Read history

A meeting across town

Following Muhammad’s death in Medina, conventionally dated 8 June 632, members of the Ansar assembled with Sa‘d ibn Ubada at the Saqifa of Banu Sa‘ida to decide who should lead the community. Ali and Muhammad’s household were occupied with the body and did not take part.

Abu Bakr, Umar ibn al-Khattab and Abu Ubayda ibn al-Jarrah went to the gathering, where some of the Ansar proposed one leader from the Ansar and another from the Meccan emigrants.

The hand that settled it

Abu Bakr argued that leadership should remain with Quraysh and offered Umar or Abu Ubayda. Umar instead took Abu Bakr’s hand and pledged allegiance to him, and others present followed.

That was not the community-wide inauguration. A hadith account places the broader public pledge at the mosque’s pulpit on the following day.

Contested from the start

The Saqifa decision established Abu Bakr as Muhammad’s political successor — leadership of the community, not succession to prophethood. The gathering did not include every interested party, and it became central to later Sunni–Shia disputes over legitimate succession.

The accounts were also written down later. A comparative University of Leeds study finds that medieval historians selected and shaped the material according to developing theological positions.

Midday in the valley

On the Day of Arafat during his Farewell Pilgrimage — 9 Dhu al-Hijjah 10 AH, converted to 6 March 632 in the Julian calendar — Muhammad left his tent at Namira after midday, rode his camel al-Qaswa to the bottom of the valley and addressed the assembled pilgrims.

In Jabir ibn Abd Allah’s account he declared the audience’s blood and property inviolable, abolished pre-Islamic blood-revenge claims and usury, addressed marital obligations, and told the listeners to hold fast to the Book of God.

Three calls to witness

Muhammad asked whether he had conveyed his message. After the audience affirmed that he had, he raised his forefinger toward the sky and pointed it toward the people, calling on God to witness — three times.

Bilal then called the prayers. Muhammad led the noon and afternoon prayers consecutively and remained standing at Arafat until sunset; the pilgrimage continued rather than ending with the speech.

The version circulating today is a composite

There is no single contemporary verbatim transcript. Later sources preserve overlapping selections, and modern “full sermon” texts often splice material from addresses reportedly given at Arafat, at Mina or elsewhere during the pilgrimage.

The strongest Jabir recension says Muhammad left his listeners the Book of God; it does not carry the expanded phrase “the Quran and my Sunnah.”

A report attributed to Umar says Quran 5:3 was revealed while Muhammad stood at Arafat on a Friday — but Sahih al-Bukhari does not say he recited it during the sermon.

No event is recorded on that day

No consulted scholarly or primary source identifies a proclamation, treaty, final battle or “official recognition” of Arabian unity on 18 March 633.

The date looks like an artifact of the calendar. The SUNY translation of al-Tabari assigns AH 11 the civil span 29 March 632 to 17 March 633, and 18 March is simply the day after that span closes.

What the sources do support

Modern educational scholarship supports only the broader statement that Abu Bakr’s campaigns had brought the entirety of Arabia under the first Islamic state’s control by 633. OpenStax gives no month and no day.

The translated al-Tabari volume likewise treats the subjection of Arabia as the cumulative result of the Ridda campaigns rather than a single-day constitutional act.

And 633 may still be early

Encyclopaedia Iranica places the defeat and death of Musaylima — central to the traditional Ridda narrative — in May or June 633, months after the supposed completion date, and warns that early conquest accounts are contradictory and precise dates elusive.

“United” also covers very different outcomes: conquest, coerced tribute, negotiated agreements, and the incorporation of groups never subject to Muhammad.

The translated al-Tabari volume describes rebellions against Muslim domination, and modern historians also identify political and fiscal disputes rather than religious apostasy in every case.

A designation written in advance

In the standard chronology Abu Bakr died in Medina on 22 Jumada II 13 AH, conventionally converted to 23 August 634. Traditional narratives say he had been ill with fever and bedridden for about two weeks; the medical cause cannot be established.

Before his death he arranged a written designation of Umar ibn al-Khattab as successor, and al-Tabari preserves reports that the document was read publicly and accepted. Some advisers reportedly worried that Umar was too harsh.

A succession with no Saqifa

Because that designation already existed, leadership passed without the kind of open succession meeting that had followed Muhammad’s death. Umar became the second caliph, inheriting an operating government and two active war fronts in Syria and Iraq.

Traditional accounts say Abu Bakr was buried in the room or house of his daughter Aisha, beside Muhammad.

How firm the date is

The 23 August 634 date is the standard traditional chronology, used by Wilferd Madelung, and derives from 22 Jumada II 13 AH. Some older works give 22 August, because Arabic dates begin at sunset and differing conversions can shift the civil date by a day.

Early sources also transmit other dates. A modern Journal of Islamic Studies analysis notes proposals in July or August 635, based on alternative reports and chronological synchronisms.

Four forces, four frontier districts

In the opening phase of 634, Abu Bakr dispatched separate forces toward Byzantine territory: Amr ibn al-As in southern Palestine, Shurahbil ibn Hasana in Jordan, Yazid ibn Abi Sufyan around the Balqa and Abu Ubayda ibn al-Jarrah in the Golan. Fred Donner puts the combined strength at about 24,000.

The sources do not establish a single overall commander, nor agree on the order in which the commanders departed or whether Abu Ubayda went with the first dispatch. Donner finds a general consensus only around early spring 634; the armies may have set out in late 633.

The first fighting, near Gaza

Amr’s force reached Dathin and Badan near Gaza. When negotiations with Gaza’s military commander failed, a skirmish followed and the Muslim force prevailed.

Abu Bakr then ordered Khalid ibn al-Walid to leave his base at al-Hira in Iraq and reinforce the armies already campaigning in Syria.

A famous march with no agreed map

Khalid emerged near Damascus and joined the other Muslim armies around Bosra. His route and chronology cannot be reconstructed securely — scholars cannot choose confidently between itineraries involving Palmyra and Dumat al-Jandal, and medieval chroniclers appear to have moved incidents from his earlier Iraqi campaigns into the story.

Reported strengths for the reinforcement run to 500, 800, 3,000, 7,000, 9,000 or 10,000, with no secure figure. His departure also did not mean Iraq was finished: Sasanian forces soon recovered ground and defeated a Muslim army at the Battle of the Bridge.

Two armies on the desert edge

At al-Qadisiyya, a frontier settlement southwest of al-Hira, an Arab Muslim army commanded by Sa‘d ibn Abi Waqqas faced a larger Sasanian army under Rostam Farrokhzad. The two forces stood opposite one another for months before the fighting began. The battle is conventionally dated to 636, but proposed dates range from 634/635 to 637.

Sa‘d’s coalition combined a Medina core force of about 4,000 with Arabian tribal levies and reinforcements from Iraq and Syria. Encyclopaedia Iranica assesses plausible upper bounds of no more than about 12,000 Arab Muslims and 30,000 Iranians.

Elephants, dust, and a severed head

The traditions describe several days of combat and the disruptive use of Sasanian war elephants, which initially frightened the Muslim cavalry and dominate the surviving battle stories.

In the Islamic narrative a dust storm rose during the final fighting, Rostam was killed and beheaded, and the announcement of his death precipitated the collapse of the Sasanian army. The details around that death differ among accounts.

The date is not settled

The victors pursued survivors toward al-Mada’in, opening the central Iraqi alluvium to Muslim occupation — though not, by itself, ending Sasanian resistance.

Islamic chroniclers variously place the battle in AH 14, 15 or 16, and modern proposals include 634/635, 636 and 637. The familiar 16–19 November 636 is a conventional reconstruction, not an attested day-by-day chronology.

Even the duration is unclear: three days in the traditional account, with fighting continuing into the night after the third — while some sources give no duration at all.

A chain of cities, not one wall

Following al-Qadisiyya, Sa‘d ibn Abi Waqqas advanced across central Iraq in 637 toward al-Mada’in — “the cities” — the Sasanian capital complex of several cities and suburbs spread along both banks of the Tigris.

Muslim forces first reduced the western-bank city called Bahurasir or Veh-Ardashir. The Sasanian court and army then evacuated Ctesiphon on the eastern bank.

Across a river running high

The Persians removed the boats and burned or disabled the bridge. Sa‘d’s soldiers crossed the swollen Tigris by a ford and on horseback, and the eastern capital was occupied without another set-piece battle.

Yazdegerd III fled toward Hulwan with his family, leading dignitaries and movable valuables. The occupiers still seized substantial royal treasure left behind, including the celebrated garden carpet known as the Spring of Khosrow.

How long, and how many

March 637 is the standard date given by Encyclopaedia Iranica, but the chronology rests heavily on contradictory later Arabic narratives, and some modern reconstructions place the campaign later.

The reported length of the operations ranges from two to twenty-eight months across the source tradition; Baladhuri alone gives nine months at Bahurasir, with an alternative report of eighteen.

No reliable consulted source verifies an army of exactly 15,000.

Morning prayer in his own capital

Early one morning in 661, while Ali ibn Abi Talib was praying in a mosque at Kufa — the city he had made his capital — the Kharijite Abd al-Rahman ibn Muljam struck him on the head with a poisoned sword. The attack was reportedly intended to avenge Kharijites killed by Ali’s forces at Nahrawan.

Ali survived the initial blow. He died from the wound about two days later, at a reported age of 63.

Which day, exactly

Early sources variously place Ali’s death — or sometimes the attack — on 17, 19 or 21 Ramadan 40 AH. Modern conversions consequently range across roughly 26 to 28 January 661 and sometimes later; Encyclopaedia Iranica dates the death to 27 January while acknowledging the conflicting Ramadan dates.

So 26 January cannot be presented as certain, and the distinction between the day of the blow and the day two days later when Ali died is frequently blurred.

What happened to the caliphate

After Ali’s death, Kufans acclaimed his son Hasan as caliph — that same day or two days later, depending on the report. Muawiya, who controlled Syria, contested the claim, and his uncontested rule followed Hasan’s later surrender agreement rather than Ali’s death.

Ali’s burial was kept secret; the sanctuary at Najaf grew around a tomb identified generations later. Encyclopaedia Iranica warns that much of Ali’s biographical material is polemical.

The traditional story

In the traditional account, the supposed 674–678 siege ended after Arab fleets failed to take Constantinople and suffered a disastrous shipwreck near Sillyon while withdrawing. Nikephoros has those fleets fighting repeated seasonal naval campaigns near the city across what he calls a seven-year war, wintering at Cyzicus.

News of that disaster, Nikephoros reports, prompted the Arab ruler to send envoys to Constantine IV and seek a tributary peace — a sworn settlement of thirty years.

Two episodes, joined later

Source-critical analysis indicates that Theophanes conflated the storm-driven wreck at Sillyon with a separate Byzantine land and naval victory in Lycia involving Greek fire. Sillyon lay on the southern Anatolian coast, not at Constantinople, and Nikephoros attributes the destruction there to storm and shipwreck rather than to fire.

Theophanes and Nikephoros both wrote more than a century after the events, synthesizing earlier and partly lost sources.

Whether there was a siege at all

Marek Jankowiak’s revision dates the first major Arab attempt on Constantinople to 667–669, led by Yazid; on that reconstruction there was no discrete 678 siege-ending engagement of the kind this entry describes. James Howard-Johnston has likewise challenged the conventional multi-year siege narrative.

So 678 is at best an approximate traditional endpoint, with no securely established autumn day for a head-on engagement.

Whatever happened did not close the question either: another major Arab siege of Constantinople followed in 717–718.

A small force, mostly Berber

In approximately April or May 711, Tariq ibn Ziyad took a small force across the Strait from North Africa to southern Iberia. It was probably no larger than 7,000 men, predominantly Berber, with Arabs forming only a small minority.

This was not the first Muslim landing in Iberia — Tarif had led a reconnaissance and raiding expedition the previous year, and its success encouraged the larger crossing. Hugh Kennedy also argues that Tariq needed conquest and booty to reward and retain newly recruited Berber troops.

A bridgehead with a way back

Tariq’s men secured the “Green Island” at modern Algeciras as a base that also preserved a route of retreat to North Africa. He did not push inland.

Roderic, the Visigothic king, began moving south after learning of the landing, while Tariq stayed near his coastal base and requested reinforcements before the decisive summer battle.

The date, and the legends

The exact landing day is not securely recoverable. A traditional Muslim chronology places Tariq’s arrival and encampment in the Calpe/Gibraltar area on 5 Rajab 92 AH, which one modern reference converts to 28 April 711; other accounts give 26 or 27 April, and the conversion depends on assumptions about the historical lunar calendar.

The Cambridge History of Islam is more cautious still, dating the encampment at Calpe and the capture of nearby Carteia only to Rajab 92 AH, or spring 711.

The detailed Arabic narratives were written well after the conquest, and the near-contemporary Chronicle of 754 supplies only a sparse outline. Kennedy states that neither Count Julian’s existence nor the story attached to him can be verified.

A battle without a located battlefield

At an uncertain location in southern Iberia, an Umayyad force commanded by Tariq ibn Ziyad defeated the army led by the Visigothic king Roderic. Later Arabic sources place it on 19 July 711; the near-contemporary Chronicle of 754 instead assigns it to 712.

Suggested sites include the Guadalete or Barbate valleys, the former La Janda lagoon, Medina Sidonia and the Chronicle’s unidentified Transductine Mountains. “Guadalete” is a conventional label, not a secure identification.

What the earliest account says

The Chronicle of 754 says the Gothic army fled during the battle and that Roderic was killed, along with rivals who had accompanied him in competition for the kingship.

Tariq’s force was predominantly Berber. One modern reference drawing on Arabic traditions gives 7,000 men including only about 300 Arabs, plus 5,000 reinforcements sent before the battle — roughly 12,000 in all.

The opening of a conquest, not its completion

After the victory Tariq moved east along the Guadalquivir valley toward Écija and Córdoba. The captures of Córdoba, Seville and Toledo were subsequent operations, not events of the battle day.

The 33,000-man Visigothic army in this entry is not securely sourced; early medieval army figures in later narratives are especially unreliable.

The later traditions about searching for Roderic and finding only his horse, saddle, cloak or sandal are likewise not contemporary evidence.

Seven days of watching

Traditionally dated to 10 October 732 — though many scholars prefer late 733 or 734 — the battle followed Abd al-Rahman al-Ghafiqi’s invasion of Aquitaine and his defeat of Duke Eudo. Eudo sought help from Charles Martel, and the Frankish and Umayyad armies met on the road between Poitiers and Tours. The Chronicle of 754 describes seven days of manoeuvring or harassment before the pitched battle.

Charles’s forces held their formation and killed Abd al-Rahman during the fighting. The chronicler calls Charles’s troops “Europeans” — one of the earliest surviving Latin uses of that collective label.

The empty camp

Fighting ended at nightfall. At dawn the Franks found the Umayyad tents still standing, but the army had withdrawn in good order.

Charles had won the field. Muslim forces nonetheless retained positions in southern Gaul and conducted later raids.

How little is actually known

10 October 732 is the traditional exact date, not a contemporary attestation; Oxford Bibliographies notes that many scholars prefer late 733 or 734 on the basis of Iberian chronology. The precise battlefield is unknown beyond the broad area between the two cities.

No trustworthy troop-strength or casualty totals survive, and the composition of Charles’s army is incompletely recorded — “Frankish and Burgundian” is not a precise order of battle. Even Abd al-Rahman’s aim, whether large raid, punitive campaign or territorial conquest, remains debated.

Charles did not demonstrably receive the name Martel that day either: the epithet was recorded later and was not linked to this battle until the eleventh century.