Though the earliest language spoken in this region was probably an Austroasiatic language with Dravidian and tibeto-burman superstratum, we know nothing about the songs and stories composed in this language. Even though pANini mentions eastern grammarians, nothing has survived of this scholarly literature either. The first attested writing comes from the maurya period at mahAsthAnagaDh.a where a brAhmI inscription shows an example of contemporary mAgadhI prAkRta. The situation of our knowledge hardly improves in the succeeding gupta period, even though the statements of chinese scholars visiting bengal (e.g. fa hien in 5th century) points to a vibrant buddhist academic community. It is not known what language the name of the tribe comes from, but an Austric root Bonga meaning sun-god has been claimed, as also a Dravidian tribe Bang from around 1000 BC.
It is only from a later period that we meet some of these scholars as they have left their names on buddhist treatises copied and preserved in tibet: an important one amongst them was shIlabhadra, disciple of dharmapAla and teacher of Hieun-Tsang (visited in 7th century), and who became the head of nAlandA. In the heydays of nAlanda, a lot of scholars lived and worked here, but we often do not know where they were from. A famous grammarian candragomin, also the author of nyAYasiddhAloka (who may not have been the same person as the vajrayAnI candragomin) was from varendrI. gauD.apAda, who wrote gauD.apAdakArikA, sAMkhyakArikA, and uttaragIta, and who was the teacher of the teacher of shaMkarAcAryya, was probably also a bengali. Eastern India was also known for its scholarship on elephant treatment, and it is possible that pAlakApya, to whom hastyAYurveda is attributed, was a real person (coincidentally, pAla of dravidian origin could mean an elephant) living somewhere on the banks of the brahmaputra. The evidence also points to widespread scholarship in the traditional disciplines: of the vedic schools, the vAsaneYI branch of yajurvedic tradition was the most common.
When it came to literature, there was a separate style called gauD.I which was known for its use of sounds, as described by vANabhaTTa: zleSaprAYamudIcyeSu pratIcyeSvarthamAtrakam | utprekSA dAkSiNAtyeSu gauD.eSvakSaraDambaram || navo-rtho jAtiragrAmya zleSo-kliSta sphuTo rasaH | vikaTAkSarabandhazca kRtsnamekatra duSkaram || daNDI also contrasted the overuse of the figures of speech and complicated twists of meaning of the gauD.I style to the smoother flowing and punny vaidarbhI style; though bhAmaha seemed to prefer the former. The highly ornamented style of the inscriptions by lokanAtha at tripurA and bhAskaravarmA at nidhAnapura attest to a similar devlopment in royal proclamations. It seems that a mAgadhI style developed out of the gauD.I at later times.
Similarly, when it came to plays, bharata mentions four major styles: avantI, paJcala mAdhyama, dakSinAtyA, and oDra mAgadhI, the last of which was practiced in eastern India. Thus, by the time of zazAGka, eastern India was already on its way to developing independently in its language, literature, and arts; though a distinct bengali identity can not to be seen at this stage yet.
The actual rise of he bengali identity probably had to await the pAla dynasty, many of whose ministers like darbhapANi, kedAramizra, and guruvamizra were renowned scholars themselves. It was during this time that literature was composed mainly in pure or buddhist saMskRta, with a strong presence of zaurasenI apabhraMsa. prAkRta was not pronounced ‘properly’ according to rAjazekhara, so much so that he imagined sarasvatI saying ‘brahman viJjApaYAmi tvAM svAdhikArajihAsaYA | gauD.astyajatu vA gAthAmanyA vA-stu sarasvatI’. The language of the common man evolved through a mAgadhI dialect to the earliest form of bengali preserved in the slightly later caryyApada, though zaurasenI dialects were probably understood easily. saMskRta poetic creations is evident in the inscriptions from this period. An early form of the bengali script is attested in manuscripts from this period. However, the pure sanskritic influence increased in scholarly writing over time, though some buddhist traditions might have maintained the other traditions. It should also be pointed out that in this period, bengali tradition includes almost all of what is now bihar, and no attempt has been made here to distinguish the two.
The literary produce of this period includes writings by one or several abhinandas. rAmacarita by abhinanda (during the time hAravarSa) contains devImAhAtmya by hanumAna. This was later followed by rAmacarita (probably finished under the reign of madanapala) by sandhyAkara nandI, son of prajApati nandI, son of piNAka nandI of puNDravardhanapura, a poem that applies equally to the story of rAma and to the recent history since the kaivarta revolt. It is not clear whether kSemIzvara, author of caNDakaushika and naiSadhAnanda, was in the court of mahIpAla of bengal or mahIpAla of gurjara pratihAra. nItivarma, author of kIcakavadha, may also have been a bengali, and similar unsubstantiated claims have been made for vizAkhadatta who wrote mudrArAkSasa.
In addition, the oldest collection of poems and snippets, called kavIndravacanasamuccaYa, was probably a bengali creation. It collects poems by kAlidAsa, amaru, bhavabhUti, rAjazekhara, gauD.a abhinanda, D(/h?)imboka, kumudAkara mati, dharmakara, buddhAkara gupta, madhuzIla, vAgoka, lalitoka, vinaYadeva, chittapa, vandya tathAgata, jaYIka, vitoka, vidyAkA (or vijjokA), vinaYadeva, vIRyamitra, vaiddoka, zubhaGkara, zrIdhara nandI, ratipAla, yogoka, siddhoka, so(n)noka, hiGgok, vaidyadhanya, aparAjita rakSita, and others. Though the story of kRSNa's dalliance with the cowherd girls (gopIs) of vraja has a long tradition, and the phlosohphical and erotic aspects already highly developed in the c. 6th century harivaMsa and 10thrAdhA and kRSNa is probably (there are isolated instances of idols of kRSNa and a lady before) in an inscription by vanamAlavarma, king of kAmarUpa, and certainly in an inscription by bhojavarma; and it appears in poems in this collection. The name rAdhA does appear in gAthAsaptazatI composed during hAla, the sAtavAhana king from the first century, and it also appears sporadically in prior prAkrRt literature. Notable contemporary mentions are found in pre-8th century bhaTTanArAYaNa's venisaMhAra, 10th century abhinavagupta's dhvanyAlokalocana, kSemendra's dhavatatacarita from around 1066, and in a benedictory verse found in inscriptions of paramAra vAkpati muJja of mAlva around 973–994. This story, of course, reached prominence in the gItagovinda much later (12th century), and further developed by vidyApati and caNDidAsa in the 14th and 15th centuries.
The saMskRta philosophical writings of this period include prakAza (a TIkA on kezavamizra's chAndogya-pariziSTa) written during the rule of devapAla by nArAYaNa, son of goNa, son of umApati of northern rADh.a; nyAYakandalI (a TIkA on the bhASya by prazastapAda of the vaizeSika sUtra padArthadharmasaMgraha) written around 980 AD for kAYastha pANDudAsa by zrIdharabhaTTa (who also wrote advaYasiddhi, tattvaprabodha, tattvasaMvAdinI, and saMgrahaTIkA), son of baladeva and AbbokA in bhUrizreSThI in sjayapalaouthern rADh.a; lakSaNAvalI, kiraNAvalI, kusumAJjalI, and Atmatatvaviveka by udaYana who is mentioned in the traditional mythology of brahmin origin in bengal; and yogavaziSThasaMkSepa by abhinanda.
The grammatical writings include vivaraNapaJjikA jinendrabuddhi, tantrapradIpa and dhAtupradIpa by maitreYarakSita, and kAmadhenu by subhUticandra, though whether these authors and vimalamati were bengali has not been determined. jitendrIYa and bAlaka probably lived in bengal in this period, and wrote about religion and morals; and yogloka may have been a predecessor to them. kalyAnavarmA of vyAghrataTI wrote sarAvalI, a treatise on astrology.
In medicine, one finds cakrapANi datta and his brother bhAnu, son of nArAYaNa, the head of kitchens of a king of gauD.a, possibly referring to jaYapAla or naYapAla. According to tradition, this cakrapANi datta of the lodhravali kulIna dynasty and student of naradatta was from vIrabhUma. He wrote AYurvedadIpikA (also called carakatAtparyyadIpikA), bhAnumatI (TIkA on zuzruta), zabdacandrikA, dravyaguNasaMgraha, and cikitsAsaMgraha. surezvara (in an unidentified king bhImapAla's court), son of bhadrezvara (doctor in rAmapAla' court), son of devagaNa (doctor in govindacandra's court) wrote zabdapradIpa, vRkSAYurveda, and lohapaddhati (or lohasarvasva). vaGgasena, son of gadAdhara of kAJjikA, wrote cikitsAsArasaMgraha. Just as doubtful is whether aruNadatta, vijaYarakSita, vRndakuNDa, zrIkaNThadatta, and gaYadAsa were bengalis. It is extremely doubtful if mAdhava, author of rugavinizcaYa and nidAna was a bengali; on the other hand, nizcalakara, who wrote ratnaprabhA, probably was and lived during rAmapAla.
The buddhist writings are difficult both to date and place; with the universities in bengal, people arrived from all around to work here. Even when the tibetan sources give place names like j(/s)Ahora or uDDIYAna, it is not clear what they refer to. To add to that, some personal names are repeated so often, that identities remain in doubt.
Thus we find buddhist teachers like zAntirakSita who composed aSTatathAgatastotra, vajradharasaMgItabhagavatastotraTIkA, and paJcamahopadeza; bodhisattva who wrote books on saptatathAgata; and zAntarakSita who composed books like tattvasaMgraha, vAdanyAYavRtti vipaJcitArtha, and madhyama kAlaGkAra kArikA; all of whom may have been the same person. He was born during the reign of gopAla, and he died during dharmapAla's reign, and had traveled to tibet under khri-srang-lde-vtsan. He was probably almost contemporary of sarorUhavajra/padmavajra, kamalazIla, and padmasambhava.
kukkuripAda from around the 8th–9th century might be the same as kukur pA, and composer of some songs in caryAgIti. kambalapAda or kambalAmbarapAda composed kambalagItika and some songs in caryAcaryavinizcaYa. zavarIpAda, who may have been the same as zavarIzvara, kumAracandra, taGkadAsa or daGkadAsa, and nAgabodhi of shivasera village who composed yamArisiddhacakrasAdhana were also from this period.
One again finds a lot of buddhist thinkers starting in th 10th century, many of whom may have been from bengal. The style of poetry used when worshipping the goddess in bengal is quite clear in compositions of scholars such as dharmAkaramati, zavarapAda, kRSNapAda, ratnAkara, zubhAkara, kuladatta, advaYavajra, lalitagupta, kumudAkaramati, padmAkara, abhaYAkara gupta, guNAkara gupta, karuNAcala, kokaradatta, anupama rakSita, cintAmaNi datta, sumati bhadra, maGgala sena, and ajita mitra.
At the end of 10th century, we find jetAri, son of garbhapAda, and teacher of atIza zrIjJyAna dIpaGkara. (There seems to be a different jetAri sometime later.) The latter was born in around 980 AD to kalyAnazrI and prabhAvatI in vikramaNipura, died in tibet at 73 years of age, and may not have been identical to the other dIpaGkaras (e.g. dIpaGkara bhadra, dIpaGkara rakSita, or dIpaGkara candra). jJanazrI mizra, abhaYAkara gupta, divAkara candra, kumAravajra, ratnAkara zAnti, dAnazIla, vibhuti candra, prajJAvarmA, mokSAkara gupta, puNDarIka or jJanavajra, and lui pA who may have been the same as mInanAtha matsyndranAtha were probably only slightly later. mInanAtha's student gorakSanAtha and his student jAlandharIpAda and his student virUpA followed. During mahIpAla's time, we find tilapa / tillapa / tillipya / tilipA / tillopA / tailopa / tollapA / telopA / tilopA / tailikapAda / teliyogI or prajJAvarmA / prajJAbhadra from TsATigÃo (caTTagrAma). His main student was nAro / nAropA / nArotpA / nAD.opA / nAD.a / nAD.apAD.A or jJAnasiddhi yazobhadra. A student of jAlandharIpAda, kRSNa / kRSNapAda / kahNu pA / kAhNa pA is also important. Others from this period include dArika, kila pA, krmAra, vINA pA, dharmapAda or guNDArIpAda, kaGkaNa, and garbhapAda.
Confusingly, there are a couple zAntidevas, one of whom is probably a mahAyAnI from 8th century saurashtra, whereas the other is probably a vajrayAnI tAntrik from 11th century who wrote shrIguhyasamAjamahAyogatantravalavidhi, sahajagIti, and cittacaitanyazamanopAYa. The latter had alternate names rAutu or bhusuku, and may have been identical to a poet of caryyAgIti. Similarly zAntipA of these songs may be identical to zAntipAda who wrote books like sukhaduHkhadvaYaparityAgadRSti. One can also guess that saraha rAhulabhadra is identical to the folk poet sarahapAda.
These scholars mainly wrote about vajrayAna, and some composed dohAs and songs. However, a few reflect on philosophy, especially scientific philosophy. Thus, during dharmapAla, haribhadra, in abhisamaYAlaGkArAvaloka, tried a synthesis of the madhyamika school of nAgArjuna and yogAcAra school of maitreYanAtha. Other notables from roughly the same period are buddhazrIjJana or buddhajJAnapAda, jinamitra (who, along with dAnazIla and zIlendrabodhi, wrote a saMskRta-tibbetan dictionary).
The songs collected in the later caryyApada (or caryyAgIti or caryyAcaryyavinizcaYa, songs about how to live life or decide between what should or should not be done: available here) were by such poets as lui-pA, kAhna-pA, jAlandharI-pA or hAD.i-pA, zavarI-pA, bhusuku, tantrIpAda, who were probably living between the tenth and 12th centuries. The songs are mainly in mAtrAvRtta with antyamila. They were set to rAgas. The paYAra (or lAcAD.I) metre was probably derived from the metres seen here (padakulaka). The language is clearly bengali, but sometimes influenced by zaurasenI and maithilI. The zaurasenI dohAs written by kAhNa, sarahapAda, and dAka (dAkArNava) also influenced bengali literature in this period, they also show influence of local bengali and maithilI dialects.
Just as the previous pala period saw the rise of bengali literature, the sena period was probably the high point of saMkRta literature in Bengal. Even then, it's actual literary output was little. This was mainly a time of a reconstruction of hinduism in bengal, but actual new scientific and philosophical thought seems very limited in bengal proper. One should also point out that occasionally there is a problem identifying a person as native of orissa or bengal/bihar region: here it has been chosen to treat a personage of the sena court as a ‘bengali’ even if there is dispute as to their birthplace.
In mAnasollAsa (or abhilaSitArthacintyamaNi), a collection made in 1051 zaka year for cAlukyarAja somezvara III of gujarat, we find some bengali songs about kRSNa with the gopIs in vRndAvana and about the avatAras of viSNu. In the fourteenth century collection of avahaThTa (apabharMza) poems, prAkRta paiGgala, a few poems are probably from this old period in bengal: they are descriptions of simple lives and loves, or about gods in human terms. Similarly, some zlokas in vidagdhamukhamaNdala are probably from this period. Same may be true of a poem in zeka-zubhodaYA.
Based on the absence of noticeable turkish influence, and the society being described, common sayings attributed to Dak, khanA, and zubhaGkar probably contain sayings that originate in the ancient period, though their language is much more modern. The same may be true of the popular stories like that of cÃd sadAgar-lakhIndara-behulA-dhanapati-lahanA-khullanA-zrImanta-kAlketu in caNDImaGgal-manasAmaGgal, lAusena and king gopIcandra, his mother maYanAmatI, and his wives aduna-paduna in songs of gopIcÃda. However, by the time of their earliest attestation, these all belong to the medieval period in the history of bengal.
The emphasis on sounds and word play of the previous period of saMskRta poetry seems to have given way to a more idea based structure. The sena court supported and the kings have been compared to kRSNa himself, and, for the first time, the kRSNa that had close physical relationship with rAdhA. In retrospect, this humanization of the sacred gods and advent of the concept of bhakti paved the way for the religious humanism which was to become the hallmark of hinduism in the medieval times. Perhaps the greatest example of this is gItagovinda a saMskRta poem that is stylistically very close to bengali. This was composed by jaYadeva (12th cent), a poet born of bhojadeva and vAmA (or rAmA or rAdhA) devI in kindubilva, and husband of dancer padmAvatI (who may have been promised to the puri temple when young). Even though this is not the first mention of rAdhA, it is the text which popularized these stories all over India. jaYadeva also wrote a few songs in apabhraMza, and a few of his songs (which suggests he was a paJcopAsak smArta brAhmaNa rather than vaiSNava) are collected in SaduktikarNAmRta, some of which are even of vIRarasa.
Sanskrit tradition at the sena court did encompass philosophical treatises dealt, but they dealt not with nyAYa as in earlier periods, but rather with mimAMsA: for example mimAMsAsarvasva by halAYudha and tautAtitamatatilaka by bhavadeva bhaTTa. bAlabalabhI resident siddhalagrAmavAsI (in rADh.a) sAmavedIYa kauThumazAkhAdhyAYI sAvarNagotrIYa bhavadeva bhaTTa ‘dvitIYa varAha’ was one of the greatest scholars of this period, and wrote about horAzAstra (none extant) and dharmazAstra (including vyAvahAratilaka, prAYazcittaprakaraNa or prAYazcitta nirUpaNa, and chAndogyakarmAnuSThAn paddhati or dazakarmapaddhati or saMskArapaddhati or dazakarmadIpikA). pAribhadrIYa mahAmahopAdhyAYa jimUtavAhana wrote kAlaviveka, vyAvahAramAtRkA, and dAYabhAga, the last of which still defines the succession law for bengali hindus. campAhaTTIYa mahAmahopAdhyAYa aniruddha, teacher of vallAla sena, wrote hAralatA and pitRdaYita. vallAla sena himself wrote AcArasAgara, pratiSThAsAgara, dAnasAgara, and adbhutasAgara (started in 1168 AD, finished by his son lakSmaNa sena). dAmuka putra guNaviSNu wrote chAndogyamantrabhASya. mahAdharmAdhyakSa mahAdharmAdhikRta dharmAgArAdhikArI halAYudha, son of vatsagotrIYa dhanaJjaYa and ujalA, was another important scholar in lakSmaNasena's court and wrote brAhmaNasarvasva (about zuklayajurvedIYa kANvazAkhAdhyAYI brAhmaNas), mimAMsAsarvasva, vaiSNavasarvasva, zaivasarvasva, and paNDitasarvasva. grammarian and lexicographer puruSottama may have been different people: the latter wrote trikANDazeSa, hArAvali, varNadezanA, and dvIrUpakoSa, of which varNadezanA describes difficult spellings due to the gauD.IYa script. Lexicographer artiharaputra vandyaghaTIYa sarvAnanda wrote TIkAsarvasva (1159 AD), a resource for early local words, and one that collects zlokas from extant sources, some of which like sAhityakalpataru, devIzataka, vidagdhamukhamaNDala, vRndAvanayamaka, and zrIpovyoka's vAsanAmaJjarI could conceivably be from bengal.
This period also saw the composition of naiSadhacarita by zrIharSa. Even though bengali tradition declares his father's name as medhAtithi (or tithimedhA), we know that he is the son of zrIhIra and mAmalladevI, and it is not even clear whether he is a bengali. naiSadhacarita is very much in the gauD.I style with its prevalence of sound play like anuprAsa, and many of its puns are based on peculiarities of eastern pronounciation (equivalence of z, S, and s, of j and y, and of N and n). The same goes for its description of rice and other food, and the marriage rituals. zrIharSa also wrote in praise of a gauD.a king (could be vijaYasena), and books like narasAhasAMka carita, sthairya vicAra prakaraNa, arNava varNanA, zivazaktisiddha, chindaprazasti, zrIvijaYaprazasti. He also had a treatise on the khaNDanakhaNDakhAdya philosophy.
Bengali tradition also claims bhaTTanArAYaNa, author of veNIsaMhAra as a bengali. The same may be true of murArI mizra, author of anargharAghava, though in either case, evidence is scarce. Before the 15th century, sAgaranandI mentions a lot of plays by bengalis, with names like kIcakabhIma, pratijJAbhIma, zarmiSThApariNaYa, rAdhA, satyabhAMA, keliraivataka, USAharaNa, devImahAdeva, urvazImardana, nalavijaYa, mAyAmadAlasA, unmatta candragupta, mAYAkApAlika, mAYAzakunta, madanikAkAmukA, jAnakIrAghava, rAmAnanda, kekaYIbharata, ayodhyAbharata, vAlibadha, rAmavikrama, and mArIca vancitaka, but it is difficult to place exact times and places to these.
A few sanskrit poetic collections are also available from this period. The foremost amongst these, saduktikarNAmRta (or sUktikarNamRta) was collected in 1206 AD, probably under the patronage of keshavasena, by zrIdharadAsa, son of zrIvaTudAsa. This collection from 854 (and a few unknown) poets was divided into five pravAha (475 zloka in 95 vIci on lilA, 895 zloka in 179 vIci on zRGgAra, 270 zloka in 54 vIci on cATu, 360 zloka in 72 vIci on apadeza, and 370 zloka in 74 vIci on uccAvaca). Studying the names (apart from the well known zaraNa, umApati dhar, jaYadeva, govardhanAcArya, son of nIlAmbara, author of a text on dharmazAstra, who wrote AryA saptazati with help from udaYana and balabhadra, dhoYI kavirAj who wrote pavanadUta, lakSmaNasena, and kezavasena, we also find jalacandra, yogezvara, vaidya gaGgAdhara, sAJcAdhara, vetAla, vyAsa kavirAja, kevaTa, papIpa, vaGgAla, candracandra, gAGgoka, vimboka, zuGgoka, mathu, sAJcAdhara, etc.), it seems likely that many of the poets were bengali, though, of course, the collection also contains poems by such well known non-bengalis as pANini, bhAsa, bhAravi, kAlidAsa, bhAmaha, amaru, vANabhaTTa, vilhaNa, bhartRhari, muJja, rAjazekhara, vAkpatirAja, and vizAkhadatta. Other examples of compositions from this period may be found in inscriptions like bhavadeva prazasti, bArAkapUra and deopAD.A prazasti under vijaYasena, naihATi prazasti under vallAlasena, and AnuliYA, govindapUra, and tarpanadIghi-zAsana prazasti and mAdhAinagara paTTolI under lakSmaNasena. The vRhaddharma and brahmavaivarta purANas are only slightly later.
On the musical front, caryApada was set to rAgAs, as were the songs of gItagovinda. In addition, locanapaNDita wrote a treatise on music called rAgataraGginI in which he quotes from the earlier tumburunATaka. Though rAgatarGginI contains obviously late (14th century) references, it is possible that the core was composed in 1160 AD under vallAlasena. locana also wrote rAgasaGgItasaMgraha. Slightly later, zArGgadeva (1210–1247) composed saGgIta ratnAkara.
Up to history of ancient Bengal