The Trojan War #3: Bronze Age evidence

Part 1. The consensus | Part 2. Homer | Part 3. Bronze Age evidence

Previously we’ve touched on the current state of scholarship concerning the Trojan War, and some points that suggest the Iliad was composed firmly from a seventh-century perspective. But when people argue for a historical Trojan War, it’s evidence from the Bronze Age that carries the most weight — where the ‘Bronze Age’ means about 1200 BCE and earlier. So that’s what we’ll look at today.

Here I’ll be taking it as read that the Homeric name for Troy, Ilios, is a Greek rendering of the Hittite name Wilusa. That equation isn’t rock-solid. There are reputable scholars who doubt it (examples: 1, 2, 3). But I’m not going to be relying on that equation to argue in favour of a historical Trojan War; so if you’ll permit, let’s say I’m allowing that equation to play advantage.

The Alaksandu Treaty: treaty between a 13th century BCE Hittite king and king Alaksandu of Wilusa

Who were the Hittites? This is an important piece of context. The Hittite empire flourished in the second half of the 1000s BCE. By the 1200s they covered most of Turkey, and large parts of Syria and Lebanon. We have a selection of texts from Hittite archives, especially at their capital, Hattusa in central Turkey, as well as diplomatic correspondence at various sites. In the decades after 1200 — the ‘Bronze Age Collapse’ — many Hittite centres were abandoned or destroyed, including the capital, Hattusa, and the empire disappeared except for a remnant in the south-east.

The empire was built out of a combination of military force and one-sided treaties. One such treaty was with king Alaksandu of Troy: it formally incorporated Troy within the Hittite realm as a vassal state and as a part of the Arzawa region. Arzawa was a group of states on the west coast of Anatolia. The treaty dates to the early 1200s BCE.

Troy’s formal position within Arzawa and the Hittite empire roughly corresponds to the period in between Bronze Age Troy at its wealthiest (Troy VIh) and post-Bronze-Age-Collapse Troy (Troy VIIb2, VIIb3) — that is, the archaeological phases known as Troy VIIa and Troy VIIb1. There are no discontinuities between these phases in terms of population or material culture, or anything like that: rather, the dividing lines are disasters which affected the architecture and layout of the citadel. Troy VIh and VIIa are separated by a major earthquake, and Troy VIIa and VIIb1 by a fire.

Timeline of Late Bronze Age/Iron Age Troy and some potentially related events

So: what does Bronze Age evidence have to tell us? Let’s look through the most popular notions: most of them are red herrings.

1. Fire in Troy VIIa

Archaeological evidence shows that the citadel of Troy was destroyed by fire ca. 1190–1180 BCE. One potential cause of fire is human agency; and if the fire was caused by human agency, military hostilities are a possible inference.

The date has a tidy correspondence with Eratosthenes’ guesstimate for the fall of Troy, but as we saw last time, that was just one guesstimate among many, and none of the ancient Greeks who give us these dates had access to any evidence that we don’t also have. 3rd century BCE testimony has zero value.

If the fire was caused by human agency (and that’s not a trivial if: Jürgen Seeher, the director of excavation at Hattusa in 1994–2005, has commented that archaeologists are overly prone to inferring military conflict from fire), even so we have no idea who started it. If it was military hostility, there’s nothing to indicate who the hypothetical enemy was. The Troy VIIa fire gives us a possible scenario for a Trojan War, but nothing that confirms anything.

One reason this fire is so attractive is because classical poems about the fall of Troy put a lot of emphasis on fire. In those poems, though, fire emphasises the point that the city is being eradicated. The real city wasn’t eradicated at all. Troy VIIb1 was a direct continuation from Troy VIIa, and the city continued to be inhabited for nearly 250 years after it supposedly went up in flames.

Troy survived after this fire for longer than the USA has existed.

2. Earthquake in Troy VIh

With the earthquake, Trojan-War-hunters have two main options: (1) Greek myth might have reimagined the earthquake as something to do with Poseidon, who was both an earthquake god and a horse god — and a horse features rather prominently in the Greek legend; (2) the earthquake might have weakened Troy’s defences to the point where Greeks attacked and destroyed the city.

The first option is the most speculative kind of Euhemerism, and even at its best Euhemerism has never been a reliable or even a useful methodology. It’s a game of speculation, not a traceable history.

The ‘weakened defences’ option is at least not ridiculous, but it’s still speculative. If we didn’t have the classical Greek legend, no one would have any reason at all to suspect hostile military action. If there was military action, then — as with the Troy VIIa fire — still we have no clues as to who the hypothetical enemy was.

3. The Tawagalawa letter

One letter in the Hittite text archives, written by a Hittite king and addressed to the king of Ahhiyawa, mentions a dispute over Troy. Most scholars nowadays accept that Ahhiyawa is a hittitised form of the Greek Achaia (or of a Bronze Age form *Achaïwi-).

Conflict involving Greeks and Troy? Must be the Trojan War! Here’s the relevant passage, in Gary Beckman’s translation (Beckman, Bryce, and Cline 2011: 114–117):

(I, the Hittite king, want you to send this message to Piyamaradu:) ‘The King of Hatti has persuaded me (i.e. the Ahhiyawan king) about the matter of Wilusa concerning which he and I were hostile to one another, and we have made peace. Now(?) hostility is not appropriate between us.’ [Send that] to him. ...

... And concerning the matter [of Wilusa] about which we were hostile — [because we have made peace], what then? If [a certain ally] confesses an offense before his ally, [because he confesses] the offense before his [ally], he does not reject [him. Because] I have confessed [my offense] before my brother, [ ... And] let it [ ... ] no further to my brother.

CTH 181 (the ‘Tawagalawa letter’), iv.7–10 and 18–26

Here are the problems: (1) This dispute wasn’t between Trojans and Greeks, it was between Hittites and Greeks. (2) The Hittite king makes it crystal-clear that the Hittites were the aggressors, not the Greeks, and he’s trying very hard to apologise. (3) There’s no powerful reason to interpret the dispute as a war. Here’s the commentary on the passage by the hittitologist Trevor Bryce:

Hostilities had apparently broken out between them over the country called Wilusa ... This is the only occasion in the Ahhiyawa corpus where there is a reference to what appears to have been direct conflict between Hatti and Ahhiyawa. In all other cases, hostile action by Ahhiyawa against Hatti appears to have been limited to support for the activities of local insurrectionists like Piyamaradu. However, we do not know what the nature or the scale of the hostilities was on this occasion, whether it amounted to outright war, a skirmish or two, or merely a verbal dispute conducted through diplomatic channels. (The verb ku-ru-ri-iḫ-ḫu-e-en used in this context could mean any of these things.)

Beckman, Bryce, and Cline 2011: 121

There are potential quibbles over the date of the letter too — Gurney argued in 2002 that it could be as early as the 1290s, Singer has more recently argued for a date in the mid-1200s: either way, it definitely had nothing to do with the fire of Troy VIIa! Whatever the nature of the dispute, it’s not a great match for the legend.

4. The name game

Various names that pop up in Hittite records have tempting similarities to names that we know from classical-era Greek stories. We’ve already seen Wilusa = ‘Ilios’, and Ahhiyawa = ‘Achaia’.

Another interesting one is Alaksandu, the king of Troy who was made to agree to the ‘Alaksandu Treaty’. His name may well be Greek: it looks awfully similar to Greek Alexandros, and Alexandros is absolutely fundamentally Greek: it’s made out of Greek roots (it means ‘man-defender’). Alexandros is also an alternate name for Paris, the legendary Trojan prince. (Again, this is way too early for a Trojan War in Troy VIIa: the Alaksandu Treaty dates to the early 1200s.)

If Alaksandu = Alexandros, though, it just raises more questions. (1) In the legend, Troy was no vassal but an independent city. (2) Paris/Alexandros never became king. Alexandros was a common name (Paris isn’t even the best-known Alexander of antiquity!), so we don’t have a strong case for linking him directly to the legendary character. (3) What on earth is a Trojan king doing with a Greek name? We don’t know the ethnicity of the Late Bronze Age Trojans, but their political and material cultures were rooted in Anatolia, not the Greek world. So, whether you think Alaksandu has anything to do with Paris or not, we have a serious problem here. How on earth did an ethnic Greek get to be king of a thoroughly non-Greek Hittite vassal city? This is a damned good question, and no one has a very tidy answer.

Next up: Attarissiya ‘of the city of Ahhiya’ is named in the ‘indictment of Madduwatta’ (CTH 147) as having attacked a Hittite vassal. If you squint just right, this may look a little bit like ‘Atreus’, Agamemnon’s father in the Greek legend. At least that’s what Emil Forrer suggested in the 1920s.

Ahhiya is certainly an early form of Ahhiyawa, so it does look like Attarissiya was genuinely Greek. (It’s disconcerting that the text uses a city determinative on ‘Ahhiya’, though.) The timeframe is a bad fit: Attarissiya lived in the early 1300s BCE, during the reign of the Hittite king Tudhaliya I/II, three or four generations before the end of Troy VIh (let alone Troy VIIa). But it’s his name that’s the real problem. The similarity, if you can call it that, isn’t nearly as good as Alaksandu or Wilusa ... and it looks like it’s not Greek. Both elements of the name pop up in Hittite names (Attarimma, Sarissiya), and a prehistoric form of ‘Atreus’ should have sounded something like *Atrew(o)-. Forrer’s idea pre-dates the decipherment of Linear B, and linguists don’t buy it these days.

Next is *Akagamuna. A letter from an Ahhiyawan king to a Hittite king (CTH 183) supposedly mentions an Ahhiyawan leader of this name, and it kind of looks a bit like ‘Agamemnon’, the Greek commander in the Iliad. This one’s just based on obsolete information. The initial a- is an old misreading of a damaged determinative sign; the name is really ‘Kagamuna’; and the text is damaged, so we can’t tell whether Kagamuna was Ahhiyawan or Assuwan. He might not even be Greek!

Popular accounts still sometimes try to link Piyamaradu to the legendary king Priam. This one’s just silly. There’s no getting Priamos out of Piyamaradu — ‘Priam’ is indeed a Luvian name, but it comes from an unrelated root, pariya- ‘outstanding’ — and Piyamaradu wasn’t a king of Troy who fought the Greeks, he was a warlord who made a base at Miletus and colluded with the Ahhiyawans against the Hittites.

Last but not least: Appaliuna, a god associated with Troy/Wilusa in the Alaksandu Treaty, definitely is linked to the Greek Apollo (Apollōn), the most important god favouring the Trojans in the Iliad (see especially Bachvarova 2016: 243–250). Out of the names in this list, Appaliuna is by far the strongest link between Homer and Bronze Age Troy. The link is confirmed by a text describing an Arzawan purification ritual, CTH 456, which mentions Appaluwa as a plague god.

Even here some care is needed. In the Archaic period the chief civic cult of Troy was that of Ilian Athena, not Apollo (Il. 6.269–311; Hdt. 7.43; copious later textual, epigraphic, and numismatic evidence). The cult of Apollo was based at Thymbra, nearby, though its location is uncertain (possibly a 6th–5th century BCE site 7 km to the southeast, on the river Kemer). Even so, Thymbra gave Apollo a special relationship with Troy.

(For the linguistic connection between Arzawan Appaluwa, Trojan Appaliuna, and Greek Apollōn see Bachvarova 2016: 246–247: the variation in -l-, -li-, -ll- is explained by Beekes’ conjecture of a palatalised l in pre-Greek, *ly, which also gave rise e.g. to both Ἀχιλεύς and Ἀχιλλεύς. Cf. dialectal forms of Apollo: Doric Ἀπέλλων, Cypriot Ἀπείλων < *Apelyōn. Beekes, aware of Appaliuna but not Appaluwa, reconstructs the proto-form *Apalyun (2010: 119). Personally I’m not convinced that the reconstructed *ly shows Appaliuna/Appaluwa was an import from the Greek world, as Bachvarova argues: we still don’t have Apollo in Mycenaean.)

The Merneptah Stele, Egyptian Museum, Cairo: Merneptah’s campaigns against the Sea Peoples, 1208/7 BCE

5. The Sea Peoples

Popular accounts often link the Bronze Age Collapse to Egyptian reports of conflicts with Sea Peoples. The relevant records are from the reigns of the pharaohs Merneptah (1213–1203 BCE) and Ramesses III (1186–1155 BCE). Sometimes these Sea Peoples get blamed for the downfall of the Mycenaeans and the Hittites.

The catch is that we only have reason to think of the Bronze Age Collapse and the Sea Peoples as firmly connected to one another in the region of Lebanon-Israel. Contrary to popular belief, the Bronze Age Collapse was not a continent-wide catastrophe afflicting all civilisations from Greece to India: the name refers to a political, economic, and demographic upheaval in mainland Greece, Crete, Turkey, Cyprus, Syria, Lebanon, and Israel. The New Kingdom of Egypt purred along for another hundred years; the Assyrian empire was untouched.

Egyptian monuments link the Sea Peoples to events specifically in Egypt, Israel, and potentially as far afield as Cyprus (if we take Merneptah’s propaganda at face value). Now, even granting that Cyprus was a hotspot of Sea Peoples activity, that’s still a loooong way from Greece or Troy (and, in 1208/7, two decades too early for the burning of Troy VIIa).

‘Ah, but the archaeologist Eric Cline wrote a book on how the Sea Peoples ended civilisation in 1177!’ Well, that’s certainly what the title of the book suggests, and there’s no doubt about Cline’s high scholarly standards. But if you actually read the book, you’ll find he’s actually very cautious about this. He specifically argues, in fact, that no single factor caused the collapse, and that there is no powerful reason to see the Sea Peoples as active in Greece and Turkey.

In both stories — the Bronze Age Collapse, and the Sea Peoples — Troy is peripheral. Plenty of Mycenaean and Hittite sites were abandoned, some violently destroyed, at the end of the Late Bronze Age (Mycenae, Tiryns, Gla, Pylos, Hattusa, Ugarit, etc. etc.); others suffered a downturn but survived (Thebes, Knossos). Troy belongs firmly to the second group. Along the more than 1000 km of coastline between Troy and Cyprus — where the Sea Peoples were active — some of the most important sites suffered only a downturn (Miletus), like Troy, or no downturn at all (Ephesus, Tarsus).

6. Solar eclipses

In 2008 two astronomers, Baikouzis and Magnasco, argued that a series of purported astronomical references in the Odyssey pointed to a date of 1178 BCE for Odysseus’ homecoming from the Trojan War (and therefore a date of 1188 for the fall of Troy). This is very much a fringe view, but it did provoke a certain amount of interest. I published a response that rejected the thesis unequivocally, showing that the argument was founded on several kinds of false assumptions, mistranslations, and cherry-picked data.

If you find their idea even faintly appealing, just take a moment before investigating to read the bit straight after Theoclymenus supposedly mentions this ‘eclipse’:

So (Theoclymenus) spoke, and then they all laughed sweetly at him.
Among them Eurymachus, Polybus’ son, led the talking:
‘He’s raving, this guest, this new arrival from abroad!
Come on, boys, throw him out of the house, outdoors
into the town square, since he says it’s like night in here.’

Odyssey 20.358–362

Path of solar eclipse in June 1218 BCE, suggested by Papamarinopoulos et al. as a supposed ‘eclipse’ in Iliad 17.366–377. (Hey, you missed one in Iliad 16.567!) Source: NASA Eclipse Web Site

In spite of this, a few other similar studies have popped up since 2008: 1, 2, 3, 4 (none written or co-authored by ancient historians, surprise surprise). They don’t fare any better. For the record, the first (Henriksson) dates the fall of Troy to 1312 BCE, the other three (Papamarinopoulos et al.) to 1217 BCE. Neither date is a match for either the end of either Troy VIh or Troy VIIa. The Henriksson article at least puts some effort in on the research front, but the others don’t even take a glance at previous research on the text and background of the Homeric poems. Baikouzis and Magnasco report that their research consisted of looking at footnotes in popular translations. Yes, seriously.

(Surprisingly, my own article has earned me a substantial section in the German Wikipedia article on Odysseus. Should I feel chuffed? Actually I do, kind of. It’s just that there’s no way the 2008 argument lives up to Wikipedia’s ‘notability’ standards, so really my response doesn’t either!)

---

An alternative perspective: literary traditions

Up to this point things are looking pretty negative. We’ve got Apollo, we’ve maybe got a kind of parallel to Paris (but with the Greek version of his name), and that’s about it. We have lots of possibilities but not much that is at all definite. What positive things can we say?

Things look very different if we stop agonising over historical events, with the question framed as ‘Was Homer based on fact?’, and instead ask a more literary-mythological question: ‘What kinds of influence from Bronze Age Anatolia can we see in Homer?’ Evidence becomes clearer and looks a lot more pertinent, and things fall into place more simply.

Earlier this year a new book appeared by the hittitologist-hellenist Mary Bachvarova (with whom I had the good fortune to co-teach a course several years ago), From Hittite to Homer (2016). Bachvarova synthesises a lot of material concerning not just historical events of the Late Bronze Age, but more importantly poetic traditions. Homer is no historical record, but literary influence from Bronze Age Near Eastern material is beyond any possible doubt. Mycenaean Greece and the Hittite world were certainly in contact —

• We have about two dozen Hittite documents referencing Ahhiyawa, including one letter from an Ahhiyawan king to a Hittite king;
• Physical evidence of trade across the Aegean, mainly with Miletus in south-western Anatolia (which was actually an Ahhiyawan city for much of the Late Bronze Age), but also to some extent in the north-west, and even a Mycenaean sword blade in the Hittite capital (Cline makes a big thing of this);
• In Hittite sources it’s easy to trace a story of ongoing Ahhiyawan meddling in western Anatolian politics: we’ve alluded already to Ahhiyawan collusion with the rebel warlord Piyamaradu, and a dispute of some kind over Troy.

In addition, Greek poetry and mythology show heavy influence from the Near East. The masterwork on this topic is M. L. West’s The East Face of Helicon (1997). Some elements include

• Story structures like the Succession Myth. Greek: the Ouranos-Kronos-Zeus succession that we see in several Theogonies; Hurrian-Hittite: the Kumarbi cycle and the Song of Ullikummi; Phoenician: Sanchouniathon’s Phoenician history (perhaps: we have no Phoenician source, only a late Greek source); Ugaritic: the Baal cycle; Babylonian: the Enuma elish.
• Stories revolving around a city being destroyed by a spectacular device, often with direct divine involvement. Greek: the Trojan War; Hurrian-Hittite: the Song of Release, about the destruction of Ebla; Hebrew: the sack of Jericho in Joshua 6; Egyptian: the Sack of Joppa.
• Other genres and incidental features, like a god fighting a chaos/water-monster (Zeus vs. Typhoeus, Marduk vs. Tiamat, Baal vs. Yamm); many of the tropes of wisdom poetry (Greek: the Works and Days; Ugaritic/Hittite: the Instruction of Shube’awilum(?) to Zurranku(?); Sumerian: the Instructions of Shuruppak; etc.); parallels in the Hymn to Hermes to a Sumerian story about Lugalbanda.

Note that we’re not talking about direct imitations. Hesiod didn’t have a copy of the Enuma elish in front of him. Rather, we’re talking about traditions: literary and mythological genres and tropes which spread because of their common appeal.

Bachvarova’s new book is tailored more specifically than West’s, and looks at how this influence was filtered through Hurrian-Hittite culture in Anatolia. The result is very compelling in places (though I don’t share her eagerness to latch onto figures like Alaksandu).

The most compelling example of influence in the poetic tradition is the argument made by the great linguist Calvert Watkins that Wilusa — and here we had better stop taking it for granted that Wilusa is Troy — had a place in a Luvian poetic tradition. A Hittite tablet describing cult songs at the religious site of Istanuwa lists the opening lines of several cult songs in Luvian. One hymn-opening reads:

ahha-ta-ta alati awienta Wilusati

When they came from steep Wilusa

CTH 772.1 (= KBo 4.11)

(Bachvarova 2016: 21 reproduces Starke’s transliteration, which has some technical differences.)

The especially striking thing is that the phrase alati ... Wilusati ‘(from) steep Wilusa’ has an identical meaning to three phrases that appear in Homer: Ilios aipeinē ‘steep Ilios’, Priamoio polin ... aipēn ‘Priam’s steep city’, and Ilion aipy ‘steep Ilios’.

If there’s a catch, it’s that the parallel is solely semantic: there’s no possibility of a formula in a shared linguistic tradition. The phrasing is different, and Luvian ala- is unrelated to Homeric aipys and aipeinos. (On the other hand, it is fairly likely that ala- does appear in the Homeric word ēlibatos ‘(steep?)-stepping’, used of high rocks.)

---

The upshot is that it’s pretty much futile to look for historical events in Homer, but in terms of mythological or poetic traditions there is a kind of continuity — very very indirect perhaps, but still traceable — between Bronze Age Anatolia and Archaic Greece. That continuity is perhaps more visible in Hesiod than in Homer.

And it is very indirect. Homer shows barely any trace of effort towards making his Trojans Anatolian. Their names are nearly all Greek: Hector, Deiphobus, Aeneas, Polyxena, Lycaon, Poulydamas, Anchises, Cassandra, Agenor, and many others are all built out of Greek roots, like Alexandros. Only a handful have non-Greek origins: Priam, Paris, Dardanus, and Assaracus. (I leave aside Ilus, Tros, and Troilus, since they’re based on place-names.) Priamos and Paris both appear to come from a Luvian root pariya- ‘outstanding’, and the name Pariyamuwa is attested in Kizzuwatna, the region of Anatolia to the north of Cyprus.

But the overall picture is of a Near Eastern poetic substrate shifting in the course of adapation, and being gradually remodelled as the centuries pass. Anatolian names like Paris get sidelined in favour of characters who are wholly Greek inventions, like Hector and Aeneas. Something comparable happened to the demographics of the real Troy throughout the Dark Age (Aslan and Hnila 2015): Wilusans kept on living there after the end of Troy VIIa and throughout Troy VIIb1, continuing to make Anatolian styles of pottery, but in Troy VIIb2–3 there are increasing signs of migration from Thrace and Bulgaria.

Eventually, when the first signs of Greek settlement start to appear in the mid-700s BCE, there must have been quite an ethnic mix. They were no longer Hittites, Arzawans, Sehans, or Mirans: the Anatolian groups mentioned in Homer are Lelegians, Maeonians, Mysians, and Phrygians. The only Trojan allies to show continuity with any Bronze Age group are the Lycians, from southern Turkey.

However, this ethnic melting pot would be an ideal setting for a legend about an ethnic conflict.

It’d be strange to imagine a 7th century Greek wanting to preserve a 500-year-old Luvian story faithfully. But a story about an ethnic conflict, in a place recently colonised by Greeks, where Greeks were running up against a diverse and entrenched non-Greek population ...?

That’s the situation in the Iliad, and it’s the situation of 8th–7th century Troy. Sure, we can trace elements from Bronze Age Anatolia (and Phoenicia, too: that’s where the Greeks got their alphabet, and the Odyssey is much more interested in Phoenicians than in Anatolians). But the Greeks took this age-old material and made it their own. The Troy of the Iliad is first and foremost an 8th–7th century construct.

References

• Aslan, C. C.; Hnila, P. 2015. ‘Migration and integration at Troy from the end of the Late Bronze Age to the Iron Age.’ In: Stampolidis, N. Chr.; Maner, Ç.; Kopanias, K. (eds.). Nostoi. Indigenous culture, migration and integration in the Aegean islands and western Anatolia during the Late Bronze and Early Iron Ages. Istanbul: Koç University Press. 185–209.
• Bachvarova, M. 2016. From Hittite to Homer. The Anatolian background of ancient Greek epic. Cambridge: Cambridge University Press.
• Beckman, G.; Bryce, T.; Cline, E. 2011. The Ahhiyawa texts. Atlanta: SBL (Writings from the Ancient World vol. 28).
• Beekes, R. 2010. Etymological dictionary of Greek. Leiden: Brill.
• Cline, E. H. 2014. 1177 B.C. The year civilization collapsed. Princeton: Princeton University Press.
• Gurney, O. R. 2002. ‘The authorship of the Tawagalawas letter.’ In: Taracha, P. (ed.). Silva Anatolica. Warsaw: Agade. 133–141.

Part 1. The consensus | Part 2. Homer | Part 3. Bronze Age evidence