The Retreat of the Ten Thousand was the withdrawal and subsequent dispersal, from 401 to 399 BCE, of the Greek force that had marched with Cyrus the Younger after he was killed at Cunaxa. The soldiers first joined Cyrus's Asian troops under Ariaeus and negotiated with Artaxerxes II through Tissaphernes. When the principal Greek commanders were seized and killed, the remaining officers and soldiers reconstructed a command, fought north along the Tigris, crossed Carduchian country and the Armenian highlands in winter, and saw the Black Sea from Mount Theches before reaching Trapezus. Their journey continued along the coast and through Thrace with weakening cohesion; many survivors entered Spartan service in Asia Minor in 399 BCE. The events are known overwhelmingly through Xenophon, who took command during the crisis and made his own rise one of the narrative's organising themes. His account records military improvisation and collective endurance, but it also turns a mercenary army's struggle for pay, provisions and employment into an exemplary story about leadership. From the Achaemenid perspective, the episode was neither a Greek conquest nor proof that the empire could not defend itself. Artaxerxes had defeated his dynastic challenger and recovered Cyrus's Asian following; the Greek commanders were then seized and killed. The survivors took the dangerous northern route through mountain warfare and winter. The Greeks survived, but their survival depended on bargaining, guides, food taken from villagers and the contingencies of geography as well as battlefield skill.
Contents
- A coalition without its prince
- An armed negotiation
- Reconstituting an army
- Along the Tigris
- Seven days among the Carduchians
- The Centrites and Armenia
- Winter and the cost of movement
- Reconstructing the route
- Theches, the sea and Trapezus
- Beyond the Black Sea
- What the retreat shows about empire
- Primary sources
- How we know
- References
Once news of Cyrus's death was confirmed, the Greeks' success on their wing at Cunaxa could not alter the dynastic outcome. Artaxerxes remained alive, and the purpose for which they had marched had failed. A royal envoy put the point directly: "The King believes that he is victor because he has slain Cyrus." The Athenian Theopompus nevertheless replied that the Greeks possessed "arms and valour" and would fight for whatever else they needed.[1] The soldiers retained their weapons and remained dangerous, but they no longer had a prince or a political objective.
Artaxerxes did not need to destroy every Greek where they stood. He had secured the throne, and his officers could work on the coalition that Cyrus had assembled. Pierre Briant reconstructs the next stage as a sequence of separation and containment: Tissaphernes recovered Ariaeus and the Asian contingents for the king, the Greek commanders were eliminated, and the remaining soldiers were channelled towards a northern route whose rivers, mountains, winter and armed inhabitants seemed likely to finish what a pitched battle had not.[2] The retreat was therefore not simply a continuous running fight between ten thousand Greeks and a single Persian army. It changed character repeatedly, from negotiation under royal pressure, through pursuit by satrapal forces, to violent passage through communities whose relations with the court ranged from service and treaty to practical independence.
A coalition without its prince
The first problem was political rather than geographical. Cyrus's Asian soldiers had withdrawn with Ariaeus, one of the prince's senior commanders. Clearchus offered him the throne on the ground that "to those who are victorious in battle belongs also the right to rule".[3] Ariaeus refused. His reported reason was not cowardice but rank: "there were many Persians of higher rank than himself and they would not tolerate his being king."[4] The reply exposes the implausibility of treating Cyrus's coalition as an army that could simply transfer its royal project to a new candidate. Cyrus had possessed an Achaemenid pedigree, a record of command and a network in western Anatolia. Ariaeus did not.
The Greeks and Ariaeus's followers instead swore to return together. Xenophon describes a mixed ritual in which a bull, a boar and a ram were sacrificed over a shield, the Greeks dipping a sword in the blood and the Asians a lance. The two parties promised that "they would not betray each other and that they would be allies", while Ariaeus's side also pledged to guide the march.[5] This was a practical compact between men exposed by the same lost rebellion, not a fusion of communities. Ariaeus knew the country and royal politics; the Greeks brought a compact infantry force. Each needed the other while the king's response remained uncertain.
Ariaeus has often been cast as a traitor. That judgement relies too heavily on the Greeks' retrospective language. He had served the dead challenger and now faced an undefeated king whose government could punish or pardon. Tissaphernes' most consequential achievement after Cunaxa was to reconcile Cyrus's Asian officers to Artaxerxes. It restored trained soldiers and commanders to royal allegiance and isolated the mercenaries for whom reintegration was harder. The arrangement was also fragile: Greek and Asian camps marched separately, suspicion followed every delay, and the alliance of necessity ended when Ariaeus accepted the royal settlement.[6]
An armed negotiation
The Greeks' first dealings with Artaxerxes were negotiations conducted under arms. The king demanded that they surrender their weapons. They refused, argued that they had been deceived about Cyrus's object, and offered service to the king. Tissaphernes then proposed to obtain safe conduct and a market for them, provided they did no damage while passing through royal territory. Oaths and right hands sealed the agreement between Tissaphernes, a royal relative, and the Greek officers.[7] The terms recognised facts on both sides. The Greeks could not be disarmed cheaply; the royal government did not intend to let an armed force plunder its way unopposed through the central provinces.
For weeks the parties marched in an atmosphere of controlled distrust. Tissaphernes' troops kept their distance, while the Greeks watched what they interpreted as delay, surveillance and preparation. The satrap's position cannot be reduced to the stock figure of an inherently treacherous Persian. He was serving the king against men who had invaded with a rebel prince and who still refused to lay down their arms. Yet Xenophon's account is unequivocal that sworn safe conduct was broken. Gabriel Danzig's study of Tissaphernes in Xenophon separates that charge from an ethnic caricature: the narrative condemns perjury because a leader who destroys his reputation for reliability loses an instrument of power, not because deceit is presented as a Persian essence.[8]
The crisis came when Clearchus sought a personal meeting to clear away mutual accusations. Some soldiers objected that all the senior officers should not go and "that they should not trust Tissaphernes". They were overruled. Clearchus, Proxenus, Menon, Agias and Socrates went to the satrap's camp with a larger body of captains and soldiers. Xenophon's compressed sentence gives the result: "at the same signal, those within were seized and those outside were cut down." The five generals were taken to Artaxerxes and, with the exception of Menon for a time, executed.[9] The killings removed men who carried Cyrus's closest Greek connections and decapitated the mercenary command in one operation.
Achaemenid evidence does not supply an independent narrative of the meeting. The event is known through the surviving side, and Xenophon was not present inside the satrap's tent. A later Greek tradition elaborated motives, speeches and royal intrigue, but cannot repair that evidential gap. What can be said is narrower. The commanders entered under pledges and did not return; the royal government thereby exchanged a difficult escort problem for a leaderless armed body. Whether Artaxerxes ordered the operation in detail or ratified it after capture cannot be recovered. The strategic effect is clearer than the chain of decision.[10]
Reconstituting an army
The seizure did not leave the Greeks literally without officers. Captains, unit structures and experienced soldiers survived. Its danger was that no accepted authority linked those units or made collective decisions possible. Xenophon stages the night after the disaster as the moment of his own emergence. He reports that the soldiers were despondent, "not less than ten thousand stadia" from Greece, with no guide, "cut off by impassable rivers which flowed across the homeward route".[11] In his telling, a dream and private reflection moved him to summon the surviving captains, speak before them and take the place of Proxenus, whose guest and follower he had been.
Xenophon wrote retrospectively, knowing that he had survived and that his leadership would become famous. Carol Atack calls the catastrophe his moment of "moral luck": Cyrus's death and the commanders' destruction gave an Athenian volunteer an opportunity that success under Cyrus would never have offered. Xenophon's speeches transform a mercenary force into a temporarily heroic community and justify his own choice to lead it. The achievement need not be invented for the self-fashioning to matter. The narrator selects the deliberations, omens and exemplary actions through which his later authority appears both necessary and deserved.[12]
Reorganisation was nevertheless collective. The surviving officers chose replacements for the five generals, and a full assembly of soldiers heard proposals and voted. Cheirisophus, a Spartan who had joined Cyrus with ships and troops, became the senior figure at the front; Xenophon and Timasion took responsibility at the rear; Cleanor and Philesius held other commands. Decisions could be proposed by one man, amended by others and ratified by raised hands. When Xenophon urged the abandonment of wagons, tents and superfluous baggage, Cheirisophus put the motion, and "They all raised their hands."[13] The army that emerged was neither a democracy without officers nor the personal following of Xenophon. It was an armed, mobile association in which generals commanded only while captains and soldiers continued to confer legitimacy.
The first formation chosen for the march was a "hollow square", with baggage and the large body of camp followers protected inside.[14] That decision shows what the famous number conceals. The Ten Thousand were not ten thousand identical hoplites moving alone. Light troops, a small cavalry arm assembled later, enslaved and free attendants, traders, women, animals and accumulated property travelled with them. Protection of this human and material train constrained routes and speed. The order to burn wagons and tents was an attempt to make the force tactically responsive, but it did not abolish its dependence on carriers, pack animals and food taken or purchased along the way.[15]
Along the Tigris
Once the compact with Tissaphernes had ended, royal and satrapal troops pursued. The force described by Xenophon was composite: Tissaphernes' own cavalry, troops associated with Orontas, men formerly attached to Cyrus, forces of the king's brother and reinforcements from Artaxerxes. The list matters because "the Persians" were not a single undifferentiated host. Royal authority mobilised household, provincial and transferred rebel formations, then coordinated them against the Greeks.[16] Cavalry and missile troops could harass the column without closing with its heavy infantry.
The Greeks initially lacked an answer. Xenophon says that "the Greek rearguard, while suffering severely, could not retaliate at all" against missiles delivered from beyond its effective range. The tactical imbalance reduced one day's movement to "not more than twenty-five stadia".[17] Survival depended on adaptation, not on an invariant superiority of hoplite courage. Rhodian soldiers who could sling lead farther than the Persians' stone slingers were organised into a unit; horses were gathered and riders appointed. "We need slingers ourselves at once, and horsemen also", Xenophon has the officers conclude.[18] Counter-skirmishing, aggressive rear-guard action and seizure of heights gradually made pursuit more costly.
These actions were not another Cunaxa. The royal forces did not offer the dense infantry collision that suited the Greek formation, and the Greeks could not convert local tactical successes into control of territory. They sought passage, not battle. Their declared policy was to do as little damage as possible if no one hindered the journey and to fight if anyone did. That formula combined prudence with a threat whose burden fell on communities along the route. Any distinction between peaceful provisioning and coercion depended on whether local people accepted the army's terms.[19]
Near the upper Tigris, the strategic frame altered. The Greeks learned that the direct passages west were blocked by rivers, defended crossings or routes they could not use. The way north led into the mountains of the Carduchians. Xenophon says these people "were not subjects of the King", yet adds that they sometimes made treaties with the satrap in the plain.[20] The two statements should be held together. The Carduchians could resist royal demands and maintain their own armed communities without inhabiting a sealed world beyond all imperial contact. Achaemenid frontiers often consisted of bargaining, tribute claims, local autonomy and intermittent force, rather than a line separating governed territory from an external wilderness.[21]
Seven days among the Carduchians
The Greeks entered Carduchian country hoping to pass without open fighting. At first they refrained from seizing more than necessary and tried to establish friendly relations. The Carduchians abandoned settlements, refused the approach and attacked from the heights.[22] Their decision had its own logic. A large foreign army, accompanied by non-combatants and compelled to eat, was not a neutral traveller. Even disciplined requisition could strip a village. The Greeks' claim that they wanted only passage did not oblige inhabitants to believe them.
For seven days the column fought through narrow ways under stones, arrows and attacks on the rear. Xenophon later judges that the army "suffered more evils than all which they had suffered taken together at the hands of the King and Tissaphernes".[23] The comparison shifts attention away from a simple contest between the Greeks and the king. The most dangerous sustained resistance did not come from Artaxerxes' field forces but from mountain communities defending their own ground. Greek success consisted in forcing passages, capturing guides, taking positions by night and keeping the column moving. It did not make the Carduchians subjects, conquer their country or establish a route for a returning administration.
Violence was reciprocal and intimate. The Greeks took prisoners to obtain information; guides could be threatened or bound; villages and stores were entered; Carduchians attacked stragglers and recoverable baggage. Xenophon renders these episodes as tests of command and ingenuity, but they were also encounters between an army's survival needs and local households' need to preserve winter food, animals and security. The people whose fields and houses supplied the march rarely speak in the text. Their actions, flight, concealment, refusal and ambush, are often the only surviving traces of their choices.[22][23]
The Carduchian passage also dissolved the idea of a single Persian pursuit. Royal troops did not follow the Greeks into terrain where their own relations with the inhabitants were hostile or uncertain. Briant argues that the abandonment was strategic rather than proof of impotence: once the Greeks had been pushed onto the northern itinerary, Persian commanders expected the environment, mountain fighters and provincial troops ahead to destroy them.[24] The calculation failed, but it was intelligible. The Greeks' survival does not mean their opponents had chosen the costliest possible means of ensuring their death.
The Centrites and Armenia
The river Centrites formed the boundary, in Xenophon's geography, between Carduchian country and Armenia. The exhausted Greeks found the far bank occupied. Xenophon identifies the force as troops of Orontas and Artuchas, including Armenians, Mardians and hired Chaldaeans.[25] This Orontas is Orontas of Armenia, a royal son-in-law and provincial governor on the king's side in 401 BCE. He is not the Orontas whom Cyrus had accused and executed before Cunaxa. The common name has encouraged accidental conflation, but the careers, offices and moments in the campaign are distinct.[26]
Briant identifies the Chaldaeans in this northern force as Chaldaeans of Armenia and notes their paid military service.[27] Kuhrt uses the same Greek-derived label for groups settled in Babylonia in the early first millennium.[28] These are separate geographical and historical contexts; similarity in English spelling does not establish identity. The Mardians remain the contingent named by Xenophon at the Centrites, without a broader identification here.
The Greeks crossed the Centrites by finding a ford, feinting at one point and forcing the passage at another. Provincial troops withdrew rather than accept close combat on unfavourable terms. That tactical result opened Armenia, but not an empty highland. Xenophon next names Tiribazus as the governor who met the army. The passage identifies him by context as the governor who met the army in Armenia under Artaxerxes II; no broader namesake identification is required. He agreed that the Greeks might pass without burning houses if they did not take more provisions than they needed, and he offered a market.[29]
The compact with Tiribazus soon broke down amid mutual suspicion. Xenophon says that scouts discovered troops gathering on the heights and that the Greeks attacked the governor's camp, taking equipment while Tiribazus escaped. Whether the governor intended an ambush, was collecting forces to shadow the march or was preparing to enforce the provisioning terms is recoverable only through the Greek narrative. The sequence repeats an important pattern: agreements were possible, but a large armed body moving without an accepted sovereign guarantee could not sustain trust for long.[29]
Winter and the cost of movement
The Armenian highlands exposed the column to severe cold and snow. Snow covered people and animals during overnight halts; Xenophon gives its depth as "a fathom" and reports that pack animals, enslaved attendants and about thirty soldiers died.[30] Hunger and exposure broke the marching order. Some men lagged behind with what the translation calls "hunger-faintness"; others lost sight to snow glare or had toes destroyed by frostbite.[31] The losses and injuries increased the pressure to obtain shelter and provisions.
Fire, food and shelter were unevenly distributed. Men who reached villages first had advantages over those guarding the rear. Officers compelled those indoors to yield space to the sick. Soldiers learned that eating relieved hunger collapse and that dark cloth over the eyes helped against snow blindness. These responses were learned during the march. Xenophon's narrative often places adaptation within the story of command, but ordinary soldiers' observations also shaped survival.[31][32]
One episode shows the pressure placed on an Armenian village chief. A Greek officer seized "all the villagers, the village chief included", together with horses and "seventeen colts which were being reared for tribute to the King".[32] The detail joins imperial extraction to army survival. The colts embodied a household's obligation to royal authority; the Greeks appropriated them while taking the people who knew the roads. The village chief was compelled to guide the march and held under threat, even as Xenophon describes negotiations, gifts and a parting relationship that he presents as cordial. Agency operated under an unmistakable imbalance of force.
The houses themselves were "underground, with a mouth like that of a well, but spacious below". People, cattle, sheep, goats, poultry, wheat, barley, pulses and barley wine were sheltered within.[33] The description is one of the richest ethnographic scenes in the Anabasis, but it is also a record of an army entering domestic space and consuming stored subsistence. The welcome that Xenophon recalls cannot be generalised to every inhabitant. Some cooperated, some fled, some were captured, and all faced the immediate demand of thousands of hungry outsiders.
Provisioning exposed the cost imposed on local communities. The Greeks sometimes bought food at markets organised by Persian officials or Greek cities. Elsewhere they took what they judged necessary, plundered settlements they considered hostile or seized guides to reach the next supply. The distinction between requisition and pillage was made by the armed party. At different points the same force was vulnerable to royal pursuit and able to coerce individual villages.[7][32][34]
Reconstructing the route
Xenophon supplies day marches, rivers, peoples and occasional distances, yet the inland route remains disputed. Ancient names do not map neatly onto modern ones; distances may be schematic or corrupted; the army's knowledge came through interpreters and compelled guides; and some labels appear nowhere else in a securely identifiable form. Christopher Tuplin and Shane Brennan stress the striking paucity of comprehensible proper names in Book 4. Between the Tigris and Trapezus, the very portion readers most want to plot, the narrative often withholds the route choices that would explain why the army moved as it did.[35]
The highland river called the Phasis cannot be equated by name alone with the better-known Phasis of Colchis. Proposed routes diverge widely at this stage; Tuplin and Brennan discuss an Araxes and Pasinler context while retaining substantial uncertainty.[36] Larisa beside the Tigris is identified with Nimrud, not the Thessalian Larisa.[36] The Chaldaeans named in the mountains belong to the Armenian context established above, not automatically to Babylonia.[27][28] Gymnias and several communities along the route remain disputed. Identification must follow sequence, terrain, distance and neighbouring places; where those controls do not converge, the location remains uncertain.
Xenophon's itinerary is a narrative rather than a survey. Tuplin and Brennan show that landscape drives Book 4, while route deliberations are largely suppressed and the human landscape is often left anonymous. The stages and parasangs suggest orderly movement but cannot validate one modern route on the ground.[37] Broad corridors can therefore be reconstructed with more confidence than a chain of exact coordinates.
Theches, the sea and Trapezus
From a summit called Theches, men at the front raised a noise that those behind first mistook for an attack. The cry became "The Sea! The Sea!" and passed down the column.[38] Xenophon says that soldiers, generals and captains embraced "with tears in their eyes" and heaped stones into a cairn.[39] It is the most famous scene in the Anabasis, and one of the most durable homecoming images in Greek literature. Yet the sighting was not arrival, and the sea was not Greece.

Hermann Vogel's late nineteenth-century illustration of the Ten Thousand sighting the Black Sea. Xenophon places the cry ‘The Sea! The Sea!’ on Mount Theches, before the army reached Trapezus; the armour, horses, terrain and dramatic staging are later reconstruction, not ancient evidence. Hermann Vogel, The Illustrated History of the World (1881–1884) · via Wikimedia Commons · Public domain source ↗
Mount Theches cannot be located with certainty, and the army still had to cross inhabited country before reaching the coast. The first secure terminus was Trapezus, "an inhabited Greek city on the Euxine Sea, a colony of the Sinopeans in the territory of Colchis".[40] The wording places Greek settlement within a landscape belonging to others, not at the edge of an empty barbarian world. Trapezuntians offered gifts and a market; the army sacrificed, held games and rested. It had reached a community through which ships, news and Greek networks became imaginable.

A silver drachm attributed to Trapezus on the Black Sea, fourth century BCE, photographed in the British Museum. It is broadly contemporary civic material from the settlement reached after the inland march; the Commons record supplies the attribution but no museum inventory number. Constantine Michailidis · British Museum · via Wikimedia Commons · Public domain source ↗
Even at this apparent deliverance, the force lived by pressure on its surroundings. Xenophon says it stayed about thirty days in Colchian villages and "from these as a base plundered Colchis".[34] Celebration and extraction occurred together. The sea solved neither transport nor pay. There were not enough ships to carry everyone, coastal cities feared the burden of thousands of armed men, and commanders disagreed over whether to sail, march, found a city or seek new employment. The coherent army of the mountain narrative began to fragment as the common danger receded.
At Cerasus, Xenophon reports a count of "eight thousand six hundred men" and says the rest had died "at the hands of the enemy or in the snow, a few also by disease".[41] The figure is Xenophon's own reckoning and should not be converted into precise loss rates from an equally uncertain starting total. It does establish the scale of survival he wished readers to grasp. Thousands had reached the Black Sea; thousands of others, including attendants barely visible in the totals, had not.
Beyond the Black Sea
Books 5 to 7 are often compressed into an epilogue because they disturb the clean shape of a homeward march. Xenophon's narrative does not complete the promised return to Greece, and its settlement projects repeatedly remain unrealised.[42] In Book 6 the soldiers resolved to appoint a single commander. Xenophon declined; the army chose Cheirisophus, but by 6.4 he had died after taking medicine for a fever and Neon had succeeded him. At Calpe the soldiers restored the former command structure and threatened death against anyone who again proposed dividing the army. Xenophon then told the soldiers that the journey would have to continue overland because they had neither ships nor provisions for a delay.[43]
In Thrace, Xenophon negotiated service under Seuthes. The ruler promised one Cyzicene to each soldier, twice that sum to captains and four times to generals, together with land, ox teams and a fortified place on the coast.[44] The offer gave the army a new employer and Seuthes an experienced infantry force. It did not provide an immediate settlement or a final homecoming. The route beyond the Black Sea was being determined by available pay, supplies and patrons rather than by a single collective destination.[42]
Seuthes did not promptly meet the wage promise. The soldiers blamed Xenophon for the unpaid money, while Seuthes became angry because Xenophon continued to press their claim; offers of coastal strongholds to Xenophon then ceased.[44] The dispute shows that command remained answerable to economic results. A general mediated between soldiers and employer, but could not make either side accept delay by authority alone. The force retained assemblies and officers, yet its cohesion depended on subsistence and credible prospects of pay.
In 399 BCE Spartan envoys arrived seeking troops for war against Persian satraps in Asia Minor. Thibron took over the remaining force, combined it with other Greek troops and campaigned against Tissaphernes and Pharnabazus.[45] The ending is therefore not a return of all survivors to their poleis. A substantial remnant passed from one employer to another and returned to western Achaemenid territory as part of a Spartan war. Xenophon's Hellenica confirms the transfer in compressed form, treating Cyrus's survivors as an available body of experienced mercenaries rather than as a nation come home.[46]
What the retreat shows about empire
The survival of the Ten Thousand became an argument that the Achaemenid Empire was militarily weak and ripe for invasion. The argument has an ancient pedigree, but it confuses the aims of the participants. Artaxerxes' immediate task in 401 BCE was not to annihilate every mercenary. It was to defeat his brother's claim, retain the throne and restore the allegiance of forces and provinces drawn into rebellion. He achieved those ends. Tissaphernes recovered Cyrus's Asian following and was restored to power in western Anatolia. The Greeks escaped without conquering a province, capturing a royal centre or installing a rival king.[47]
The route also displays reach rather than simple absence. Royal and satrapal forces mobilised across different commands; provincial troops waited at crossings; tribute horses were being raised in an Armenian village; mountain communities dealt intermittently with satraps; Greek cities on the Black Sea negotiated around imperial and local powers. Control was uneven, negotiated and sometimes resisted, but unevenness is not the same as non-existence. The empire could decide that channelling an armed residue into dangerous country was preferable to repeated close battle. That expectation proved wrong without proving that the imperial system had collapsed.[16][32][21]
J. G. Manning gives the methodological warning in memorable form: "Neither Artaxerxes nor Tissaphernes had read the Anabasis" and learned the later lesson that Persian armies were powerless against the Ten Thousand.[48] The apparent demonstration of weakness exists only because the survivors' commander wrote the story, later readers knew of Alexander and success became evidence that success had always been predictable. At the time, the Greeks were often hungry, divided, immobilised by missiles or bargaining for guides. Their eventual escape was an achievement under pressure, not the revelation of an obvious military law.
The army's experience does illuminate constraints on Achaemenid power. Heavy infantry that maintained cohesion was costly to destroy, and cavalry harassment did not always prevent adaptation.[17][18] Officials could not guarantee obedience in every mountain district, while large bodies of Greek mercenaries had become important participants in dynastic and interstate war.[21][47][45] But the same evidence reveals the empire's resources, the density of provincial politics and the danger of campaigning across its distances. Shane Brennan has argued that the Anabasis can be read as a negative paradigm for panhellenism: its scale, hazards and threats to Greek identity counsel caution as readily as conquest.[49] The evidence supports a narrower conclusion: the foreign contingent of a defeated rebel army improvised its survival, but the episode was not a rehearsal for a Macedonian future its participants did not know.
Primary sources
The ancient evidence, and what each source attests.
- Xenophon, Anabasis 2–7
- The continuous participant narrative from the negotiations after Cunaxa through the army's transfer to Thibron. It supplies the detailed sequence of commanders, marches, fighting, provisioning and coastal employment used throughout the account; quotations follow the held Brownson translation.[50]
- Xenophon, Hellenica 3.1.4–6
- A compressed account written by the same author, useful chiefly for the end of the story: Thibron's recruitment of Cyrus's surviving soldiers into the Spartan war against Tissaphernes and Pharnabazus in 399 BCE. It is not independent confirmation of the inland march.[46]
- Diodorus Siculus, Library 14.26.4–5; 14.27.4
- A later parallel for Artaxerxes' reward of Tissaphernes, the plan to recover Ariaeus and the point at which Tissaphernes left the Greeks at the Carduchian frontier. The held direct Diodorus translation begins with Book 15, so these Book 14 passages are used only through the held modern carriers, Kuhrt for 14.26.4 and Briant for 14.26.5 and 14.27.4. They do not independently reconstruct the highland route.[51][52]
- Plutarch, Artaxerxes 18
- A much later, Ctesias-derived story about Clearchus and the captured commanders in detention, their execution and manoeuvring at the royal court. It is evidence for later court-centred reception and variant storytelling, not an independent reconstruction of Tissaphernes' meeting or the northern route.[53]
- Achaemenid and Babylonian records
- Within the held source corpus, no contemporary royal inscription, administrative record or Babylonian chronicle narrates the retreat. That evidentiary silence does not prove that the episode was insignificant to every Persian official, but it leaves Xenophon's survivor-account without an indigenous narrative counterpart.[54][55]
How we know
The retreat survives under an extreme imbalance of testimony. Xenophon was participant, commander, advocate and literary craftsman. His Anabasis supplies the route, speeches, tactical sequence, ethnography and nearly every individual motive. The work is exceptionally rich evidence precisely because it is not a transparent campaign diary. Xenophon can report a decision made in an assembly, preserve an ordinary soldier's practical discovery and describe a village interior, then frame the same material so that his own foresight becomes the key to collective survival. Carol Atack's account of the crisis as Xenophon's "moral luck" is valuable because it neither dismisses his actions nor mistakes self-presentation for an unfiltered record: the catastrophe made possible the career whose virtue the narrator then demonstrates.[56]
The most persistent older reading treated the march as a revelation of Persian decadence. On this view, a small Greek force had penetrated the empire, defeated everything it met and shown the road that Agesilaus and Alexander would later take. Briant and Manning supply the essential correction. The dynastic rebellion was already over when the retreat began; Artaxerxes retained the throne, recovered Cyrus's Asian forces and used diplomacy, selective killing, pursuit and geography rather than seeking another Cunaxa. The Greeks survived but did not seize or hold imperial territory. Reading their survival backwards from Alexander substitutes hindsight for the choices available in 401 BCE.[57][48]
The opposite correction can also be overdrawn. Royal forces shadowed the Greeks with intermittent attacks but did not destroy the column before leaving it at the Carduchian frontier.[47] Farther north, Kuhrt's dossier shows Carduchians resisting royal authority while also making treaties with a lowland satrap, evidence for negotiated rather than uniform control.[21] Carduchians, Armenian villagers and Trapezuntians acted from local interests rather than as interchangeable imperial agents.[23][32][40]
Geographical reconstruction remains a field of confident disagreement. Tuplin and Brennan's analysis of Book 4 explains why: the narrative's proper names, distances and suppressed route choices do not permit a single exact line to be recovered from the Tigris to the sea. The location of Theches, the identity of the highland Phasis and the placement of several peoples remain disputed.[58][36] Maps should therefore distinguish secure points from conjectural corridors rather than converting one plausible reconstruction into a fixed route.
Reception has repeatedly remade the march. It has been read as adventure, military handbook, school text, autobiography, Socratic inquiry, imperial warning and panhellenic promise. Modern work represented by Brennan and the essays edited by Tim Rood and Melina Tamiolaki has shifted attention towards narrative form, apology, leadership and the unstable community made by the soldiers. That approach restores the untidy Books 5 to 7 to the event. The force does not complete a simple nostos to Greece; it remains an armed labour body seeking ships, cities, food and patrons. The evidential imbalance remains: Xenophon records the march in unmatched detail, but his arrangement around his own leadership and the limited testimony available for communities along the route must remain visible.[59]
References
Source key: primary ancient evidence · secondary modern scholarship · consensus an explicitly labelled overview, not a specific citation. Open Source details for the full record.
- ↑primary Xenophon — Anabasis 2.1.11–12; held Brownson physical page markers 63–64
Source details
Reference. Xenophon, Anabasis 2.1.11–12
Evidence used. Phalinus' statement that the king claimed victory after Cyrus's death and the Athenian Theopompus' reply that the Greeks retained 'arms and valour'. Translation quoted: C. L. Brownson, Loeb Classical Library 90 (Cambridge, MA, 1922), physical pp. 63–64 in the paged corpus
- ↑secondary P. Briant 2002 — p. 633
Source details
Reference. P. Briant, From Cyrus to Alexander (Winona Lake, 2002), p. 633
Evidence used. Tissaphernes' recovery of Ariaeus and Cyrus's Asian forces, the elimination of the Greek command and the effort to channel the survivors onto a northern itinerary expected to prove fatal
- ↑primary Xenophon — Anabasis 2.1.4–5; held Brownson physical page marker 62
Source details
Reference. Xenophon, Anabasis 2.1.4–5
Evidence used. the Greeks' offer to place Ariaeus on the throne and the argument that rule belongs to the victors; Brownson translation, physical p. 62
- ↑primary Xenophon — Anabasis 2.2.1; held Brownson physical page markers 66–67
Source details
Reference. Xenophon, Anabasis 2.2.1 — Ariaeus' reported refusal because higher-ranking Persians would not accept him as king; Brownson translation, physical pp. 66–67
- ↑primary Xenophon — Anabasis 2.2.8–10; held Brownson physical page markers 68–69
Source details
Reference. Xenophon, Anabasis 2.2.8–10
Evidence used. the reciprocal compact with Ariaeus, its mixed sacrificial ritual and the pledge to guide the Greeks; Brownson translation, physical pp. 68–69
- ↑secondary A. Kuhrt 2007 — pp. 366–368; Sources 26–27
Source details
Reference. A. Kuhrt, The Persian Empire: A Corpus of Sources from the Achaemenid Period (London, 2007), pp. 366–368, Sources 26–27 and commentary
Evidence used. the commanders' seizure, Tissaphernes' forces and the separation of the Greek contingent from Cyrus's former Asian army
- ↑ a bprimary Xenophon — Anabasis 2.3.18–29, 2.4.1–8; held Brownson physical page markers 73–79
Source details
Reference. Xenophon, Anabasis 2.3.18–29 and 2.4.1–8
Evidence used. Tissaphernes' offer of safe conduct and a market, the exchange of oaths and right hands, and the suspiciously separated march; Brownson translation, physical pp. 73–79
- ↑secondary G. Danzig 2007 — pp. 27–45
Source details
Reference. G. Danzig, 'Xenophon's wicked Persian, or What's wrong with Tissaphernes? Xenophon's views on lying and breaking oaths', in C. Tuplin (ed.), Persian Responses: Political and Cultural Interaction with(in) the Achaemenid Empire (Swansea, 2007), pp. 27–45
Evidence used. the construction of Tissaphernes' reputation, the modern Orientalising stereotype and Xenophon's prudential critique of perjury
- ↑primary Xenophon — Anabasis 2.5.29–32; held Brownson physical page markers 87–88
Source details
Reference. Xenophon, Anabasis 2.5.29–32
Evidence used. objections to trusting Tissaphernes and the coordinated seizure and killing of the Greek officers and attendants; Brownson translation, physical pp. 87–88
- ↑secondary A. Kuhrt 2007 — pp. 366–368; Source 26 (Xenophon, Anabasis 2.5.31–33, 2.6.1)
Source details
Reference. A. Kuhrt, The Persian Empire (2007), pp. 366–368, Source 26
Evidence used. Xenophon 2.5.31–33 and 2.6.1 on the commanders' capture and execution, with source-critical commentary
- ↑primary Xenophon — Anabasis 3.1.1–3; held Brownson physical page marker 97
Source details
Reference. Xenophon, Anabasis 3.1.1–3
Evidence used. the soldiers' reaction after the commanders' loss, the claimed distance from Greece, lack of guides and obstructing rivers; Brownson translation, physical p. 97
- ↑secondary C. Atack 2022 — pp. 63–81
Source details
Reference. C. Atack, 'Xenophon's Moral Luck: Crisis and Leadership Opportunity in Anabasis 3', in T. Rood and M. Tamiolaki (eds.), Xenophon's Anabasis and its Reception (Berlin and Boston, 2022), pp. 63–81
Evidence used. the crisis as an opportunity for Xenophon's self-presentation and the incomplete transformation of the mercenary force into a heroic community
- ↑primary Xenophon — Anabasis 3.1.38–47, 3.2.1–39; held Brownson physical page markers 106–116
Source details
Reference. Xenophon, Anabasis 3.1.38–47 and 3.2.1–39
Evidence used. appointment of replacement generals, the army assembly, proposals for discipline and the unanimous vote to abandon impedimenta; Brownson translation, physical pp. 106–116
- ↑primary Xenophon — Anabasis 3.2.36; held Brownson physical page marker 116
Source details
Reference. Xenophon, Anabasis 3.2.36
Evidence used. the hollow square and the placement of baggage and camp followers; Brownson translation, physical p. 116
- ↑primary Xenophon — Anabasis 3.3.1–2; held Brownson physical page marker 117
Source details
Reference. Xenophon, Anabasis 3.3.1–2
Evidence used. burning wagons and tents and abandoning superfluous baggage; Brownson translation, physical p. 117
- ↑ a bsecondary A. Kuhrt 2007 — pp. 367–368; Source 27 (Xenophon, Anabasis 3.4.13–16)
Source details
Reference. A. Kuhrt, The Persian Empire (2007), pp. 367–368, Source 27
Evidence used. Xenophon 3.4.13–16 on the composite royal and satrapal forces pursuing the Greeks, including Tissaphernes' cavalry, Orontas' contingent, Cyrus's former troops and royal reinforcements
- ↑ a bprimary Xenophon — Anabasis 3.3.6–10; held Brownson physical page markers 118–119
Source details
Reference. Xenophon, Anabasis 3.3.6–10
Evidence used. the inability of the rear guard to answer long-range missiles and a day's progress of no more than twenty-five stades; Brownson translation, physical pp. 118–119
- ↑primary Xenophon — Anabasis 3.3.3–5; held Brownson physical page markers 117–118
Source details
Reference. Xenophon, Anabasis 3.3.3–5
Evidence used. the answer to Mithradates: passage with the least possible damage if unhindered, and armed resistance if opposed; Brownson translation, physical pp. 117–118
- ↑primary Xenophon — Anabasis 3.5.15–17; held Brownson physical page marker 134
Source details
Reference. Xenophon, Anabasis 3.5.15–17
Evidence used. the northern route into Carduchian country, the claim that the inhabitants were not royal subjects and their treaty dealings with the satrap in the plain; Brownson translation, physical p. 134
- ↑ a b c dsecondary A. Kuhrt 2007 — pp. 837–839; Sources 3–4
Source details
Reference. A. Kuhrt, The Persian Empire (2007), pp. 837–839, Sources 3–4 and commentary
Evidence used. the Carduchians as a case of mountain resistance, treaty contact with lowland satraps and the limits of direct imperial control
- ↑ a bprimary Xenophon — Anabasis 4.1.4–10; held Brownson physical page markers 136–138
Source details
Reference. Xenophon, Anabasis 4.1.4–10
Evidence used. the attempt to pass without open conflict, limited initial provisioning, Carduchian evacuation and attacks; Brownson translation, physical pp. 136–138
- ↑secondary P. Briant 2002 — p. 633
Source details
Reference. P. Briant, From Cyrus to Alexander (2002), p. 633
Evidence used. the Persian expectation that the northern itinerary, natural barriers, mountain communities and forces in Armenia would prove fatal to the Greeks
- ↑primary Xenophon — Anabasis 4.3.1–7; held Brownson physical page markers 148–149
Source details
Reference. Xenophon, Anabasis 4.3.1–7
Evidence used. the Centrites boundary, the force of Orontas and Artuchas, and its Armenian, Mardian and Chaldaean contingents; Brownson translation, physical pp. 148–149
- ↑secondary A. Kuhrt 2007 — p. 651
Source details
Reference. A. Kuhrt, The Persian Empire (2007), p. 651
Evidence used. Orontas, royal son-in-law and governor of Armenia in 401 BCE; contextual control distinguishing him from the Orontas executed by Cyrus before Cunaxa
- ↑ a bsecondary P. Briant 2002 — p. 794
Source details
Reference. P. Briant, From Cyrus to Alexander (2002), p. 794
Evidence used. the Chaldaeans named with Orontas' Armenian force as 'Chaldeans of Armenia', their mercenary service, and the broader category of northern communities outside regular royal subjection
- ↑ a bprimary Xenophon — Anabasis 4.4.4–22; held Brownson physical page markers 154–158
Source details
Reference. Xenophon, Anabasis 4.4.4–22
Evidence used. Tiribazus as governor, the passage-and-provisions agreement, scouts' report of a gathering force and the Greek attack on his camp; Brownson translation, physical pp. 154–158
- ↑primary Xenophon — Anabasis 4.5.3–5; held Brownson physical page marker 159
Source details
Reference. Xenophon, Anabasis 4.5.3–5
Evidence used. a snow-covered bivouac, snow said to be a fathom deep, and deaths among animals, attendants and soldiers; Brownson translation, physical p. 159
- ↑ a b c d eprimary Xenophon — Anabasis 4.5.24, 4.6.1–3; held Brownson physical page markers 162–166
Source details
Reference. Xenophon, Anabasis 4.5.24 and 4.6.1–3
Evidence used. seizure of the Armenian villagers, village chief and seventeen tribute colts, and the chief's compelled service as guide; Brownson translation, physical pp. 162–166
- ↑primary Xenophon — Anabasis 4.5.25–27; held Brownson physical page markers 162–163
Source details
Reference. Xenophon, Anabasis 4.5.25–27
Evidence used. underground houses, animals, stored food and barley wine; Brownson translation, physical pp. 162–163
- ↑secondary C. J. Tuplin and S. Brennan 2022 — pp. 105–125, especially pp. 113–115
Source details
Reference. C. J. Tuplin and S. Brennan, 'From the Tigris to the Sea: The Problematic Geography of Anabasis Book 4', in T. Rood and M. Tamiolaki (eds.), Xenophon's Anabasis and its Reception (Berlin and Boston, 2022), pp. 105–125, especially pp. 113–115
Evidence used. suppressed route choice, paucity of securely identifiable names and the difficulty of reconstructing the march
- ↑ a b csecondary C. J. Tuplin and S. Brennan 2022 — pp. 106–108, 124–125
Source details
Reference. C. J. Tuplin and S. Brennan, 'From the Tigris to the Sea' (2022), pp. 106–108, 124–125
Evidence used. Larisa on the Tigris identified with Nimrud; large divergences among proposed routes at the highland Phasis; the possible Araxes/Pasinler context; and the continuing circularity of route and place identification
- ↑secondary C. J. Tuplin and S. Brennan 2022 — pp. 108–109, 121–126
Source details
Reference. C. J. Tuplin and S. Brennan, 'From the Tigris to the Sea' (2022), pp. 108–109, 121–126
Evidence used. Book 4 as landscape-driven narrative; the anonymous human landscape and evasive route discussions; and the limits of the stage-and-parasang record as a cartographic guide
- ↑primary Xenophon — Anabasis 4.7.21–25; held Brownson physical page marker 174
Source details
Reference. Xenophon, Anabasis 4.7.21–25
Evidence used. ascent of Theches and the cry 'The Sea! The Sea!'; Brownson translation, physical p. 174
- ↑primary Xenophon — Anabasis 4.7.25–27; held Brownson physical page markers 174–175
Source details
Reference. Xenophon, Anabasis 4.7.25–27
Evidence used. the emotional response, embraces and cairn after sighting the sea; Brownson translation, physical pp. 174–175
- ↑primary Xenophon — Anabasis 5.3.3; held Brownson physical page marker 194
Source details
Reference. Xenophon, Anabasis 5.3.3 — the count of 8,600 survivors at Cerasus and the stated causes of loss; Brownson translation, physical p. 194
- ↑ a bsecondary S. Brennan 2022 — p. 52
Source details
Reference. S. Brennan, Xenophon's Anabasis: A Socratic History (Edinburgh, 2022), p. 52
Evidence used. the rebellion ends in Book 1, while the promised homecoming is never completed within the narrative and settlement projects repeatedly remain unrealised
- ↑primary Xenophon — Anabasis 6.1.21–33, 6.4.9–13; held Brownson physical page markers 231–233, 245
Source details
Reference. Xenophon, Anabasis 6.1.21–33 and 6.4.9–13
Evidence used. the decision to appoint a single commander, the choice of Cheirisophus, his death and Neon's succession, followed by Xenophon's statement that the army lacked ships and provisions and must travel by land; Brownson translation, physical pp. 231–233 and 245
- ↑ a bprimary Xenophon — Anabasis 7.2.32–38, 7.5.4–10; held Brownson physical page markers 279, 295
Source details
Reference. Xenophon, Anabasis 7.2.32–38 and 7.5.4–10
Evidence used. Seuthes' promised pay, land and other rewards, followed by the soldiers' unpaid wages, their blame of Xenophon and his insistence that Seuthes pay; Brownson translation, physical pp. 279 and 295
- ↑ a bprimary Xenophon — Anabasis 7.6.1, 7.8.24; held Brownson physical page markers 369, 390
Source details
Reference. Xenophon, Anabasis 7.6.1 and 7.8.24
Evidence used. Thibron's envoys, promised pay and eventual takeover of the army for war against Tissaphernes and Pharnabazus; Brownson translation, physical pp. 369 and 390
- ↑ a bprimary Xenophon — Hellenica 3.1.4–6; held Brownson printed page marker 177
Source details
Reference. Xenophon, Hellenica 3.1.4–6
Evidence used. Thibron's arrival in Asia and the joining of the soldiers who had marched with Cyrus; used as a same-author parallel for the army's entry into Spartan service. Translation: C. L. Brownson, Loeb Classical Library 88 (Cambridge, MA, 1918), held in the page-marked corpus
- ↑ a b csecondary P. Briant 2002 — pp. 632–634
Source details
Reference. P. Briant, From Cyrus to Alexander (2002), pp. 632–634
Evidence used. Cunaxa as a clash of royal armies, Artaxerxes' political victory, Tissaphernes' recovery of Cyrus's Asian troops and the limits of using Greek survival as proof of imperial weakness
- ↑ a bsecondary J. G. Manning 2021 — p. 283
Source details
Reference. J. G. Manning, 'Armed Force in the Teispid-Achaemenid Empire', in B. Jacobs and R. Rollinger (eds.), A Companion to the Achaemenid Persian Empire (Hoboken, 2021), p. 283
Evidence used. warning against reading Persian military decisions through hindsight supplied by Xenophon's later narrative
- ↑secondary S. Brennan 2022 — p. 20
Source details
Reference. S. Brennan, Xenophon's Anabasis: A Socratic History (Edinburgh, 2022), p. 20
Evidence used. the Anabasis as a possible negative paradigm for panhellenism, emphasising the empire's intractability, resources and risks to Greek identity
- ↑primary Xenophon — Anabasis 2.1–7.8; held Brownson physical page markers 61–391
Source details
Reference. Xenophon, Anabasis 2–7 — the continuous participant narrative used for the retreat, from the first negotiations after Cyrus's death through the army's transfer to Thibron. Translation: C. L. Brownson, Loeb Classical Library 90 (Cambridge, MA, 1922), held in the page-marked corpus
- ↑primary Diodorus Siculus — Library 14.26.4; Kuhrt 2007, p. 366
Source details
Reference. Diodorus Siculus, Library 14.26.4
Evidence used. Artaxerxes' reward of Tissaphernes; the passage is mediated through A. Kuhrt, The Persian Empire (2007), p. 366
- ↑primary Diodorus Siculus — Library 14.26.5, 14.27.4; Briant 2002, p. 633
Source details
Reference. Diodorus Siculus, Library 14.26.5 and 14.27.4
Evidence used. the plan involving Ariaeus and Tissaphernes' withdrawal at the Carduchian frontier; the passages are mediated through P. Briant, From Cyrus to Alexander (2002), p. 633
- ↑primary Plutarch — Artaxerxes 18.1–7; held Perrin physical page markers 5488–5490
Source details
Reference. Plutarch, Artaxerxes 18.1–7
Evidence used. the Ctesias-derived captivity and royal-court story concerning Clearchus and the generals seized by Tissaphernes. Translation: B. Perrin, Loeb Classical Library, vol. XI (Cambridge, MA, 1926)
- ↑secondary A. Kuhrt 2007 — pp. 349–368, section 9A
Source details
Reference. A. Kuhrt, The Persian Empire (2007), pp. 349–368, section 9A
Evidence used. the held source dossier for the war of the brothers relies on Greek literary witnesses for the retreat and presents no contemporary Achaemenid royal, administrative or Babylonian narrative of it
- ↑secondary P. Briant 2002 — pp. 632–634
Source details
Reference. P. Briant, From Cyrus to Alexander (2002), pp. 632–634
Evidence used. the retreat reconstructed from Xenophon and Diodorus without a contemporary indigenous narrative counterpart
- ↑secondary C. Atack 2022 — pp. 63–81
Source details
Reference. C. Atack, 'Xenophon's Moral Luck', in Rood and Tamiolaki (eds.), Xenophon's Anabasis and its Reception (2022), pp. 63–81
Evidence used. the literary construction of Xenophon's crisis, leadership opportunity and temporary community
- ↑secondary P. Briant 2002 — pp. 632–634
Source details
Reference. P. Briant, From Cyrus to Alexander (2002), pp. 632–634
Evidence used. Artaxerxes' dynastic success, Tissaphernes' recovery of Cyrus's Asian forces and the strategic separation of the Greek and Asian contingents
- ↑secondary C. J. Tuplin and S. Brennan 2022 — pp. 105–125
Source details
Reference. C. J. Tuplin and S. Brennan, 'From the Tigris to the Sea', in Rood and Tamiolaki (eds.), Xenophon's Anabasis and its Reception (2022), pp. 105–125
Evidence used. why exact modern route reconstruction remains underdetermined
- ↑secondary T. Rood and M. Tamiolaki 2022 — pp. 1–17; compare Brennan pp. 20, 52
Source details
Reference. T. Rood and M. Tamiolaki, 'Introduction', in Xenophon's Anabasis and its Reception (Berlin and Boston, 2022), pp. 1–17
Evidence used. the work's literary sophistication, autobiographical and apologetic dimensions, disputed panhellenic purpose and diverse reception; read with Brennan 2022, pp. 20, 52
Cite this entry
“The Retreat of the Ten Thousand (401–399 BCE)”, in Achaemenica: An Encyclopaedia of the Achaemenid Persian Empire (entry the-retreat-of-the-ten-thousand), https://achaemenica.org/articles/the-retreat-of-the-ten-thousand, version of 2026-07-20.
Show BibTeX and RIS
@misc{achaemenica-the-retreat-of-the-ten-thousand,
author = {{Studio Daric}},
title = {The Retreat of the Ten Thousand (401–399 BCE)},
year = {2026},
howpublished = {\url{https://achaemenica.org/articles/the-retreat-of-the-ten-thousand}},
note = {Achaemenica: An Encyclopaedia of the Achaemenid Persian Empire. Version of 2026-07-20}
}TY - ELEC AU - Studio Daric TI - The Retreat of the Ten Thousand (401–399 BCE) T2 - Achaemenica: An Encyclopaedia of the Achaemenid Persian Empire PB - Studio Daric PY - 2026 DA - 2026/07/20 UR - https://achaemenica.org/articles/the-retreat-of-the-ten-thousand ER -
The article's metadata is embedded on the page, so Zotero, Mendeley and other reference managers detect it automatically.
Corrections and suggestions
Every entry is checked against its sources, but errors and omissions are always possible. If you have found an inaccuracy, or can suggest an improvement, please email a correction. Each note is read and checked against the sources before any change is made.
Related entries
Cyrus the Younger · The Battle of Cunaxa (401 BCE) · Artaxerxes II · Xenophon · Warfare & the Army · The Satrapy System · The Royal Road · Sardis · The Achaemenid dynasty · The King of Kings · Tissaphernes
Last updated 2026-07-20.