The Development of Persian Culture under the Early Ghaznavids

Author: C. E. Bosworth

Source: Iran, Vol. 9 (1968), pp. 33-44

Published by: British Institute of Persian Studies

I

The Ghaznavids were a dynasty of Turkish slave origin who, in the last quarter of the tenth century, established themselves in what is now eastern and southern Afghanistan, at first as local governors on behalf of the Samanid Amirs of Transoxania and Khurasan, and then as independent sovereigns. In the course of a thirty-two years' reign, from 998 to 1030 AD, the greatest ruler of the line, the dynamic Sultan Yamin ad-Daula Mahmud b. Sebuktigin, extended his empire by force of arms until it stretched from western Persia to the Ganges valley of India, thereby earning an immense contemporary renown as the champion of Sunni orthodoxy and the hammer of the pagan Hindus. This vast empire was entirely a personal creation, and it endured for only a decade after his death; in 1040 AD his western conquests fell into the hands of a wave of Turkmen nomads from the steppes, the Seljuqs and their fellow-tribesmen of the Oghuz. However, Mahmud's descendants kept possession of eastern Afghanistan, Baluchistan and north-western India for a further century and a quarter, although this truncated empire became necessarily oriented more towards the Indian than to the Persian world.

This present paper deals not with the early Ghaznavid Sultans themselves or their policies, but with one aspect of their age, that is, the stimulus which the constitution of the Ghaznavid empire on the eastern fringes of the Iranian world gave to Perso-Islamic culture in that region. But before embarking on a discussion of this topic, it will be useful to bear in mind certain other facts. A consideration of almost any aspect of inner Asia, including this region where the Iranian and Indian worlds meet, requires a close reference to geography and topography. It may seem surprising that any degree of political or cultural unity is possible at all in Afghanistan, the Ghaznavid heartland, for the northern part of the country, in particular, is a meeting place of great mountain massifs and bare upland plateaux. Yet the measures of unity achieved by the great empires which have straddled this region, and the crossings of it by groups so widely separated in time as the first Indo-European invaders of India and the armies of Turkish conquerors like Timur and Babur, show that the central ganglion of the Hindu Kush and the Pamirs have never been a serious barrier to the passage of armies or peoples.

To take the Pamirs as a specific example: although these mountains rise to over 25,000ft., there has been in historical times a continuous local movement of peoples, with attendant cultural influences, from the valleys on one side to those of the other, and the results can be seen today in the linguistic geography of the region. Closely related Iranian languages appear on both sides: Wakhi, Munji and Shughni on the Oxus headwaters, Yidgha in Chitral. One finds lexical correspondences between the totally unrelated Wakhi on the upper Oxus and Burushaski in Hunza and Nagir at the extreme northern tip of West Pakistan, and between the only distantly related Indo-Iranian languages of Iranian Wakhi and Dardic Khowar in Chitral.

Modern Persian has filtered across the Pamirs and has become the language of literacy and official usage in parts of Hunza, including amongst the Burusho, whose own language is ofcourse a non literary one.

In pre-Islamic times, semi-nomadic confederations like the Kushans and Hephthalites or White Huns straddled the plateaux and fertile valleys of eastern Afghanistan, linking this region with their Central Asian homelands.

On the religious plane, there were strong Buddhist elements in northern and eastern Afghanistan, and these regions were thus linked with Tibet and China on the one hand and with Buddhist northern India on the other. All over the western parts of Central Asia, Buddhism decayed during the course of the seventh and eighth centuries, a process accelerated by the appearance of the Arabs and their new, assertive faith of Islam, although the great Buddhist centre of Bamiyan does not seem to have become definitely Muslim till the early Saffarid period, i.e. the later ninth century, or even later. Links with the Indian world nevertheless continued after the decay of Buddhism. Eastern Afghanistan was not properly secured for Islam till the end of the ninth or more probably the tenth centuries, and for the preceding two or three hundred years it was ruled by local princes, epigoni of the Hepthalite rulers of Zabul.

These were most likely themselves ethnically Iranian, but had close connections with India. This is clearly demonstrable in regard to the Hindushahi dynasty of Kabul and the upper Indus valley, whose power was uprooted by Sebuktigin and his son Mahmud. The local rulers of Ghazna during the first half of the tenth century, the shadowy Lawiks (displaced by the incoming Turkish slave commander Alptigin in 962 AD) were related to the Hindushahis. Nor is it unduly hazardous to link these Lawiks with the Zunbils who ruled in Zabulistan and Zamindawar, the region stretching between Ghazna and Bust, in the pre-Saffarid period, and who for long formed a powerful barrier against Muslim expansion there. Zamindawar had in the sixth and seventh centuries, if not indeed down to later times than this, a great shrine and pilgrimage centre devoted to the god Zun or Zhun, the Su-na of Chinese Buddhist travellers in these parts.

The origins and nature of this cult are highly obscure, except that it was clearly not Buddhist or Zoroastrian; Marquart plausibly argued that the cult of Zun might be connected with the shrine of the Hindu Sun-God Aditya at Multan, and recently, Bussagli has suggested links with the pre-Buddhist religious and kingship practices of Tibet. Thus it was really only in the tenth century that eastern and southern Afghanistan, the area between Zamindawar and Kabul, was fully islamized and integrated with the Muslim lands farther west. It is true that Sistan, in south-western Afghanistan, was invaded by Arab armies as early as the Caliph 'Uthman's reign and speedily Islamized, as the early appearance of scholars and traditionists with the nisba of " as-Sijistani ", and the transfer thither of the Arab politico-religious disputes between Khariji sectaries and orthodox Sunnis, show. Soon after the first appearance of the Arabs in Sistin, their forces passed through Bust into Zamindawar and clashed with the local ruler there, the Zunbil; but these were exploratory and plunder raids only.

Two centuries later, the Saffarids Yaqub and Amr b. Laith penetrated as far as Kabul, but these raids too were primarily for plunder and slaves. It is far from certain that the Lawiks of Ghazna were Muslims, despite the Islamic names given to them in the sources. As for the region of Ghur in central Afghanistan, this was a pagan terra incognita until the Ghaznavid expeditions there in the early eleventh century.

When in 900 AD Amr b. Laith was defeated in Khurasan by the Samanids, the succeeding Saffarids were speedily reduced to the status of a minor, tributary dynasty in Sistan alone, and eastern Afghanistan fell under Samanid suzerainty. The process, begun briefly by the first Saffarids, of drawing this region into the Perso-Islamic world, was continued under the political and military leadership of Turkish slave commanders of the Samanids, who assumed defacto power on the peripheries of the Samanid empire.

Qaratigin Isfijabi and other Turks established themselves in the south at Bust, and Alptigin and eventually Sebuktigin at Ghazna and Kabul, pushing back Indian influence down the Kabul river valley and then launching raids on the plains of northern India. Sebuktigin's governorship merged into the beginnings of the Ghaznavid empire, which speedily united Afghanistan with Khurasan and central Persia as far west as Jibal and the borders of Dailam.

Ethnically, Afghanistan had long been predominantly Iranian. There were some elements on the plateaux of southern and eastern Afghanistan, named in such sources as Istakhri, Ibn Hauqal, the Hudud al-alam and Utbi, as being Turkish tribesmen of the Khalaj and Oghuz groups; these may well have been human debris left behind by nomadic confederations who had held the region. These Turkish groups survived intact into the Ghaznavid period and were enrolled into the Sultans' armies; the Khalji Sultans who ruled in Delhi at the opening of the fourteenth century derived from them, and the name " Khalaj " may further survive in that of the modern Ghilzai Afghans.

Linguistically, Afghanistan was likewise predominantly Iranian, the main exceptions being the Dardic and Kafiri groups of languages and peoples in the valleys and mountains to the north of the Kabul river, the modern Nuristan (Kafiristan) and Chitral. These two groups were probably more territorially extensive in earlier times than they are today, for the trend in Afghanistan during the tenth and eleventh centuries was towards linguistic uniformity, with the New Persian or Farsi branch of north-eastern Iranian emerging as dominant over the other Iranian languages. The age of the Samanids saw the forging in Khurasan and Transoxania of New Persian as a fine instrument for literary expression, early seen in the verse of Rudaki, Daqiqi, Firdusi and others, and in the court literature of the early Ghaznavid poets.

So far as our knowledge of Pashto, the other great representative in Afghanistan today of the eastern Iranian linguistic family, is concerned, everything before the sixteenth century remains undocumented prehistory, despite recent pronouncements from Afghanistan which would push back our knowledge of written Pashto several centuries to the Ghurid period; but the hypothesis may tentatively be put forward that the beginnings of Pashto should be traced to the language of such early Iranian invaders of Afghanistan as the Sakas.

The spread of Farsi in Afghanistan is a little-known process, yet there are indications that the Ghaznavid period was an important one here. Admittedly, there are only one or two pointers. One concerns the remote and backward region of Ghur.

When in 1020 AD Mahmud's troops invaded Ghur, the leader of the expedition, Prince Masud, had to take with him interpreters; clearly, standard Farsi had not yet penetrated to this province. The next century saw the remarkable rise to power in eastern Iran and Afghanistan of the Ghuri chieftains of the Shansabani family, but there is no mention in any literary or historical source of any linguistic aberrancy in their native territory. Like the Ghaznavids whom they supplanted, the Ghurids had their court poets, and these wrote in Persian. What, then, could the old language of Ghur have been ? The late Prof. V. Minorsky once mentioned to the present writer that one fantastic theory had linked it with Burushaski. More feasible is the suggestion of Prof. Georg Morgenstierne, that it was one of the south-eastern Iranian group of dialects which have gradually been eliminated from Afghanistan by the spread of Farsi and Pashto. Parachi (in a few villages to the north of Kabul) and Ormuri (in the Logar valley and at Kaniguram in Waziristan) may represent the last remnants of some of these dialects.

A second pointer, admittedly atenuous one, may lie in a study of present day isoglosses in Afghanistan. Ghazna and Kabul are still salients of Persian speech in eastern Afghanistan, separated by a wedge of Pashto in the Logar and Wardak valleys. This may be due to the influence in medieval times of Kabul and Ghazna as centres of Persian culture and learning, their effect still significant over the centuries, whereas farther south, so far as we can see, Pashto has extended westwards through Qandahar (a place of comparatively recent importance and of lesser cultural significance) towards Farah and Isfizar or Sabzawar. This, however, is speculative; we may have a firmer base for speculation when data has been assembled by the Linguistic Survey of Afghanistan at present being undertaken by Kabul University under the general guidance of Prof. Morgenstierne.

II

It has just been suggested in general terms that the establishment of the Ghaznavid empire brought the central and eastern parts of the Iranian world under a single political authority and at the same time gave a fillip to the supremacy of Perso-Islamic culture on the far eastern fringes of that world. This thesis now requires elaboration.

In many ways, the empire of Mahmud was a successor-state to that of the Samanids in Transoxania and Khurasan. Mahmud and his father Sebuktigin both began their careers as military commanders of the Samanids, only abandoning their masters when it was evident that the Samanid empire was disintegrating under the internal stresses set up by rebellious generals and by external pressure from beyond the Syr Darya by the Turkish Qarakhanids. To the very end, Sebuktigin clung to his legal and official status as a slave provincial governor on behalf of the Samanid Amirs, and he never formally claimed the independence which he in reality enjoyed; the inscription on his tomb at Ghazna names him as al-Hajib al-Ajall " Most exalted commander ", and not as Amir. Down to 999 AD, Sebuktigin and his two sons Ismail and Mahmud all acknowledged the Samanids on their coins.

From the point of view of administrative structure and techniques, the continuity of the Ghaznavid empire with the Samanid one can clearly be demonstrated. Ruling from Bukhara, the Samanids evolved a highly-developed system for administering their territories, drawing taxation from the agriculture and industry of the oases of Khurasan and Soghdia, and controlling the long-distance caravan trade between the Islamic lands and the Far East; above all, the Amirs grew rich on the traffic in slaves from the Turkish steppes.

According to the historian of Bukhara,Narshakhi, the Samanid administration comprised nine separate diwans or government departments, including not only the basic financial, secretarial and military ones, but also departments responsible for espionage and police services and for the postal and intelligence network. Administrative techniques were advanced, and Abu Abdullah Muhammad al-Khawrazmi in his dictionary of technical terms, the Mafatih al-ulum " Keys of the sciences ", mentions that the Samanids had twenty-six different kinds of official registers for recording financial and other business. The ultimate model for all this administrative structurewas, of course, the Caliphal administration in Baghdad.

The first Ghaznavids developed an administrative system in Ghazna which was based on that of the Samanids, firstly because they knew of no other model to take, and secondly because there were no strong Islamic administrative traditions in the Ghazna region for them to build on otherwise. There must have been some coming and going of Samanid officials from Bukhara or Nishapur during the years when Turkish slave governors were ruling in Ghazna on behalf of the Amirs.

During Sebuktigin's tenure of power (977-997 AD), we can see the emergence of three basic government departments, those of finance, correspondence and military affairs, under the Vizier, Chief Secretary and Arid respectively. The organization of the Diwan-i Risalat or Correspondence Department was the work of one of the most noted literary stylists of the age, Abul-Fath Ali al-Busti, known for his scintillating use of paronomasia and other rhetorical devices as the Sahib at-Tajnis, and the friend of Badi az-Zamin al-Hamadhani, Abu Mansur ath-Thaalibi, Abu Bakr al-Khawrazmi and the Sahib Ismail b. al-Abbad.

Under Mahmud, the Ghaznavid empire expanded enormously, and the machinery of government inevitably grew more complex. Two new diwans appear, one a subdivision of theVizier's department, an accounting office under the Mustaufi, and the other that of the Ishraf, responsible for communications, postal services and espionage--a very necessary department for such a far-flung empire, whose Sultan was a despot ruling by fear rather than by consent.

Continuity with Samanid administrative practices was ensured by a continuity in many cases of actual personnel, for the Ghaznavids inherited many members of the old Samanid bureaucracy. Some of these merely remained at their posts in the Diwan of Khurasan at Nishapur when Mahmud took over there. Other Persian officials migrated from Transoxania when the Qarakhanids moved in and allowed much of the old Samanid administration to run down and fall into disuse; a continuator of Narshakhi says that taxation and expenditure every where decreased when the Qarakhanids came.

The power of the new and dynamic Ghaznavid empire to attract able men extended all over the Iranian world. An anecdote in chapter XLI of the Siyasat-nama of Nizam al-Mulk describes how a group of unemployed Buyid officials and secretaries in Ray contemplate emigrating to Khurasan, where they believe that Mahmud's well-known appreciation of scholars will bring them recognition and employment. Although this particular story is of dubious authenticity, we do have definite instances of men passing from Buyid into Ghaznavid service. The Qadi Abul-Hasan Ali Shirazi, head of the civil administration in northern India during the reign of Masud b. Mahmud, came originally from Buyid circles. However, it was the Samanid bureaucracy which provided by far the greatest number of Ghaznavid officials.

Mahmud's Vizier Abul-Abbas al-Fadl Isfaraini had been a secretary in Samanid Khurasan, as had been the father of the Vizier of Masud and his son Maudud, Ahmad b. Abd as-Samad; and the list could be prolonged. Since the Ghaznavid empire depended so much on the Samanid inheritance for its political structure, it is not surprising that literary, cultural and artistic trends under Mahmud also followed the patterns established in the eastern Iranian world by the Samanids.

It was the court of Bukhara which gave material backing for the literary florescence of New Persian, whilst at the same time remaining a great centre for the traditional Arabic theological, legal and philological sciences. Tha'alibi's praise of Bukhara at this time. prefixed to the fourth section of his literary anthology the Yatimat ad-dahr fi mahasin ahl al-asr " The unique pearl concerning the elegancies of contemporary people ", is well known: " In the time of the Samanids, Bukhara was the meeting-place of all nobility, the centre of all authority, the place where the outstanding people of the age congregated, the rising-place of the stars of the learned scholars of all the earth and the place of pilgrimage for all the brilliant men of the time."

But lesser courts of eastern Iran also contributed to this cultural revival by providing patronage and

shelter for scholars. Amongst the Ziyarids in Gurgan and Tabaristan, the Amir Qabus b.Vushmagir himself achieved repute as a poet and prose stylist, as the subsection on him of the Yatimat ad-dahr shows.

The poet Farrukhi Sistani, eventually one of the adornments of Mahmud of Ghazna's court, won his

initial fame at the petty court of the Muhtaji Amir of the Chaghaniyan on the upper Oxus.

The unknown author of the geographical treatsie, the Hudud al-alam " Limits of the world seems to have worked in Guzgan, the small principality of the Farighunid family in north-western Afghanistan. It was therefore unlikely that Mahmud would allow his own court to fall short of the standards set by the Samanids and by these minor Iranian princes. New buildings in Ghazna were financed out of and adorned by the spoils of India, such as precious metals, and the finds of the Italian Archaeological Mission in Afghanistan at Ghazna show that Hindu statues and other objects were incorporated into the fabric of palaces as trophies of war. On the intellectual level, libraries and mosques in Ghazna were enriched by the spoils of Khurasanian and other Persian collections.

The library of the madrasa attached to the splendid mosque built by Mahmud, called the Arus al-Falak or " Bride of Heaven ", was fitted out in this way. When Mahmud captured Ray from the Buyid Majd ad-Daula in 1029 AD, fifty loads of heretical books of the Batiniyya and Mutazila were burnt, but the more harmless ones were spared and carried off to Ghazna. The rich collections of Ghazna doubtless perished in the sackings and burnings of the Ghurids and Mongols a century or two later.

Yet it was not enough to found libraries; great luminaries of the literary and scientific worlds had to be brought to Ghazna if it was to become the Baghdad of the east. The polymath Abu Raihan al-Biruni, after working in his native land of Khawrazm and then wandering to the Ziyarid court in the Caspian coastlands, returned to Khawrazm and the patronage of the Ma'munid Shahs of Gurganj. When Mahmud's troops marched into Khawrazm in 1017 AD, al-Biruni joined the exodus of scholars from there to Ghazna, apparently of his own free will. At Ghazna he finished his days as a kind of scientific adviser cum court astrologer to Mahmud,his son Masud and then the latter's son Maudud.

It was through this position that al-Biruni was able to accompany Ghaznavid armies into India, delve deeply in Sanskrit and the Prakrits tongues of northern India and acquire a unique knowledge of Indian religious and social practices, all of which he put into his magnum opus on India, the Tahqiq ma li l-Hind. It is indeed fortunate for human knowledge that a man with such an omnivorous and all-enquiring mind should have found himself placed in Ghazna at a time when Muslim arms were making their first major break-through into the plains of northern India. For facilitating this alone, we should be grateful to the early Ghaznavids.

However, not all scholars were entranced by the prospect of life at Mahmud's court. As E. G. Browne truly noted, "Sultan Mahmud has often been described as a great patron of letters, but he was in fact rather a great kidnapper of literary men, whom ... he often treated in the end scurvily enough ".

A story in Nizami Arudi's Chahar maqala, which may be apocryphal but which nevertheless reflects Mahmud's strong-arm methods, describes how he sent an ultimatum to the Ma'munid Khawrazm-Shah Abul-Abbas Ma'mun, demanding that the Shah send to his court certain famous scholars. One of these, Ibn Sina or Avicenna, had been reared in an atmosphere sympathetic to Ismili Shi'ism, and he did not relish a summons to the conservative and orthodox Islamic milieu of the Ghaznavid court; he preferred to flee across Persia and end his days as Vizier to the Kakuyid ruler of Isfahan, Ala ad-Daula Muhammad. The cases of Ibn Sina and one or two others with heterodox sympathies were probably exceptional.

Mahmud cultivated the image of himself as a Maecenas, and scholars and literary men made their way to Ghazna from all over the eastern Islamic world. In particular, the Ghaznavid court acquired a group of fine Persian poets, presided over by Unsuri, who allegedly had the title of Amir ash-Shuara, Poet Laureate. The comparative cultural poverty of the region of Ghazna and Zabulistan is shown by the fact that no major figure of this group of court poets seems to have been a local man: Farrukhi came from Sistan, Unsuri from Balkh, Manuchihri from Damghan, Ghadairi from Ray, and so on.

The volume of verse turned out must have been very great, much of it being panegyric addressed to various members of the royal family, to high officials and to military commanders, and what we have left of it today is generally of a high literary standard. The style is comparatively simple, certainly when compared with the intricacies of Seljuq and later Ghaznavid verse, and the use of imagery and the depiction of nature are usually fresh and delicate. The literary forbears of this eulogistic and lyrical poetry were such samanid poets as Abul-Hasan Shahid of Balkh, Abu Bakr Khusrawi of Sarakhs, Abul-Hasan Kisa'i of Merv and, above all, Rudaki. Most of these poets were dhawu al-lisanain,equally proficient and productive in Arabic or Persian, and we can thus trace the stylistic origins of early Ghaznavid poetry through the Samanid poets to the Arabic qasida. Unfortunately for purposes of comparative and stylistic study, we have extant today only a small part of this early Ghaznavid poetry, the visible tenth, as it were, of a poetic iceberg.

The only diwans which we possess in anything like completeness are those of Unsuri, Farrukhi and Manuchihri; the other poets are known only through fragments or citations. This is in contrast to the later Ghaznavid period, from which we have several more or less complete diwans, such as those of Mas'ud-i Sa'd-i Salman, Abul-Faraj Runi, Sayyid Hasan and 'Uthman Mukhtari.

Because of these movements of men, ideas and literary concepts eastwards to Ghazna, the Ghaznavid empire gradually became integrated with Khurasan and the eastern Iranian world in general. Once this seed was implanted, Ghazna and eastern Afghanistan began to develop a Persian culture of their own, a culture which survived the Ghurid take-over of the Ghaznavid empire in the mid-twelfth century and endured right down to the Mongol invasion, when Ghazna's role as a historic centre ended for ever.

III

So far, the implanting of outside ideas within Afghanistan during Mahmud's reign has been discussed. Although it has been noted above that the region of Ghazna and Zabulistdn was not, at the opening of

the eleventh century, culturally so richly developed as say Transoxania or Khurasan, Afghanistan as a whole was not entirely a cultural desert. It was pointed out at the beginning of this paper that it lies at

the meeting-point of several civilizations, those of Iran, of India, of the Central Asian steppes and of China and Tibet. And since the Hindu Kush and the Pamirs have never inhibited the movement of people and ideas across the heartlands of Asia, and since some at least of these peoples and ideas have halted within Afghanistan, it would be surprising if the region had nothing whatever to contribute to the flowering of Perso-Islamic civilization during Mahmud's reign.

It is suggested that such a contribution may in fact be traced in the stimulus given to popular epic literature and romances during the early Ghaznavid period. Our surviving information relates wholly to such literature in Persian, but there are indications that a similar literature existed also in Turkish, probably in a less-polished literary form and circulating at an oral level. It is easy to forget that the Ghaznavid Sultans, though speedily Persianized in culture and outlook, were ethnically Turks, not far removed in time from the Central Asian steppes. Sebuktigin was born in paganism, apparently at Barskhan by the Ishiq Gol of the modern Kirghiz Republic. Persian and Arabic dominated the bureaucracy, the religious institution and the world of scholarship, but Turkish remained an everyday language for the Sultans and their intimates, certainly down to Masud's reign and probably longer.

The army was the great and enduring stronghold of Turkishness, being in large part composed of Turks

from the steppes, and there were always new arrivals from this quarter thus keeping up the barbarian

and unsophisticated element in the ruling institution. These Turks must have brought with them

their tribal legends, stories and poetry. The Persian poet Manuchihri mentions Turkish poetry in one

of his own verses; this would doubtless be popular poetry and not the product of cultured circles such as

Manuchihri and his colleagues formed. An interesting point here is that Manuchihri speaks of poetry in both " Turki " and " Ghuzzi ", a very early division of the Turkish dialects into an eastern, Qarluq and Uighur group, and a western, Oghuz and Qipchaq one, a division laid down, with supporting phonological evidence, in Mahmud Kashghari's Diwan lughat at-turk thirty or forty years later.

Mahmud of Ghazna's name is linked in popular imagination with that of Firdusi, author of the supreme version of the Persian epic, the Shah-nama or Book of Kings. A well-known version of the relations between the two men is that given by Nizami Arudi: that Firdusi spent twenty-five years on his masterpiece, that he brought it to the Ghaznavid court, but that Mahmud repulsed him and offered only a miserable present. The poet then wandered westwards to the court of the Bawandids in Tabaristan. In the end, runs the story, Mahmud recognized Firdusi's genius and wished to atone for his shabby behaviour by sending a magnificent gift of 60,000dinars worth of indigo; but when this arrived at Tus, the poet was already dead. The truth of the matter remains unclear.

Arberry thinks that Firdusi made an unfortunate choice in offering what he calls " his vast epic in praise of Zoroastrian Persia " to Mahmud, " the fanatical conformist".

Admittedly, the Shah-nama is not Islamic at all in its inspiration, but neither is its ethos specifically Zoroastrian, and the Ghaznavids were not averse to being connected with the glories of old Persia. When obliging genealogists concocted a pedigree for the dynasty, they were unable to get round the fact of Sebuktigin's pagan origin, but they did manage to connect him with the last Sasanid emperor, Yazdagird III, whose family had allegedly fled into the Central Asian steppes after the Arab invasions and had settled there, intermarrying with the local Turks. Rypka's reconstruction of the relations between Sultan and poet as being dependent on Ghaznavid political policy, firstly pro-Iranian against the Turkish Qarakhanids, and then pro-Arabic as links were strengthened with the Baghdad Caliphate, is attractive but hard to prove. He is right, however, in regarding the Shah-nama as the climax of the " feudal epic " which had arisen from the

Samanid milieu, and in regarding the epics of the succeeding Ghaznavid period as constituting a new genre, the " romantic epic ".

It is the role of Afghanistan in general and Mahmud's own court in particular as the nurturers of this " romantic epic " that requires further examination. The probable survival of a popular Turkish literature amongst the extensive groups of Turks in the Ghaznavid court and army has just been mentioned. Some of these Turks became Persianized, but most of them must have been eager for literature and poetry to suit their own tastes. Their attitudes and requirements may well have affected the nature of this epic and romantic literature in Persian. Within this literature, two streams of development may be discerned. One follows directly on from Firdusi, but with a romantic element injected into the epic, pushing the heroic element further and further into the background; the other stream is that of purely romantic and lyrical idylls in verse. But a stimulus to both was, it is suggested, a contribution of the early Ghaznavid period to Persian literature.

For the Shah-nama was in many ways a culmination. If literature is to be regarded as the expression of a society, then the Shah-nama must obviously be attached to the old Persian landowning classes of eastern Iran, the dihqans. This class maintained its power and influence in khurasan Transoxania and Khawrazm until the downfall of the Samanids.

The tenth century geographers expatiate on the heroic virtues still cultivated by the people of these regions--the munificence of their entertainment of strangers, their general liberality, and their military prowess. Thus Maqdisi describes the people of Khawrazm as " men who practice hospitality, with great appetites, courageous and impetuous in battle ". Pride in the Iranian past, its heroes and their achievements, lived amongst this class, for the spirit of this past long survived the strong social and cultural pressures of Islam; Zoroastrianism itself persisted in the more mountainous and inaccessible regions of Persia, and it was, after all, in the ninth century, two centuries after the coming of Islam, that much of the Middle Persian and Zoroastrian literature known to us today was copied. It was for an important Khurasanian dihqan, Abu Mansur b. Abd ar-Razzaq of Tus, that translations into New Persian were made in 951 AD from the Pahlavi texts of the national epic, and these texts were later utilized by Firdusi and perhaps by his predecessor Daqiqi.

But from 999 AD onwards, Khurdsdn was in the hands of the Turkish Mahmtid of Ghazna, and Transoxania in those of the Qarakhanids. Supreme political leadership in eastern Iran passed out of native Persian hands, and although the dihqin class was still for a time influential at the local level, the long-term trends of the new period were definitely unfavourable for it. Under the Ghaznavids, Khurasan was ruthlessly exploited by successive Viziers and governors, themselves hard-pressed by the Sultan their master, who was avid for money to finance his campaigns in India and elsewhere.

A disastrous famine, followed by plague and much rural depopulation, is recorded for 1011AD. Two or three decades later, the depredations of the Seljuqs caused a disruption of trade and agriculture; land values in such places as the Nishapur and Baihaq oases plummeted, with ruinous effect on the fortunes of the dihqan and small landowner classes. By the time of the Mongol invasion, the term dihqan had declined into the simple meaning of " peasant ".

The decreased taste for purely heroic epics of the Shah-nama type may therefore be a consequence of the gradual ruin of the Iranian landed classes, although the diversion of this taste into new literary patterns was probably also a result of an increased sophistication and refinement of manners in early Ghaznavid times, expressed in a preference for the romantic over the heroic.

In fact, the example of the Shah-nama released a spate of Persian epic poems which has continued almost down to the present day. The late Maryan Mole quoted the early twelfth century anonymous Persian history, the Mujmal at-tawarikh, in which the Shah-nama is called the tree and all the other poems its branches.

The Shah-nama and most of its successors are based on national traditions current in the eastern Iranian world, in particular, those centred on the regions of Sistan, Zamindawar and Zabulistan. These regions had long been famous for breeding hardy and pugnacious fighting men; Rustam is described as " the Zabuli hero ", gurd-i Zavuli, not only in the Shah-nama, but also in the Middle Persian historical epic of 400 years before, the Karnamak-i Artashir-i Papakan.

The Shah-nama centresof course round Rustam and his father Zal, but the slightly later Garshasp-nama, which will be considered presently, and a host of later poems, revolve round Garshasp, the Avestan Krsaspa, " He who has lean, swift horses ", and a forebear of Rustam. These poems in which Garshasp and his family, rather than Rustam, are featured, are specifically attached to Sistan and Zabulistan; and there was, according to a lost Kitab-i Garshasp by Abul Mu'ayyad Balkhi, cited in the Tarikh-i Sistan, a fire temple at Karkuya near the capital Zarang, which was dedicated to Garshasp and had a cult in his honour.

Under the early Ghaznavids, these regions of southern and eastern Afghanistan were for the first time in centuries the heart of a vast empire, and they supplied troops for armies which campaigned from the lower Oxus to the Ganges; contingents of Afghans (thus named in Utbi's at-Tarikh al-Tamini: al-Afghaniyya) are specifically mentioned, as are the archers and infantry men of Sistan.

One might have expected that the exploits of Mahmud and his troops would have stirred popular imagination to the extent that epics on his campaigns might have been composed and the whole topic grafted on to the existing Iranian epic tradition popular there. Yet we have no surviving contemporary epics woven round the figure of the warrior-Sultan. It is true that there is mention of a metrical composition, the Taj al-futah " Crown of victories ", dealing with Mahmud's exploits; since the poet Unsuri praises this work highly, Nazim thought that this was probably written by himself. Whoever wrote it will probably never be known for certain, since it is not extant. It was not until well after Mahmud's death that a considerable literature, half romantic epic, half hagiography, grew up round Mahmud and his son Masud as champions of Islam against the infidels in India. In this literature, historical fact diminishes as the miraculous religious element increases. This is seen in the legends accumulating round the exploits of the semi-legendary Ghaznavid soldier Sipahsalar Masud Ghazi, allegedly a nephew of Sultan Mahmud; these legends were brought together in the early seventeenth century by the Miras-i Masudi of Abd ar-Rahman Chishti.

The absence of any contemporary literature specifically crusading in tone and glorifying Mahmud as the spearhead of Islam in India seems to confirm the impression from the historical sources, that Mahmud was not a fanatic bent on imposing Islamic religion on the Hindus, but was activated more by an imperialist love of power and a lust for gold. What does occasionally seem to have happened is that actual incidents or names of people involved in the Sultan's campaigns were transplanted into the epics of a few decades later. Thus in the Faramurz-namas (at least two epics with this title are known), probably dating from the latter part of the eleventh century, one of Faramurz b. Rustam's enemies is called Jaipal, this being the name of the Hindushahi Raja of Waihind and the western Panjab, the great opponent of Sebuktigin and Mahmud. The Sultans who succeeded to Mahmud and Masud were also concerned with expansion into the plains of India, for a continuous stream of plunder was necessary for the economic well-being of the empire. It was the exploits of a later Ghaznavid, Masud III b. Ibrahim b. Masud I (1099-1115AD) which formed the occasion for the composition of a remarkable panegyric in Persian on the military achievements of the whole line of Sultans.

This was written in the mutaqarib metre (that employed in the Shah-nama) and inscribed on slabs forming part of a dado round the main courtyard of the palace at Ghazna completed for Masud III in 1112AD. Of it, Prof. Alessio Bombaci, director of the Italian Archaeological Mission's excavations at Ghazna, says, " The epic tone is inspired by two ideals clearly set forth in the very few verses that have survived: the Islamic ideal celebrated in the work of Mahmud and the ideal of legendary Persian tradition embodied in the obvious adherence to the pattern of the Firdusian epic. Hence, the characteristics of the champion of the faith are confused with those of the Iranian hero ". Here, it would seem, is the Iranian epic genre at last specifically adapted to the greater glory of the Ghaznavid dynasty, but a century had to elapse after the most spectacular conquests, those of Mahmud, before this poem was put together. The list of successors to and imitations of the Shah-nama in the eleventh and twelfth centuries is long.

It includes the Garshasp-nama, the interminable Barzu-nama, which runs to over 65,ooo couplets; the Shahriyar-nama; the Bahman-nama; the Fardmurz-namas; and the Kush-nama. To this period belongs also the Jahangir-nama,probably by Qasim Madih of Herat, of very strong Islamic religious inspiration; but the genre continues into Mongol and Timurid times and beyond. Most of these authors are either unknown or else are very shadowy figures. One of the manuscripts of the Shahriyar-nama attributes the poem to the Ghaznavid court poet Farrukhi; if this is correct, it must have been the first of the post-Firdusian epics to have been composed, clearly ante-dating the Garshasp-nama. Except for one or two, all these epics are localized in the Sistan-Zamindawar-Zabulistdn region, as being the home of Rustam and his line. Their aim is to supplement and fill out the work of Firdusi, who had by no means exhausted the whole contents of the national epic, but their emphasis tends to be much more on the miraculous and incredible aspects of the heroes' exploits, exploits which take place in remote lands like India, Malaysia, Byzantium or some vague " western lands ", where the atmosphere is thick with marvels and portents. Moreover, their ethos becomes definitely Islamic, whereas the earliest epics, the Shah-nama and also the Garshasp-nama have a monotheistic ethos which could be indifferently Islamic or Zoroastrian.

Our concern here is with the Ghaznavid stimulus to this epic literature, and of especial interest to us is the epic which comes forty years or so after the Shah-nama and is very definitely rooted in the region of Ghazna and Zabulistan, the Garshasp-nama. In this instance, the author is indisputably known. Ali b. Ahmad Asadi Tusi (1080AD) is celebrated as the author of the oldest known New Persian vocabulary, the Lughat-i Furs, and Henri Masse considers his literary talent as little inferior to that of Firdusi. Asadi Tusi wandered extensively from his Khurasanian home, and in 1064-66AD composed his epic at the court of the Amir of Nakhchevan in Arran. Its inspiration, nevertheless, is strongly Zabuli.

Garshasp is the ancestor of Rustam, and his family lives in Zabulistan; and Zahak, a tyrant reigning for a thousand years in the Shah-nama, appears in the Garshasp-nama in a favourable light, ruling as a legitimate king with the approval of the nobles of Ghazna. Only a few decades later than Asadi Tusi's time, we find the Shansabani rulers of Ghur tracing their ancestry back to Zahak, with the story that Zahak fled into the fastnesses of Ghur for refuge when Feridun overthrew his rule. The name Zahak, Arabized into ad-Dahhak, seems to have been popular in Zabulistan at this period; it was borne by the father of Gardizi, the early Ghaznavid historian. It seems that Asadi Tusi used local traditions and artificially attached them to the mainstream of the Iranian epic by conveniently marrying Jamshid to a daughter of the king of Zabulistan. The Garshasp-nama therefore reflects popular beliefs of the Ghazna area, and it does not seem unreasonable to view the elaboration of these beliefs into a form utilizable by the poet as an expression of local pride in the region's role as the centre of the Ghaznavid empire.

It was mentioned that there is a second stream of development within Persian literature from early Ghaznavid times onwards, that of the poem or tale treating of romantic love and suffused with a lyrical strain. The best-known example of this is the romance of Vis u Ramin by Fakhr ad-Din As'ad Gurgani, written in the hazaj metre around 1050AD for the Seljuq governor of Isfahan and based, according to Minorsky, on a much earlier Middle Persian version, used by Gurgani either in the Pahlavi or in an early Farsi translation.

However, primacy in time goes to others. The court poet 'Unsuri wrote a romance in the mutaqarib metre, Wamiq u 'Adhra', based on a much earlier Persian romance allegedly dedicated to the Sasanid Khusrau Anushirvan. This work of 'Unsuri is, unfortunately, only known through a few citations in works on rhetoric and style and in the later Ottoman Turkish versions of Lami'i and Jami'i. But there has recently come to light in an Istanbul manuscript the important Persian romance of Warqa u Gulshah by 'Ayyuqi, an obscure figure otherwise unknown except for two citations in Asadi Tusi's Lughat-i Furs. The late Ahmed Atesh showed that these citations from a mid-eleventh century author, the style of the romance, the use of the mutaqarib metre, not used for romantic subjects after the early Ghaznavid period, and the actual dedication of the work as a Mihrgan present to " Sultan Abul Qasim Mahmud ", make its attribution to Mahmud of Ghazna's court circle quite certain.

The theme is based on the unhappy love-life of the early Arabic poet 'Urwa b. Hizam al-'Udhri, which early gave rise to a cycle of popular romances in the Islamic world. The theme of the two lovers separated by death, but with Gulshah miraculously restored to life and reunion with his beloved by the Prophet Muhammad, passed via Spain into medieval French literature as the romance of Floire and Blanchefleur. 'Ayyuqi's version of the popular romance was probably the first verse one. Atesh further underlined the importance of Warqa u Gulshah for Turkish literature; the subject appears early in popular literature, and in the fourteenth century a version in classical Turkish was made by Yusuf-i Meddah.

Can we then see a strand of Turkish influence, stimulating the composition of this early Ghaznavid work? Such an influence seems not unlikely. The numerous Turkish courtiers, soldiers and domestic slaves in Mahmud's retinue at Ghazna have already been touched upon, and their presence implies a receptive audience for both epic and romantic literature of a type familiar to the Turks in their Central Asian homeland, the type of epic woven round tribal origins which crystallized later into the Oghuz- nama.

It is known that, well before Firdusi's time, the Iranian national epic made a great impression on the Turkish mind in Transoxania and elsewhere. It was because of this that the Turks identified their national hero Alp Er Tonga with the Afrasiyab of the Iranian epic, Kai Khusrau's enemy.

Consequently, the Qarakhanids styled themselves " the House of Afrasiyab", and a genealogy was ater-compiled for the Seljuqs tracing the family back to Afrasiyab. Conversely,the Turanians of the Iranian epic, by which name were originally meant the Indo-European nomadic peoples of the steppes like the Scyths and Massagetes, were identified with the contemporary Turks who had by then taken over the steppes. But the Turks also attached themselves to the Arabo-Islamic past by weaving Turkish tales round such events as the part of Abu Muslim in the 'Abbasid Revolution of 750AD. Abu Muslim became a figure venerated by the Turks of Central Asia, and accordingto Vambery, this reverence was still discernible amongst the Turkmens and Uzbeks during the last century. Eventually, Turkish tradition made the Khurasanian Abu Muslim into a Turk himself, attaching him to the Oghuz tribe.

Amongst several epic prose tales in Turkish attributed to one Abu Tahir-i Tusi is a historical romance on the Abu Muslim story.One manuscript says that the story-teller was blind, korgozi, but blind bards and story-tellers are common figures in all literatures and one should not perhaps take too much notice of this. One point that is always mentioned in the tales attributed to Abu Tahir-i Tusi is his connection with Mahmud of Ghazna's court, the Sultan being his patron and listener to his tales. The language of these tales, according to the manuscripts, was originally Persian, but Turkish versions must speedily have followed.

Characteristic of these tales is the miraculous and magical element, with much intervention by fairies and demons, all this appealing, it would seem, to the unsubtle Turkish mind. Although there are anachronistic features in the stories attributed to Abu Tahir-i Tusi which clearly belong to Seljuq and even later times, is the conclusion of our greatest authority on the Abu Muslim cycle, Mme. Irene Melikoff, that Aba Tahir-i Tusi was a historical figure, who flourished at Mahmud's court and whose renown lasted well into succeeding centuries.

IV

To sum up, there was a definite resurgence of Persian culture in the early Ghaznavid period, and there are a few small indications that the age was also not without significance for the development of a nascent Turkish literature and culture. This Persian culture, arising in the first place out of the Samanid and Khurasanian inheritance, began in Mahmud's reign to have a life of its own and to evolve along distinctive lines. Of significance in this process are firstly,the tastes of both the Sultan's Persian officials and advisers and his Turkish soldiers and courtiers, and secondly, the enduring local traditions of Zabulistan, the geographical heart of the Ghaznavid empire, now linked politically with the wider Persian world. The literature surviving today from the early Ghaznavid period is not extensive, compared with what we know was in fact written.

One thinks of the works of many poets, known only by the mention of their names in such works as the tadhkiras, of al-Biruni's lost history of his native Khawrazm, of the missing volumes of the bureaucrat Abul-Fadl Baihaqi's enormous Mujalladat, and of other works which would doubtless throw light on the history and culture of these far eastern fringes of the Islamic world. Yet there is enough literature surviving to show how the cultural trends went, and it is hoped that the present sketch of this evolution has thrown some light on Persian culture at this time.