The historical role of Khurāsān

In order to understand just how irreparable a loss the eclipse of Khurāsān was to Islamic civilization, one must first understand the region's unique role in the centuries leading up to the fall of the Great Seljuq Sultanate. The collapse of ʿAbbāsid political rule in the Islamic East in the mid-ninth century marked not only the permanent political autonomy of Khurāsān and its Central Asian dependencies, but also the opening of an outstandingly brilliant period in its history. Over the ensuing three centuries, Greater KhurāsānFootnote ⁶ became the seat of the leading political and military powers of the Sunni world, and assumed primacy in its cultural, religious, and intellectual life. The dynasties based here – the Ṣaffārids, Sāmānids, Ghaznavids and Seljuqs – not only provided a military and political bulwark against the non-Sunni groups dominating more westerly lands, but were also the prime bearers of the banner of jihād and territorial expansion in this period.Footnote ⁷

Politically, the culmination of this era was reached during the Great Seljuq Sultanate, which is remembered in medieval Islamic historiography as the high point of the post-caliphal period stretching between the ninth century and the coming of the Mongols – understandably, since the territorial extent of this dynasty's realm was far greater than that of any other Sunni dynasty had been since the political crumbling of the unitary caliphate.Footnote ⁸

This was widely recognized by the medieval chroniclers. Thus, for instance, Nīshāpūrī writes:

While statements of this sort are useful for establishing the relative status of the Seljuqs, and the regard in which they were held, they are rather general; fortunately, al-Ḥusaynī’s encomium provides more specific supporting detail:

As can be seen from this last description, Khurāsān between the ninth and the twelfth centuries – including the twelfth century, an important point to which we shall be returning – was also one of the wealthiest of the Sunni lands: not just from agriculture, manufacturing, and mining, but also from, for instance, Sāmānid, Ghaznavid, and Seljuq conquests in the Central Asian steppes, India, and the Caucasus, and from trade with both Europe and eastern Asia.Footnote ¹²

Most importantly, the efflorescence of Khurāsān during these three centuries was not purely physical: throughout this period Khurāsān and its associated territories in Islamic Central Asia constituted the religious, cultural, and intellectual heart of the Islamic world as well. In the religious sciences in particular, Khurāsān is remarkable for having produced the most outstanding legal scholars, theologians, and Sufīs, culminating in the brilliance of the Seljuq period, which witnessed the activities of luminaries such as al-Qushayrī, Imām al-Ḥaramayn al-Juwaynī, al-Ghazālī, and Ahmed-i Jām. All of this dazzling cultural and intellectual production, and Khurāsān's pre-eminent place in the Sunni Islamic world, came to a rather abrupt end in the middle of the twelfth century; the question is why.

Richard Bulliet's “Big chill theory”: problems of conception and scientific evidence

The most popular theory that has been advanced to account for the eclipse of Khurāsān – and, in fact, the only theory to date in monographic formFootnote ¹³ – one that is perhaps anachronistically influenced by present-day concerns, is that proposed by Richard Bulliet in Cotton, Climate and Camels. In his own summary of his thesis:

In other words, Bulliet is positing that the mid-twelfth-century eclipse of Khurāsān was due to a gradual and progressive economic collapse over the course of a century and a half, caused by a catastrophic decline in both the cotton trade and what he thought to be the camel-dependent Turkic steppe nomadic economy as a result of supposed climatic cooling in Central and Western Asia; which in turn caused, and also combined with, the nomadic invasions from the steppes that destroyed Khurāsān. Let us leave aside for the moment the question of whether or not Bulliet is overstating the importance of cotton as opposed to all other agricultural and manufacturing products, and also the problem that the large-scale nomadic invasion from the steppes, after the initial destruction and dislocation, actually brought about prosperity rather than decline for over a century. Khurāsān, according to eyewitness testimony, thrived throughout most of the Seljuq period, yet the Seljuqs and a goodly portion of their armies and followers were Turkmen nomads.

Let us, rather, focus for the moment on the heart of his argument, which is that Khurāsān experienced a “Big Chill” or climate cooling at this time, and that the steppe migrations beginning in the eleventh century, culminating in the complete eclipse of Khurāsān in the mid-twelfth century, were due primarily to this supposed cooling, and a concomitant human inability to adapt to it over the course of time.Footnote ¹⁴

The first problem with Bulliet's thesis is that he does not establish that there even was a “Big Chill” in Khurāsān, at least of the type he posits. Let us examine Bulliet's treatment of the rather scarce and problematic scientific evidence (although as we shall see below, there are several ice core studies he did not utilize that are germane to the issue). Essentially, his theory rests on two studies of tree rings from Western Mongolia, both conducted at the Tree-Ring Lab at the Lamont-Doherty Earth Observatory of Columbia University. However, scientifically speaking, the tree ring record from Western Mongolia is not indicative even of the climate in Mongolia as a whole, let alone of the climate some four thousand kilometers awayFootnote ¹⁵ – as Bulliet himself observes, “Western Mongolia is a long way from Irān”.Footnote ¹⁶ In any given era, there is a great deal of regional variation, even across far smaller distances, in aridity and precipitation.Footnote ¹⁷

Moreover, Bulliet himself admits the speculative nature of his interpretive use of this very limited data:

It should be noted, however, that his interpretation of what effect a possible cooling of a degree or two on the Central Asian steppes might have had on the nomads living there is also open to doubt. Even today, in an era of global warming, average winter temperatures on the steppes of Uzbekistan average around −8° Celsius/18° Fahrenheit, and in the Fergana valley of Kyrgyzstan, −10° C/10° F. As anyone living in, say, the American Midwest can aver, if one is prepared for −10° C one is also prepared for −13° C; and the differences under consideration here were smaller than that.Footnote ¹⁸ Any climate variability, in other words, that did not cause an Ice Age or desiccation to the point of desertification is not enough to explain the wholesale eclipse of Khurāsān – and this still leaves aside the question of human adaptability to gradual processes of weather and climate change.Footnote ¹⁹

The second problem with the “Big Chill” theory is that it depends on what Jürgen Paul has shown to be an unwarranted assumption: namely, that the Oghuz specialized to a large extent in camel breeding, and that these particular herds were sensitive to cold.Footnote ²⁰ Paul has marshalled convincing evidence, however, that “the Seljuq Ghuzz held a mixed herd of the well-known Central Asian type” (Paul Reference Paul2016: 514), comprising mostly sheep, but also, in descending order of importance, goats, horses, cattle (which seem to have been the only animal absent in the Seljuq case), and camels (less than 3 per cent of the overall herds). It is clear from the literary sources, moreover (including those regarding Oghuz tax payments in the critical period, the mid-twelfth centuryFootnote ²¹) that sheep “were the ‘bread and butter’ business, the animals by which the group survived. And there can be little doubt that sheep indeed made up a very large proportion of Ghuzz herds.”Footnote ²² Whenever the Seljuq mounts are described, moreover, they are named as horses, not camels. In fact, a contemporaneous anonymous author, writing in the year ad 982 of “the Ghuzz Country”, states explicitly: “Their wealth is in horses, cows, sheep, arms, and game in small quantities” (Anon. Reference Minorsky and Bosworth1982: 100); camels are not mentioned.

The “Big Chill” and the literary evidence

The third problem with the “Big Chill” theory, and the most critical for a historian, is that the literary evidence cited to bolster it also falls short, both in its nature and in the historical use made of it. Our examination will therefore focus on this evidence – the actual historical memory of both Seljuq rule and the eclipse of Khurāsān in the mid-twelfth century – and the manner in which it has been used, supplemented by the scientific studies and primary sources that have not previously been considered.

The literary evidence Bulliet adduces for the period is scant, consisting essentially of a handful of citations from Ibn al-Jawzī’s chronicle (of which only two are unambiguously exceptional weather events); two citations, of dates only one year apart, in a translation of Ibn Faḍlān's travelogue; and a scattering of individual citations from Bayhaqī and ʿUtbī – all limited to a period that is too early, from 920–1029.Footnote ²³ The attestations from Ibn al-Jawzī dating to the years 1027 to 1029 can be immediately dismissed, since ice core research has shown that a major eruption of the Baitoushan volcano (located between today's North Korea and China) led to a sulfate spike in 1027, producing between 1028 and 1032 in the northern hemisphere the coldest winters in a thousand-year record (Oppenheimer Reference Oppenheimer2003: 424). This was, in other words, demonstrably a short-lived volcanic winter weather anomaly;Footnote ²⁴ this is a good illustration of one of the pitfalls of attempting to hypothesize about climate using anecdotal evidence from chronicles: climate is a scientific matter requiring, for instance, systematic ice-core analysis, and cannot be accurately evaluated by gleaning references to anomalous weather events recorded by chroniclers and travellers.Footnote ²⁵

Let us now examine the other reports from Ibn al-Jawzī, together with the two attestations from Ibn Faḍlān. One immediately notices that they are, first, all far too early: Ibn Faḍlān's account, together with at least two of Ibn al-Jawzi's, refers to snow in the mid-920s. The relevant passage from Ibn al-Jawzī refers to snowfall in Baghdad in the year 926 – but, once again, the unusually cold weather of this year is known from ice core research to have been tied to volcanic action rather than climate change (Castellano et al. Reference Castellano2005). The second instance reported by Ibn al-Jawzī refers to an isolated severe weather event, snowfall in Baghdad in 1007. Once again, this is far too early (as well as geographically too distant from Khurāsān), for Bulliet's timeline; if the climate had worsened significantly and persistently to a degree that affected agriculture and pastoral nomadism, then there should have been similar attestations from the later eleventh and twelfth centuries when, according to Bulliet's hypothesis, the cooling culminated, and these attestations should have become more, not less, frequent and of greater reported severity; further, it would not take a century and a half for the effect of drastic climatic worsening to show up – it does not take one hundred and fifty years for camels to get cold or cotton to die – and yet the eclipse of Khurāsān transpired in the mid-twelfth century, not when Ibn al-Jawzī describes the snow as having fallen. In fact, one of the few studies undertaken by climatologists involving medieval Baghdad has shown that while there was perhaps a period of cooling, this occurred in the tenth century, in accord with Ibn al-Jawzī’s evidence, not in the eleventh and twelfth, which were significantly warmer, not cooler (Dominguez-Castro et al. Reference Dominguez-Castro2012: 76–82).

Another significant problem with adducing particular instances of severe cold weather, such as the Baghdadi winter of 1007, as evidence for climate cooling is that isolated weather events do not equal climate: Bulliet states that “Today a five-hour snow shower in Baghdad would be as startling and memorable as it was in 904/291” (Bulliet Reference Bulliet2009: 71). But in our own day, snow did indeed fall in Baghdad in 2008 for the first time in 100 years; and Jerusalem experienced in 2013 a record 8 inches of snowfall, and freezing temperatures for five consecutive days – yet we live in an era of global warming, not of a “Big Chill”. Most importantly, moreover, Baghdad, although closer than Mongolia, is still geographically too far away from Khurāsān to be of relevance in this type of climate discussion; the literary attestations for Khurāsānian climatic severity should not have to be dredged up from places that were a thousand miles distant.

The other, more ambiguous, literary evidence Bulliet cites evinces similar analytical problems. To wit, his primary literary evidence apart from the extreme weather events consists of accounts of crop failure and famine; yet Bulliet does not examine the historical context of those reported famines. Such an examination of the broader context, however, suggests that these reported cases might well be due more to human action than to climate factors. Thus, Bayhaqī’s report regarding the ruins in which Nīshāpūr lay in 1040, which Bulliet interprets as having been due to climate change (Bulliet Reference Bulliet2009: 81), was surely due not to extraordinary weather, but rather, first, to ongoing Ghaznavid misgovernment and mulcting of Khurāsān, the result of which was that “the land became ruined and the population dispersed”;Footnote ²⁶ and, second, to the prolonged Seljuq conquest and years of warfare and marauding that preceded it, which meant there was no agricultural margin for the inevitable bad harvest or unfavourable weather event, and that supply lines and the agricultural hinterland had been ruined.Footnote ²⁷ Bayhaqī, in fact, speaks of physical destruction, which was obviously due to warfare, marauders, and successive occupations, not to cold: “Nīshāpūr this time was not as I had seen it [in the past]; everything had been destroyed, and little wealth [or] population remained” (Bayhaqī: 738).

Moreover, when one examines the immediate context of the passage Bulliet cites, one sees that this food scarcity clearly refers to Nīshāpūr alone, and was wrought by the Seljuq occupation of the city with their army; this scarcity is contrasted with the abundance found in parts of Khurāsān that had not come under Seljuq occupation or attack, most notably Ghazna; it is thus clear that the cold is mentioned merely as an additional hardship, not as the root cause of the ruinous state of the city. Bayhaqī in fact contrasts the dearth of food in war-plagued Nīshāpūr with the abundance of food in Ghazna at the very same moment: “We have such a huge amount of corn in Ghazni, and such dearth and want here!” (Bayhaqī: 741; trans. Bosworth Reference Bosworth2011: 2: 301). In light of this reported abundance of food crops in parts of Khurāsān that were not war-torn, Bulliet's interpretation of Bayhaqī’s passage becomes untenable; for all the cases of famine during the 1030s that he cites are limited to war-torn and Turkmen-ravaged areas of Khurāsān, and abundance is reported where the warring armies and raiders did not reach.Footnote ²⁸

This lack of historical contextualization also leads Bulliet in some cases to confuse cause and effect, attributing, for instance, the Seljuq civil war over the succession (1092–1105) that ensued upon Malikshāh's untimely death, to “a heightened competition for dwindling resources” (Bulliet Reference Bulliet2009: 86). This is implausible: not only are there numerous instances throughout human history of succession struggles among underage child princes and their regents after the untimely death of a sovereign, occurring in every era and therefore not attributable to “Big Chills”, but also, the restoration of law and order in Khurāsān after 1106 and the resultant second Seljuq golden age under Sanjar thereafter, simply could not have resulted in the great prosperity noted by all the literary sources, had climate-induced food scarcity been the cause of the previous disorders.Footnote ²⁹ Sanjar's time should in fact have been even more disorderly and poverty-stricken according to Bulliet's theory, but according to the historical record the contrary is the case.Footnote ³⁰

Another case of interpretation divorced from the necessary historical context is Bulliet's citing of al-Ghazālī’s letter of 1106 asking for tax remission for the people of Ṭūs: the reasons Bulliet quotesFootnote ³¹ are that the people have “suffered boundless injustice, and [their] grain was destroyed by cold and drought” – note that even in this excerpt, the primary blame is the “boundless injustice”; the cold was just the coup de grace.Footnote ³² The paragraph from the letter reads as follows:

To what does that “injustice and fate/misfortune” refer? The answer is given in the same letter collection, in a missive al-Ghazālī addressed to Sanjar's vizier Fakhr al-Mulk, in which primary blame for the problems of Ṭūs is clearly attributed to bad government and “oppression”:

Wholly apart from what appears to have been the problem of specific local officials and notables, one must not ignore the fact that, in addition to the brutal and highly destructive civil war that was ongoing until the death of Berkyaruq in 1105, there was also an Ismaili campaign of destruction taking place at this time – and this almost surely played some kind of role in what al-Ghazālī is describing.

We discover by opening any chronicle of the period that precisely this area, in the years leading up to al-Ghazālī’s request, had experienced devastation from unchecked Ismaili activity that had flourished in the disorder caused by the long Seljuq civil war – indeed, warring against the Ismailis was to be one of the first campaigns Sulṭān Muḥammad Ṭapar undertook in 1105–06 after the death of his brother and the end of the succession struggle that had lasted for over a decade.Footnote ³⁴ Regarding this Ismaili activity, Ibn al-Athīr reports that a large group of Ismailis, originally from Bayhaq:

Clearly the environs of Nīshāpūr and Ṭūs were among those areas of Ismaili activity, since in the very year of al-Ghazālī’s request, 1106, in Nīshāpūr itself, this very same vizier, Fakhr al-Mulk, Niẓām al-Mulk's most competent son and Sanjar's vizier, was assassinated by Bāṭinīs.Footnote ³⁵

In other words, to put these letters and al-Ghazālī’s request in perspective: if the agricultural areas around Aleppo in 2017 were to have experienced an unusually severe winter and suffer a poor harvest, it would be historically and climatologically indefensible to blame this on a supposed “Big Chill” rather than the ongoing Syrian civil war. The only additional literary source from the Seljuq period adduced in support of the “Big Chill” theory is one sole report in Ibn al-Athīr that cold weather in the winter of 1098 caused a rise in prices (Bulliet Reference Bulliet2009: 84), something that can be found periodically in any annalistic record, whether it be written during a period of climate warming, cooling, or stasis.

The historical memory of the later Seljuq period: the twelfth century and Sanjar

The most important objection to the “Big Chill” thesis, however, is that it stands in opposition to the testimony of the written primary sources for the twelfth century, both in its timeline and in its attribution of cause. If there really had been a one hundred and fifty year period of cooling in Western Asia, with the accompanying adverse agricultural effects that the theory posits, reaching its apex in the twelfth century, then Khurāsān throughout the Seljuq period, but especially in the first half of the twelfth century under Sanjar, should have witnessed increasing agricultural difficulties and impoverishment; indeed, the Seljuq period in its entirety should have been recorded by medieval historians as one of cold, penury, and dwindling resources; and both Malikshah's and Sanjar's periods should have been more straitened than, say, Toghril Beg's and Alp Arslān's – yet, as we shall see, both the people of the time and those who lived in succeeding generations reported precisely the opposite: Seljuq rule was regarded as puissant, wealthy, and glorious, and these two reigns in particular, of Malikshāh and Sanjar, are viewed as high points, not only of Seljuq rule, but of the entire Persianate dynastic period. There is, in other words, not only a lack of historical evidence that supposed long-term climate cooling caused gradual economic decline and destitution, but an actual refutation of this thesis in the testimony and perceptions of both the people who lived through it and the historians who viewed it in perspective after it had ended.Footnote ³⁶

In part, this error rests on a mistaken timeline of decline, which first arose from the dearth of research on the twelfth century, and led to the hitherto prevailing scholarly assumption that the power of the Great Sultans never recovered from the civil war over the succession to Malikshāh that began in 1092 and lasted for over a decade, until 1105. In other words, as a result of the almost complete absence of research into the latter half of Seljuq rule, earlier scholars simply assumed that the political weakness of the great sultanate that marked the succession crisis must also have characterized the entire second half of the Saljūq polity.Footnote ³⁷

In short, the conventional wisdom for a long while simply assumed that the entire last century of Seljuq rule was in effect one long Gibbon-like decline, and that the sultans were impotent puppet-rulers – a view, as we shall see, not supported by the primary sources, and no longer the consensus view among scholars;Footnote ³⁸ rather, the power and authority of the Great Seljuq Sultans was restored after the civil wars ended in 1105, and particularly after Sanjar b. Malikshāh's assumption of the Supreme Sultanate in 1118, thereby reversing, in Greater Khurāsān at least (that is, the areas under Sanjar's direct control, rather than that of suffragan sultans), the centrifugal tendencies which had taken root during the succession dispute. Part of what has confused the issue was the fact that the centre of the Great Sultanate after 1118 shifted very far to the east, so that the Baghdadī chroniclers upon whom most modern historians relied were now recording events only under weaker subordinate sultans on the periphery of the empire, rather than at its centre (a point that several scholars did not take into account) – that, however, is outside the scope of this study.Footnote ³⁹

Returning to the “Big Chill” theory, however, let us examine what the literary sources reveal about the Seljuq period generally, which according to the “Big Chill” should have been markedly poorer than those that preceded it, and growing ever more so. As we noted above, the sourceş do not bear this out – nor does the historical record: the fact remains that the Seljuq empire was by far the largest territorially to arise in the Islamic heartlands since the ʿAbbāsid collapse, and it underwent even further expansion (in both India and Central Asia) during Sanjar's time in the early twelfth century. This sheer size and scale of rule over the course of a century could not have been possible had the resources of Seljuq government been fewer than those at the disposal of the Sāmānid and Ghaznavid administrations.

In fact, the literary record – whether written towards the end of the Seljuq period, within a century after the end of their rule, or more than a century later – clearly and unanimously conveys the fact that the Seljuq period overall was one of material prosperity rather than growing indigence.Footnote ⁴⁰ Indeed, several historians not only rate the Seljuqs highly, but do so in relation to past Muslim rulers; for instance, the earliest historian of the period whose work survives, Nīshāpūrī, who wrote during the reign of the last Seljuq sultan, Toghril b. Arslān (d. 1194), avers as a universally known fact, writes that, since the ʿAbbāsid heyday:

Even allowing for panegyrical excess (after all, the author did not have to be enthusiastic to such a degree), this is not the eulogy of a dynasty that was considered to be poorer or less grand than its predecessors – and this is confirmed by an examination of the sources composed after the Seljuq period had ended, and there would have been no question of sycophantic flattery.Footnote ⁴² Thus, for instance, returning to al-Ḥusaynī’s statement,Footnote ⁴³ written within a few decades of the end of Seljuq rule, which was cited earlier:

Rāvandī, also writing around 1204 (Meisami Reference Meisami1999: 239), similarly extols Seljuq rule, echoing Nīshāpūrī: “In the nation of Islam, after the Rāshidūn caliphs, throughout the reign of the Banī ʿAbbās, there were no kings more pious or greater than the Seljuqs” (Rāvandī Reference Rāvandī and Iqbāl1945: 65). And, giving some historical perspective from the first part of the fourteenth century, Mustawfī Qazvīnī states:

Obviously, the Seljuq sultans could not have been liberal had they not also been wealthy; and the sources make quite clear that they were indeed wealthy rulers – including, most importantly, and to the detriment of both “The Big Chill” theory and Gibbon's Law, Sanjar.

For the literary sources for the Seljuq period all maintain in no uncertain terms that Sanjar's reign was a high point in Islamic history, in terms of both prosperity and power; the only other Seljuq ruler whose reign they write about with anything close to the enthusiasm and encomia they accord to Sanjar's is Malikshāh. Thus, Nīshāpūrī states that:

Nīshāpūrī also points out, most importantly, that “for nigh unto forty years he made nineteen conquests, during which at no time did he display weakness, nor was he defeated”, concluding that “He was a blessed king; the Shadow of God, strong, of auspicious aspect. The boundary of Khurāsān in his time became the desire of mortals and the native soil of the sciences, the fountain of literary attainments and the mine of knowledge” (Nīshāpūrī Reference Nīshāpūrī and Morton2004: 56). Indeed, another source informs us: “all the rulers of the world submitted to him and became obedient to his command”.Footnote ⁴⁸

The other major sources corroborate this assessment. Al-Ḥusaynī, for instance, states: “Sultan Sanjar was successful in all his actions; victorious in his military campaigns [ghazawātihi] … and he ruled great kingdoms, the likes of which no one before him and no one after him ever ruled, except his father the Sultan Malikshāh” (Al-Ḥusaynī Reference Al-Ḥusaynī and Iqbāl1984: 90). He further describes Sanjar as “The mightiest monarch whom God had [ever] made king” (Al-Ḥusaynī Reference Al-Ḥusaynī and Iqbāl1984: 92) and describes his great wealth and munificence – on one occasion, he is said to have distributed 700,000 dīnārs over the course of five days (Al-Ḥusaynī Reference Al-Ḥusaynī and Iqbāl1984: 125). This picture of magnificence, wealth, and power rather than decline, is further confirmed by Rāvandī, who adds:

Another late-twelfth-century author, the Arabic chronicler Ibn al-Jawzī, likewise states that after Sanjar became Supreme Sultan in 1118, “… his power was exalted and [he] was awe-inspiring … so the country was peaceful in his time” (Ibn al-Jawzī Reference Ibn al-Jawzī and ʿAṭā1412/1992, 18: 121). Ibn Khallikān, too, writing in the thirteenth century, reports Sanjar's treasurer as stating:

And, finally, the important fourteenth-century sources confirm that this reputation held true even with the perspective of distance: Mustawfī Qazvīnī writes in similar mode that Sanjar was “The Sultan of the Sultans of the world … Amongst the sultans of Islam he was as Parvīz among the Khusroes, through many conquests, high dignity and pursuit of success”;Footnote ⁵⁰ and al-Yazdī, after heralding him as “Lord, ṣubduer of enemies, King of kings, benevolent; strengthener of religion, defender of truth, aid of [all] creatures” (Yazdī Reference Yazdī and Shamsī1388/1968: 83) effuses:

Lest one think that all of this is uncritical fawning, or its echo down the centuries, it should be noted, first, that the chroniclers could not have invented wholesale Sanjar's string of military victories – and that includes not only his decades of success prior to 1141, but also, crucially, his victories subsequent to the debacle of Qaṭwān. Second, even supposing that all the authors had, mysteriously, been in universal collusion across the centuries tendentiously to inflate Sanjar's success for their own didactic purposes, they need not have done it to such a degree; merely placing him on a par with, say, Toghril Beg or Maḥmūd of Ghazna would have done the trick.

Finally, it is significant that the appraisals of Sanjar's puissance are universal across the centuries, and also include the commentary of authors who were demonstrably uninhibited in their frank evaluation of his power, which makes them particularly valuable in this context. Thus, for instance, Bundārī, writing of Sanjar's own first years as ostensible governor of Khurāsān, when Sanjar was a mere boy of about ten,Footnote ⁵² has no hesitation in describing Sanjar as having been a cypher at that time, “not assuming responsibility for governing … but possessing only the name of government … ” (Bundārī Reference Al-Bundārī and Houtsma1889: 259). Therefore when this same author writes to the contrary regarding Sanjar's reign in the early years of the twelfth century, showing that as Sanjar matured he became a vigorous and effective monarch, the statement carries weight: “[Sanjar's] rule continued in Khurāsān, and his sultanate became ever stronger and his power held sway…” (Bundārī Reference Al-Bundārī and Houtsma1889: 262). After 1105, so Bundārī enthuses, “[Sanjar's] success continued and his affairs were fortunate, his dawns illuminated, and his light broke forth” (Bundārī Reference Al-Bundārī and Houtsma1889: 262).

Indeed, the proof of Sanjar's real strength is found in the fact that, even after his sole pre-Oghuz military defeat, which took place in 1141 in battle with the infidel Qara Khitai at Qaṭwān, he still remained the supreme Muslim power of the age, and was able to put down the ensuing rebellions by his most powerful liegemen. Yet, the fact that the defeat at Qaṭwān was a major blow to the Saljūq state – militarily, financially, and prestige-wiseFootnote ⁵³ – has been erroneously interpreted by some modern historians, in the teeth of all evidence, to mean that Saljūq rule disintegrated after the battle of Qaṭwān. Once again, however, an examination of the course of events between 1141 and Sanjar's downfall in 1153 proves the opposite: Sanjar remained the most formidable ruler of his time, despite the blow the Saljūq state had received, right up until his abrupt and surprising downfall at the hands of the Oghuz. In fact, the sources state this explicitly. Such attestation to the re-establishment of Seljuq power after Qaṭwān includes the following several examples: “The power of the king's rule was established anew” (Nīshāpūrī Reference Nīshāpūrī and Morton2004: 59); “His status remained high and his success ever growing until he was taken prisoner by the Oghuz” (Al-Ḥusaynī Reference Al-Ḥusaynī and Iqbāl1984: 95); and “[Sanjar's] rule recovered its original splendour, and the prestige of the holy warrior [ghāzī] sultan was firm in the hearts of both the obedient and the rebellious, the nearest and the farthest” (Yazdī Reference Yazdī and Shamsī1388/1968: 88).

These appraisals are also validated by the empirical evidence of the actual course of events: in the twelve years after Sanjar's defeat at Qaṭwān, he crushed every rebellion and bid for power – and there were surprisingly few: that of the governor of Khwarazm for the second time in 1141, in the immediate aftermath of the defeat;Footnote ⁵⁴ and that of the vassal ruler of the province of Ghūr in 1152,Footnote ⁵⁵ in alliance with one of Sanjar's important magnates.Footnote ⁵⁶ In fact, the Ghūrid ruler and Sanjar's overly ambitious lieutenant seem to have committed the same error as modern historians: “They thought [Sanjar] would no longer be victorious in war after the battle with the infidel Qara-Khitai, and that conquest would no longer come to him as usual” (Yazdī Reference Yazdī and Shamsī1388/1968: 88). They were proved wrong – in the case of the disloyal magnate, fatally so; he was beaten in two under a banner or milestone.

The rehabilitation of Sanjar's might was confirmed by this last victory, which “restored the majesty and grandeur, which had suffered a setback after the unsuccessful war with the [Qara-]Khitai, and [Sanjar's] power was restored afresh” (Nīshāpūrī Reference Nīshāpūrī and Morton2004: 60). Clearly, bitter as the defeat at Qaṭwān was, it did not break the power of the Saljūq state. On the contrary, Sanjar was still demonstrably the strongest ruler in the Islamic world: he defeated quite handily in battle the only other two contenders for that title, the rulers of the Seljuq provinces of Khwarazm and Ghūr, whose polities were subsequently to assume supremacy in the eastern Muslim world after Sanjar's death.

In sum, the description of the Seljuq period during the first half of the twelfth century, and especially throughout Sanjar's long reign in Khurāsān, that is presented in the historical record is in no way compatible with the “Big Chill” theory of long and gradual decline. Sanjar's period is, on the contrary, explicitly depicted as one of outstanding and continuing prosperity, not growing impoverishment, and a high point not only of Seljuq times, but of Islamic history generally. Furthermore, even after the defeat of Qaṭwān in 1141, Sanjar demonstrated on the battlefield that he was still the supreme power of the central and eastern Islamic world. It should be noted that Cotton, Climate, and Camels does not cite any of these literary sources.

The Eclipse of Khurāsān in the historical record

We must look elsewhere, then, for an explanation of the eclipse of Khurāsān in the mid-twelfth century, which was, according to the primary sources, sudden and abrupt. Happily, the literary sources for the twelfth century have preserved the historical memory of the destruction of Khurāsān at this time – indeed, they supply an explicit explanation for the precipitous end of Khurāsān's greatness, and especially for the sharp drop in ʿulamā’ and intellectuals coming from that province after the 1150s. That explanation lies in the unexpected and disastrous captivity of Sanjar in the hands of a group of Oghuz Turkmen from 1153 to 1156, who then proceeded systematically to loot and ravage the province. While a lengthy analysis of this disaster and its causes lies outside the scope of this article, it suffices to say that Bundārī supplies the background explanation for Sanjar's abrupt and unexpected downfall, and the attendant evil consequences this had for Khurāsān: enfeeblement due to old age and a concomitant weakening of Sanjar's control over his own magnates.

At a time when power depended on the force of personality of an individual ruler, rather than on the strength of public institutions, any relaxing of the ruler's personal grip over his jockeying magnates could have catastrophic consequences. This is quite different from an extended imperial decline; rather, it is the personal failure of the figure at the centre of the political order, who played a role similar to that of a hub in a wheel; the weakness of such a leader, under the right circumstances, would not require very many years to bring about the unravelling of the polity. Thus, according to Bundārī, as Sanjar became senescent:

This was in fact what is described as having been the deciding factor during Sanjar's battle with the Turkmen Oghuz rabble; the magnates were too busy with their own quarrels to fight the Oghuz: “When [Sanjar's] battle with the Oghuz occurred, the Oghuz had not the ability to fight even one of his amirs; but envy of the Amīr … Yaranqush brought [the other amīrs] to forsake [Sanjar] while he was in battle.”Footnote ⁵⁸

What then ensued was frightful; during the years of Sanjar's captivity, the nomads went on a three-year rampage of pillage and rapine, in the course of which they not only destroyed the physical infrastructure of the province, but are also said specifically to have targetted the ʿulamā’ for wholesale extermination, since they were seen as being hand-in-glove with the Seljuqs (see Tor Reference Tor2016: 398–401); and, finally, the province's subsequent political dismemberment, which began during the period of Sanjar's imprisonment and continued after his death in 1157, six months after his escape from captivity.Footnote ⁵⁹

The primary sources are rife with explicit statements to this effect. Some of these include the Tārīkh-i guzīdah, which laments:

Nīshāpūrī similarly writes, regarding the targetting of the ʿulamā’, that “the [Oghuz] killed them with torture; palates and mouths which had for so many years been the revealers of the Sharʿī sciences and the founts of religious ordinances, they stuffed with earth [until they died]” (Nīshāpūrī Reference Nīshāpūrī and Morton2004: 65). According to Sibṭ Ibn al-Jawzī and Yazdī, in Nīshāpūr alone, the Turkmens are said to have killed 30,000 people, and a river of blood flowed in the streets (Sibṭ Ibn al-Jawzī Reference Sibṭ Ibn al-Jawzī and al-Jubūrī2013, 13: 29; Yazdī Reference Yazdī and Shamsī1968: 91–2). Indeed, according to Nīshāpūrī, so severe was the destruction there that “no one recognized his own quarter and house, and those places where the familiar mosques and madrasas of religious knowledge and assemblies [had stood] … were become pastures for sheep and hiding places for wild beasts and serpents” (Nīshāpūrī Reference Nīshāpūrī and Morton2004: 66).

The capital, Marv, was looted for three solid days, after which “in all the city nothing remained except the stuffing of cushions and mattresses … and that also they [then] took. Most of the people of the city they took captive.” (Nīshāpūrī 2003: 63.) Al-Yazdī reports in a similar lamentation:

These descriptions of the utter ruination of the province by Oghuz plundering, as well as the targetted killing of the class of religious clerics, are echoed and memorialized in one of the most famous of all Persian poems, Anvarī’s so-called “Tears of Khurāsān” (Anvarī Reference Anvarī and Riḍavī1959: 201–5), which was written contemporaneously with the destruction of Khurāsān, as a desperate appeal to the ruler of Samarqand to intervene and save the province, noting in particular the targetted killing of the ʿulamā’: “The Friday mosque in every town is become a stable for animals, roofless … . They do not make the khuṭba throughout Khurāsān in the name of the Oghuz because there is now no preacher, no minbar.” (Anvarī Reference Anvarī and Riḍavī1959: 202.)

Khurāsān was so physically destroyed and depopulated (including of its religio-intellectual elite) in fact, that Sanjar was unable to reconstitute his realm when he finally escaped to freedom – and the sources inform us of that as well, stating explicitly that “his realm had been destroyed”.Footnote ⁶⁰ Sanjar soon realized that the task of rebuilding Seljuq rule from the shattered ruins of his domains was impossible, and is said simply to have given up and died in despair:

The result was that by the end of the 1150s, the brilliance of Khurāsān had been extinguished and its leading role annihilated: From having been the centre of kingdoms and the seat of culture of the entire mashriq for over three centuries, it became instead a politically unsettled subordinate province, control over which was fought over by forces from the periphery, the Khwārazmshāhs and the Ghūrids – one of which, the Khwārazmshāhs, eventually triumphed, even though, if Bulliet's thesis had been correct, their realm should have been even more impoverished and depopulated.

In sum, there is no good evidence that there was a “Big Chill”, let alone that it caused a steady economic decline that led over the course of a century and a half to the political and cultural eclipse of Khurāsān – although Bulliet deserves all credit for drawing attention to the fact that the eclipse of Khurāsān actually occurred in the mid-twelfth century, decades before the Mongol catastrophe. There is, on the other hand, a great deal of explicit evidence in the written primary sources testifying that it was human agency which led to the chain of events that resulted in Khurāsān's blighting. Namely, Sanjar's increasing senescence and infirmity gave scope to the inherent factionalism that afflicted all medieval magnates, in both the Islamic and Christian worlds; and their resultant infighting allowed what was by all accounts a militarily unprepossessing group of nomadic Turkmens to defeat the Seljuq army and ravage unchecked for years, with spectacularly ruinous consequences, not just for the Great Seljuq Sultanate, but for the land of Khurāsān itself.

Let us give the last word to the great historian Ibn Khaldūn, who first noted this eclipse and its cause, the triumph of nomadic over sedentary culture: