p. 361.
and map p. 176). He bases this theory largely on the 63 Laroche, RHA 69 (1961): 68, no. 46.
debatable assumption that Arzawa occupied the 64 Huxley, Achaeans and Hittites,
pp. 11-12.

Journal of Near Eastern Studies 1974.33:395-404.

Downloaded from www.journals.uchicago.edu by Uppsala University on 05/16/20. For personal use only.
402 JOURNAL OF NEAR EASTERN STUDIES

geographical identification. G.G. attach little weight to the linguistic argument, and
largely on the basis of documentary evidence supported by archaeological finds, favor
the identification of Millawanda with Miletus.65 There is fairly widespread support for
this proposal,66 especially in view of the archaeological discoveries dating back to the
Middle Minoan III Period (i.e., ca. 1600 B.C.) which indicate that there was a flourishing
settlement at Miletus during the period covered by the Hittite texts.67 Now if the Mill-
awanda-Miletus equation is correct, then G.G. may be right in identifying Iyalanda with
Alinda.68 Such an identification would certainly satisfy the condition that Iyalanda
must be close to Millawanda, and as G.G. further show, the physical characteristics of
Alinda may be reflected in the Tawagalawas letter (A U I, I, 16 if.) where the Hittite king
refers to the "difficult terrain" of Iyalanda.69 The so-called "Fremdldinder" list70
probably has an important bearing on the whole question. Iyalanda occurs in this list-
which, as Goetze points out, appears to move in a clockwise direction, and terminates with
the Kaska lands in the north71-in close proximity to Masa and Galkisa (Karkisa?).
Moreover, Lukka is associated with Masa and Karkisa in the "Alaksandus Treaty."72
It seems very likely, then, that Masa, Karkisa, Iyalanda and Millawanda all lay in the
same general area. And it is perhaps noteworthy that all these places are mentioned
directly or indirectly in the Tawagalawas letter in connection with the activities of
Piyamaradus.73 It also seems likely that Masa was situated in the northern half of western
Asia Minor, for it was involved in attacks on Hittite dependencies in the north during the
reign of Suppiluliumas, notably Kassiya and the Hulana River Land.74 Cornelius argues
for a location in the region of Phrygia,75 whereas Goetze locates Masa between the Her-
mos and Maeander Rivers in the region of Lydia,76 which was apparently occupied by the
Maiones in Homeric times." Masa and Karkisa are frequently associated with each other
in the texts,78 and there can be little doubt that they lay close together.79

65 G.G., pp. 80-81. Their evidence, however, de- whether this reference to Lukka should be assigned to
pends overmuch on the relationship between Mill- proposal A or proposal B.
awanda and Ahhiyawa. The Ahhiyawan question is 73 E.g., Sommer, A U I, I, 35 ff; III, 52 ff.
such a vexed one that it is probably better to leave it 71 2 BoTu 34 and 35, E. Cavaignac, Les Annales
aside in attempting to fix the location of Millawanda. de Subbiluliuma (Strasbourg, 1931), p. 6, sec. 26,
66 E.g., Huxley, Achaeans and Hittites,
pp. 11 ff., Goetze, JCS 14, p. 46, H. G. Giiterbock, JCS 10 (1956):
Stubbings, CAH 2, new ed., fasc. 26 (1964), p. 22, 65, MacQueen, "Geography," p. 174.
Albright, AJA 54 (1950): 168. 75 Cornelius, "Geographie," pp. 241 and 397.
67 See, e.g., C. Weickert, Neue Ausgrabungen im 76 Kleinasien2, map.
Mittelmeergebiet und im Vorderen Orient (Berlin, 77 See, e.g., Albright, AJA 54: 168. As Barnett
1959), pp. 181-96, Istanbuler M1litteilungen 9/10 points out (CAH 2, new ed., fasc. 56 [1967], p. 24),
(1959-60): 1-96, and Stubbings, CAH 2, new ed., the origins and early history of the Maeones are lost
fasc. 26 (1964), p. 20. in legend. He goes on to suggest that the name
68 G.G., p. 78. Cf. Huxley, Achaeans and Hittites, "Masha" may be reflected in the name of the Lydian
p. 12. See also Laroche, RHA 69 (1961): 65, no. 34. eponymous hero Masnes (n. 3). Cf. also Barnett,
69 G.G. comment: "The ruins of Alinda... are seen CAH 2, new ed., fasc. 68 (1969), p. 4.
to be one of the strongest fortified positions in Caria; 78 E.g., the "Alaksandus Treaty", sec. 14, and the
the city is built upon a granite crag, and flights of "Tawagalawas document" A U I, III, 52 ff.
wide steps hewn in the rock lead to the main gate."
79 This is not the view of G.G. (see map 1). As far
70 KUB XV 34, translated in ANET, p. 352. as Karkisa is concerned, Forrer identified it with
71 Goetze, JCS 14 (1960): 46. Coracesium (Forschungen 1/1, p. 81). More recently,
72 Friedrich, Staatsvertrdge vol. 2, pp. 66 ff. sec. 14, however, G.G. and MacQueen have proposed locations
G.G., p. 102. As indicated in proposal A, Lukka also in the far northwest of Asia Minor (G.G. p. 108 and
appears next to Masa in KBo XI 40 vi 17 ff. (cf. map 1, and MacQueen, "Geography," p. 174 and p.
MacQueen, "Geography," p. 173). But this document 176, map.), although a more widely held view is that
does not necessarily indicate close proximity between of Sidney Smith (e.g., in a letter to The Times, re-
Masa and Lukka. All that one can assume is that the ferred to by G.G. p. 108) and Goetze (Kleinasien2,
Land of Lukka mentioned here lay somewhere be- map) who identify Karkisa with Caria. Cf. Huxley,
tween Masa and Kaska, as Otten implies (JCS 15 Achaeans and Hittites, p. 33 and Houwink ten Cate,
[1961]: 112). In view of this, it is difficult to decide Luwian Population Groups, p. 196; and contrast

Journal of Near Eastern Studies 1974.33:395-404.

Downloaded from www.journals.uchicago.edu by Uppsala University on 05/16/20. For personal use only.
 THE LUKKAPROBLEM-AND A POSSIBLESOLUTION 403
If Masa and Karkisa have been correctly located on the central Aegean coast, then the
case for identifying Millawanda with Miletus is considerably strengthened. This in turn
would almost certainly mean that Iyalanda and other Lukka settlements were situated
in Caria.
Admittedly, this location poses certain problems:

1. In the text of KUB XXI 6a, the name Wallarimma appears immediately after
the Lukka lands. I have already suggested that the Lukka lands referred to here have been
correctly assigned to Lycaonia,80 and if the Wallarimma of this text is the same as that
which is associated with Iyalanda in the Madduwattas document, it may be difficult to
justify the location of Iyalanda and other Lukka settlements in western Caria.

2. Waliwanda appears in the Tawagalawas letter on the Hittite king's route from
Sallapa to Iyalanda.81 Sallapa almost certainly lay somewhere to the east of the Salt
Lake82 and the document KBo XXI 26 iv83 suggests that Sallapa and Waliwanda lay
relatively close together. This must cast serious doubts on G.G.'s proposal to identify
Waliwanda with Alabanda.84 It would also mean that the Hittite king was several
hundred miles from Iyalanda (if Iyalanda lay in western Caria) when he issued his
ultimatum to Tawagalawas.85

3. KUB XXI 6a mentions Kuwalapassa in the list of conquered states, whereas on

the basis of KUB XXIII 83, Kuwalapassa (along with Dalawa) appears to have been
situated relatively close to Iyalanda. This would mean that Kuwalapassa lay much further
west than the other territories mentioned in KUB XXI 6a-if Iyalanda lay in western
Caria.86
Very likely, many of these place names were duplicated, or transferred from one area to
another, especially in view of the unsettled conditions under which the Lukka people
seem to have lived, as Tritsch points out.87 This may well explain some of the apparent
inconsistencies which occur in the texts. Nevertheless, the problems referred to above do
serve to highlight the uncertainties which cloud the whole Lukka question.
In spite of these uncertainties, there seems to be fairly substantial support in the
Hittite texts for locating a number of Lukka settlements in western Caria near the site
of Miletus.
It is difficult to see how a satisfactory compromise can be reached between proposals
82 See Goetze, JCS 14: 48, who proposes to locate
Cornelius, "Geographie," p. 396. See also W. Eilers, Sallapa somewhere in the vicinity of Kaneg. In spite
Orientalische Literaturzeitung 1935, pp. 201-13, Som- of MacQueen's doubts ("Geography," p. 177), the
mer, A U p. 157, and Albright, AJA 54: 168. Huxley, texts to which Goetze refers lend considerable support
Achaeans and Hittites, p. 20, draws attention to the fact to his proposal.
that the Persians probably called Caria Krk, and refers 83 Cited by Houwink ten Cate in JNES 25 (1966):
to E. Herzfeld, Altpersische Inschriften (Berlin, 1938), 31.
no. 14, p. 27, line 28. G.G. (p. 108) object to the identi- 84 G.G., pp. 78-79. As Laroche indicates in RHA
fication of Karkisa with Caria, on the grounds that it 69 (1961): 61, no. 15, the equation is also improbable
is irreconcilable with the inclusion of Karkisa in the from a linguistic point of view.
"Assuwan Confederacy" which appears in the so- 85 Sommer, A U I, I, 16 ff.
called Chronicle of Tudhaliyas IV (KHUB XXIII 11 86 It would seem from other references that Dalawa
and 12). It seems possible that both Masa and Karkisa at least did lie relatively close to the Aegean coast,
lay between the Maeander and Hermos rivers, with especially if one accepts Goetze's location of Masa,
Karkisa lying slightly to the north. which follows Dalawa in the Fremdliinder list. It is
80 See proposal A above. possible, then, that Kuwalapassa lay a considerable
distance east of Dalawa.
81 Sommer, AU I, I, 6. 87 Tritsch, Archiv Orientdlni 18 (1950): 499.

Journal of Near Eastern Studies 1974.33:395-404.

Downloaded from www.journals.uchicago.edu by Uppsala University on 05/16/20. For personal use only.
404 JOURNAL OF NEAR EASTERN STUDIES

A and B,88 just as it is difficult to find one single location for Lukka which is consistent
with all the relevant texts. The Carian location put forward in proposal B cannot easily
be reconciled with the Hittite texts mentioned in proposal A (especially KUB XXI 6a;
XXVI 12 ii 14ff.; and XXIV 3 ii 38 ff.) which suggest that Lukka territory lay further
east than Caria. On the other hand, any proposal which restricts Lukka territory to the
region of Lycaonia would lead to several very problematical conclusions, notably:

(a) Since Lukka territory must have extended as far as the coast, it probably reached the
Cilician coast, opposite Cyprus. However, it is difficult to see what purpose could have
been achieved by sending the Ugaritic fleet to this area.
(b) The proposed identification of Millawanda with Miletus must be rejected. And one
would need to reconsider the question of whether or not Millawanda was a coastal city.89
If so, then it should probably be located on the Cilician coast, with Iyalanda situated
nearby.
(c) Almost certainly one would have to reject, or drastically modify, Goetze's proposed
locations for Masa and Karkisa on or near the Aegean coast, in view of their apparently
close geographical relationship with Iyalanda etc. But Masa can hardly be shifted any
further south in view of its activities against countries in the north, as indicated in the
Annals of Suppiluliumas.90

The problems involved in attempting to settle on a single location for the Lukka lands
lead me to believe that there were in fact two main groups of Lukka people-one in the
vicinity of Lycaonia, the other in Caria. Quite conceivably, Lycaonia was the original
home of these people, some of whom later moved west, possibly along the Maeander
valley,91 and established themselves on the Aegean coast. Admittedly this theory does
not in any way give a complete answer to the problems associated with a study of the
Lukka people. However, it may provide a partial explanation for the many apparently
conflicting and confusing references to Lukka in the documentary sources of the second
millennium B.C.

88 Cf. Houwink ten Cate, Luwian Population and the interior, as D. Magie points out (Roman Rule
Groups, p. 196. in Asia Minor (Princeton, 1950), pp. 39-40), although
89 See G.G., p. 80, n. 3. he suggests that the Maeander route was later in
90 See n. 74 above.
91 The Maeander origin than the route which followed the valley of
valley provided one of the "time the Hermos.
honoured lines of communication" between the Aegean

Journal of Near Eastern Studies 1974.33:395-404.

Downloaded from www.journals.uchicago.edu by Uppsala University on 05/16/20. For personal use only.