MYS0D464EDY48B2·Mystery · event

The Late Bronze Age collapse

Arnaud Groignard · Public domain

When: Mid-13th to 12th century BC; most destructions about 1200–1150 BC⁠[3]
Who fell: Hittite empire, Ugarit, Mycenaean palace states; Egypt survived much weakened⁠[1]
Northern Levant destructions: 1192–1190 BC (radiocarbon, Gibala-Tell Tweini)⁠[4]
Anatolian drought: Three consecutive dry years about 1198–1196 (±3) BC in the Gordion tree rings⁠[7]
Regional climate: About 300 years of drier, cooler conditions from about 3200 years ago⁠[13]
Egyptian account: Ramesses III, Year 8: the Peleset, Theker, Shekelesh, Denyen and Weshesh⁠[2]
Migration: European-related ancestry in early Iron Age Ashkelon, gone in later generations⁠[8]
Status: Partly explained
Questions: 0 of 6 answered

People Ramesses IIIPharaoh of Egypt's 20th Dynasty; recorded the Sea Peoples battles at Medinet Habu Eric H. ClineArchaeologist, George Washington University; author of 1177 B.C.Sturt W. ManningArchaeologist and dendrochronologist, Cornell Tree-Ring Laboratory; led the Gordion drought study Suppiluliuma IILast recorded Great King of the Hittites; reigned from about 1207 BC

Catalog keyMYS0D464EDY48B2

In the 13th century BC the eastern Mediterranean was a connected world of great kingdoms: Egypt, the Hittites in Anatolia, Assyria and Babylonia, the trading city of Ugarit, Cyprus, and the Mycenaean palaces of Greece, linked by diplomacy, royal marriage and a long-distance trade in copper, tin, grain and luxury goods⁠[1]. Within a few generations around 1200 BC most of this system was gone. Hattusa was abandoned, Ugarit was destroyed and never resettled, the Mycenaean palaces burned and Linear B writing ceased, and Egypt under Ramesses III recorded a great battle against a confederation of seaborne peoples⁠[1]⁠[2].

The explanations have accumulated for a century: invasion by the "Sea Peoples", earthquakes, drought and famine, internal revolt, the failure of trade, and "systems collapse" in which several stresses compounded in tightly linked states⁠[3]. The evidence has changed in the last fifteen years. Radiocarbon sequences now fix destruction horizons to within a few years in places⁠[4]; pollen cores from Cyprus, coastal Syria and the Sea of Galilee show a long dry phase⁠[5]⁠[6]; and juniper timbers from Gordion give an annually resolved moisture record across the collapse, with an exceptional three-year drought at about 1198–1196 BC⁠[7].

Climate is now part of every account, but how much it explains is still argued. The low-frequency drying lasted centuries and societies are adapted to single bad years; the Gordion authors argue that what breaks such systems is several failed harvests in a row⁠[7]. Critics warn that the destructions spread over a century and do not line up neatly with any one event, and that precise climate dates can lend false precision to complex histories⁠[3]. Ancient DNA from Ashkelon shows a migration at the transition, but not whether migrants were a cause or a consequence⁠[8]. Of the 6 questions on this site, 0 are answered. The most likely answer to “Were the Sea Peoples a cause of the collapse or a consequence of it?”, not yet confirmed: populations displaced by famine and the failure of their own states, moving with their families.

Key questionsset by the editors

Were the Sea Peoples a cause of the collapse or a consequence of it?Most likely: populations displaced by famine and the failure of their own states, moving with their families

What was the principal cause of the end of the Late Bronze Age palace system in the eastern Mediterranean?Most likely: no single cause; several stresses (climate, famine, migration, war, earthquakes, internal unrest) struck…

Why was Hattusa abandoned?Most likely: the court moved away for political reasons, leaving the capital to decline

All 6 questions

What happened

The collapse was a sequence, not a single day. Knapp and Manning place the changes between the mid to late 13th century and the 12th century BC, at different times in different regions⁠[3]. In Anatolia the last known Hittite king, Suppiluliuma II, began his reign about 1207 BC; after him no ruler is recorded, and Hattusa, long thought sacked, now appears to have been emptied by its own administration and burned later⁠[7]. On the Syrian coast, the terminal destruction at Gibala-Tell Tweini, a dependency of Ugarit, is radiocarbon-dated to 1192–1190 BC⁠[4], and the first burning of Gibala is contemporary with the destruction of Ugarit itself⁠[9]. In Greece the palace of Nestor at Pylos was destroyed at the transition from Late Helladic IIIB to IIIC, about 3150–3130 years BP⁠[10].

Linear B tablets from the Palace of Nestor, PylosZde. CC BY-SA 4.0Artifact

Egypt survived. In his eighth year Ramesses III recorded that the foreign countries "made a conspiracy in their isles", that "no land could stand before their arms, from Hatti, Kode, Carchemish, Yereth, and Yeres on", and that their confederation of Peleset, Theker, Shekelesh, Denyen and Weshesh was defeated on land and at sea⁠[2]. Depending on the Egyptian chronology used, that year falls about 1188 or 1177 BC⁠[7].

Map of the countryOpen the full map

+−

Hattusa (Boğazkale)City

Within 1 km

Gordion, with the Midas Mound tumulusArchaeological site

Within 1 km

Ugarit (Ras Shamra)City

Within 1 km

Gibala-Tell TweiniArchaeological site

Within 100 m

Larnaca Salt Lake, CyprusLake

Within 10 km

Eastern MediterraneanRegion

A region

Sea of GalileeLake

Within 10 km

Soreq CaveCave

Within 100 m

AshkelonCity

Within 1 km

Palace of Nestor, PylosPalace

Within 1 km

Medinet Habu, mortuary temple of Ramesses III, ThebesTemple

Within 1 km

MappedWithin 1 kmA regionNatural Earth

Timeline

Timeline

Squeezed where little happened. Each mark is a published claim, by how it stands:AcceptedMost likelyStill possibleDisputedRuled out

The accounts

The contemporary written record is royal and fragmentary. The Medinet Habu inscriptions and reliefs give the Egyptian view of the invaders, with their names, dress and ships, and the claim that the Hittites and others had already fallen⁠[2]⁠[11]. From Ugarit come letters of its last years, including appeals and reports of enemy ships on the coast, and from Hittite and other archives letters about grain shipments and shortages in Hittite lands in the 13th century⁠[1]⁠[7]. The Gordion authors note that these texts lack detail and context, and that some studies of 13th-century Hattusa do not read them as signs of impending crisis⁠[7]. The Hittite archives themselves are searchable through the Hethitologie-Portal Mainz⁠[12].

The Year 8 inscription and Sea Peoples reliefs of Ramesses IIIOltau, 1177. CC BY-SA 4.0Artifact

The records stop when the palaces stop. Linear B in Greece and Hittite cuneiform in Anatolia end with the administrations that wrote them, so the decades after the fall are known mainly from archaeology⁠[1].

What was recorded

The physical record has three parts. The first is the destruction horizons: burned levels at palaces and cities from Greece to the Levant, some now dated by radiocarbon to within a few years⁠[4]⁠[3]. The second is the climate archives. Pollen from the Larnaca Salt Lake in Cyprus and from Gibala shows drying from the late 13th or early 12th century BC⁠[5]⁠[9]; a quantitative reconstruction for coastal Syria gives a 300-year dry and cool period beginning about 3200 years ago with winter rain down about 40%⁠[13]; the Sea of Galilee core shows the driest episode of the Bronze and Iron Ages at about 1250–1100 BCE⁠[6]; and Soreq Cave isotopes provide a long regional reference⁠[14]. An early synthesis of Mediterranean sea-surface temperatures, speleothems and pollen argued that cooler seas reduced rainfall over the Aegean and Levant at the same time⁠[15]. Not every local record fits: a stalagmite near Pylos shows generally wetter conditions at the time the palace burned, with a short dry spell about 3200 BP and drying afterwards⁠[10].

The terminal destruction layer at Gibala-Tell TweiniKaniewski et al., 1192. CC BY (PLOS ONE's Creative Commons Attribution licence, 2011)Artifact

The third is the Gordion tree-ring chronology, the only annually resolved record from the Hittite heartland. Built from 23 juniper samples of about 18 trees and fixed in time by radiocarbon wiggle-matching, it runs from about 1775 to 748 BC⁠[7]⁠[16]. Ring widths and δ13C both mark about 1198–1196 (±3) BC as an unusually severe run of consecutive dry years⁠[7]. A 2026 re-reading of the same data counts fifteen significant droughts across the last century of the kingdom⁠[17]. Finally, genomes from Ashkelon show that the early Iron Age population there carried European-related ancestry absent before and diluted after⁠[8].

The explanations

Drought and famine. The climate reading holds that a long drying, punctuated by extreme multi-year droughts, cut harvests below what storage and trade could buffer, so that famine, unrest and movement followed⁠[5]⁠[7]⁠[13]. It is supported by the coincidence of the dry phase with the collapse in several records and by the Gordion drought falling in the years before the end of Hattusa⁠[6]⁠[7]. Its weaknesses are that the destructions spread over decades, that some local records show no drought at the moment of destruction, and that a climate stress explains why states were vulnerable more clearly than why each fell when it did⁠[3]⁠[10].

Invasion and migration. The oldest reading takes the Egyptian texts at their word: a coalition of peoples from the Aegean and Anatolia destroyed the cities of the Levant and attacked Egypt⁠[2]. Ancient DNA shows that some newcomers did settle at Ashkelon⁠[8]. Whether they were invaders who caused the collapse or refugees displaced by it is the open question; Kaniewski and colleagues treat the Sea Peoples as a proximate cause downstream of famine⁠[5].

Earthquakes. Nur and Cline found that nearly all destroyed sites lie in zones of strong shaking and suggested an "earthquake storm" in 1225–1175 BC, while calling the evidence inconclusive⁠[18].

Systems collapse. The synthesis most specialists now favour holds that several of these stresses struck tightly connected palace economies within a short period, and that the interdependence that had made the system rich made it fragile⁠[1]⁠[3]. Its strength is that it fits the diversity of the evidence; its weakness is that it is hard to test, because almost any combination of causes can be accommodated.

Published claims

Published claims and their status

Year	Who	Claimed	Status	Reason
1936	Ramesses III, Year 8 inscription (translated by Edgerton and Wilson)⁠[2]	A confederation of peoples from the islands overran Hatti, Cyprus and Syria before Egypt defeated it on land and at sea.	Disputed	Ancient DNA shows that some people did migrate, but Hattusa was emptied by its own rulers rather than sacked, and invasion as the principal cause has been largely abandoned⁠[8]⁠[7].
1966	Rhys Carpenter⁠[19]	A change of climate brought repeated drought and crop failure to southern Greece, which ended the Mycenaean palaces and drove their people away.	Disputed	Regional proxies do show drying around 1200 BC, but the stalagmite record near Pylos shows generally wetter conditions when the palace burned⁠[15]⁠[10].
1982	B. Weiss⁠[20]	Drought drove Luwian peoples out of western Anatolia early in the 12th century BC, and they were among the Sea Peoples who attacked Egypt.	Still possible	Drying around 1200 BC is now well attested and migration is shown at Ashkelon, but no study has tied a drought to the movement of particular peoples⁠[5]⁠[8].
1993	Robert Drews⁠[21]	Raiders fighting on foot with javelins and long swords defeated the chariot armies of the kingdoms; they caused the catastrophe rather than following it.	Disputed	Later reviews find the material evidence for attackers ambiguous at most sites and treat warfare as one stress among several⁠[3].
2000	Amos Nur and Eric H. Cline⁠[18]	Nearly all the destroyed sites lie in zones of strong shaking, and an earthquake storm in 1225–1175 BC may have destroyed many of them.	Still possible	No systematic archaeoseismic survey of the destruction layers has tested it, and the authors themselves call their evidence inconclusive⁠[18].
2010	David Kaniewski and colleagues⁠[9]	Pollen at Gibala shows drier conditions from the late 13th or early 12th century BC, and the first burning of Gibala was contemporary with the destruction of Ugarit.	Accepted	Pollen records from Cyprus and the Sea of Galilee show a dry phase at the same time⁠[5]⁠[6].
2011	David Kaniewski and colleagues⁠[4]	The terminal destruction of Gibala-Tell Tweini, and with it the collapse in the northern Levant, dates to 1192–1190 BC.	Accepted	The date rests on a stratified radiocarbon sequence tied to texts; whether other destructions cluster with it is the open question⁠[4]⁠[3].
2012	Brandon L. Drake⁠[15]	Cooler Mediterranean seas around the end of the Late Bronze Age reduced rainfall over the Aegean and the Levant.	Most likely	Later pollen reconstructions agree on a cool, dry period from about 3200 years ago, though some local records differ⁠[13]⁠[10].
2013	Dafna Langgut, Israel Finkelstein and Thomas Litt⁠[6]	The driest episode of the Bronze and Iron Ages in the southern Levant fell at about 1250–1100 BCE, followed by recovery in Iron Age I.	Accepted	It agrees with the pollen records from coastal Syria and Cyprus⁠[13]⁠[5].
2013	David Kaniewski and colleagues⁠[5]	Climate-driven famine was the ultimate cause of the crisis, and the Sea Peoples' invasions of Cyprus and Syria followed from it.	Still possible	Drought around 1200 BC is now well attested, but its timing varies by region and most specialists prefer several interacting causes to one first cause⁠[3]⁠[10].
2014	Eric H. Cline⁠[1]	A perfect storm of climate change, famine, earthquakes, invasion and the breakdown of trade brought down an interconnected system of states.	Accepted	The view most specialists now favour, though which stress came first is still argued⁠[3]⁠[7].
2016	A. Bernard Knapp and Sturt W. Manning⁠[3]	The changes came at different times from the mid 13th to the 12th century BC, and the evidence supports neither a single cause nor a precise trigger date.	Accepted	The standard critical review; later work on Gordion and Pylos adds detail without overturning it⁠[7]⁠[10].
2017	Martin Finné and colleagues⁠[10]	Near Pylos, conditions were generally wetter when the palace was destroyed, with a short dry spell about 3200 BP and drying only afterwards.	Accepted	A well-dated local record of 24 U-Th dates; how widely the pattern applies outside the Peloponnese is untested⁠[10].
2019	Michal Feldman and colleagues⁠[8]	European-related ancestry arrived at Ashkelon at the Bronze to Iron Age transition, showing a migration, and was diluted in later generations.	Accepted	Direct genetic evidence that people moved, though it does not show whether they were a cause of the collapse or a consequence⁠[8].
2019	David Kaniewski and colleagues⁠[13]	Coastal Syria had about 300 years of cooler, drier climate from about 3200 years ago, with winter rain down about 40%.	Most likely	The dry phase matches other Levantine records, but the size of the change rests on one pollen-based reconstruction⁠[6]⁠[9].
2023	Sturt W. Manning and colleagues⁠[7]	Three consecutive severe dry years about 1198–1196 BC in central Anatolia could have overwhelmed Hittite food storage and brought down the state.	Most likely	The drought is well dated and fits the abandonment of Hattusa, but a record 230 km away does not prove the cause⁠[7]⁠[17].
2026	Graham Hagens⁠[17]	Fifteen significant droughts over the 13th and 12th centuries, not a single episode, drove the long decline of the Hittite kingdom.	Still possible	A re-reading of the same Gordion data, published this year and not yet answered in print⁠[17].

What would settle it

More annually resolved and absolutely dated climate records from Greece, Cyprus and the Levant, of the Gordion kind, would show whether the 1198–1196 BC drought was regional or Anatolian. Radiocarbon sequences at the main destroyed sites, modelled together, would show whether the destructions cluster in a few years after the drought or spread across decades. Ancient DNA from more Iron Age sites would show how many people moved and from where. And a dated corpus of the grain and shortage texts would let the written evidence be placed against the climate record year by year. None of this is likely to name a single cause; it could show which stress came first.