asserts.

Instead,he claims, they insist thatthe souls of every person, the souls
of every thing, and God himself are all one and the same. According to the
somewhatcuriousaccounttold to GiuseppeFelice by his Nepalese informant,
God first createda woman named Manasa,29who soon began to long for off-
spring. God then createdthe gods Brahma,Vishnu, and Mahesa (Siva). Man-
asa invitedBrahmato matewith her,but he refused.Next she askedVishnuand
he too refused.Mahesa,however,accepted,providedManasawould changeher
form to that of Parvati.Brahma,Vishnu, and Mahesa each assumed different
functions:Brahmawould concernhimself with spiritualmattersand scripture,
Vishnuwith conservingand governing,Mahesawith punishmentand death.
GiuseppeFelice then notes the Hindusregardall the thirty-threecroremale
gods to be transformationsof these three gods while the many goddesses are
transformationsof Manasa.He gives a list of the names of nine goddesses that
correspondsroughlyto the list of emanationsof Devi found in the well-known
Puranatext, theDevi-mahatmya.Next, he turnsto a discussionof the ten avatars
of Vishnu, a standardtopic of most Puranas.After this he gives a descriptive
list of the gods associated with the various planets, the north star (Dhruva),
eclipses (Rahu),followed by a list of some of minor gods such as Yama,Ku-
bera,Indra,Varuna,Agni, Vayu,Kumara,Ganesa,andthe eight Bhairavas.The
discussionthen turnsbrieflyto the Brahmins'venerationof the sacredcow, fol-
lowed by the story of the origin of the Brahmins,the Kshatriyas,the Vaishyas,
and the Shudrasfrom God's mouth, shoulders,thighs, and feet respectively.
GiuseppeFelice concludeshis accountwith a descriptionof the fouryugas and
the Hindus'belief in enormouslylong time periods.
One-hundredyears earlier,in the first part of the seventeenthcentury,the
abovementionedJesuits de Nobili, Stephens, and FernandesTrancoso were
workingandwritingin southIndia,while the AugustinianManriqueworkedin
Bengal and elsewhere. Ratherthanreview the works of these authors,howev-
er, I want to use the example of a less well-informedtractfrom aboutthe same
period, A Discoverie of the Sect of the Banians, published by the Anglican chap-
lain Henry Lord, in 1630. Lord describes, with mixed success, the beliefs and
practicesof some Banias he encounteredduringa stay at the East India Com-
pany factory at Suratin Gujaratin the early seventeenthcentury.He claims to
have gone throughtheir Bible, the so-called "Shaster,"with the help of inter-
preters.30
29
As the editor,Petech, notes, this goddess apparentlyhas nothingto do with the serpentgod-
dess Manasaof Nepal. Rathershe seems to be equivalentto God's power or sakti.Theredoes not
seem to be any identifiabletextualsource for this story.
30 Lord 1630:[A]13. Lord arrivedin Suratin 1624. It is curious that John ZephaniahHolwell
claims thatduringthe captureof Calcuttain 1756, he "lost many curiousGentoo manuscripts,and
among them two very correctand valuablecopies of the Gentoo Shastah"as well as "a translation
I made of a considerablepart of the Shastah, which had cost me eighteen months hardlabour."
(Marshall1970:46). Shaster and Shastah are of course equivalentto sastra, a generaltype of text
and not a specific title. I suspect thatHolwell may have Lord'sShasterin mind.
I thatthoughtmy observancewould bee well tooke, if I could presentmy Countreymen
with any thing new from these forraigneparts,begun my worke, and essayed to fetch
materialsfor the same out of theirManuscripts,andby renewedaccesse, with the helpe
of Interpreters,made my collections out of a booke of theirscalled the Shasterwhich is
to them as their Bible, containingthe groundsof theirReligion in a writtenword.
Lord then presents those scattered bits of information that managed to pass
through the filter of his interpreters. He begins with a confused account of God's
creation of the world and of Pourous (Purusa) and Parcoutee (Prakrti) that reads
like a twisted version of the creation in Genesis.
For this cause that Almighty consulted with himselfe, about the making of this great
worke which men call the world or Universe, and as the Ancient (say they) have deliv-
ered; the Lord made foure Elements as the ground-workeof this mighty frame;to wit,
Earth,Aire,Fire, and Water,which foure Elements were at first all mingled togetherin
a confusion, but the Almighty separatedthem in mannerfollowing.
First, it is delivered, that by some great Cane or like instrument,hee blew upon the
Waters,which arose into a bubble,of a roundform like an egge, which spreadingit selfe
furtherand further,made the Firmamentso cleare and transparent,which now compas-
seth the world about.
[GodthencreatedEarth,Sun,Moon, the pointsof the compass,andfinally Man.]That
this creaturemightnot bee alone, who was madeby naturesocieable;God secondedhim
with a Companion,which was Woman;to whom not so much the outwardshape, as the
likenesse of the minde and disposition seemed agreeing:and the first mans name was
Pourous, and the womans name was Parcoutee,and they lived conioyned together as
Man and Wife, feeding on the fruitesof the earth,withoutthe destructionof any living
creature.

Next Lord claims that the "Banians" have a social system based on descent
from the four sons of Pourous and Parcoutee named Brammon, Cuttery, Shud-
dery, and Wyse (Brahmin, Kshatriya, Shudra, and Vaishya). This account, how-
ever confused, seems to be indirectly based on the sacrifice of the primeval Pu-
rusa in the Rg-veda hymn x. 90. Although Lord inverts the names of Vaishya
and Shudra, his description of the division of labor among these four varnas is
otherwise more or less accurate. He also manages a rather confused but recog-
nizable account of the cycle of transmigration and its connection to vegetarian
ahimsa; of the trio of gods Bremaw (Brahma), Vystney (Vishnu), and Ruddery
(Rudra-Siva); and of the four yugas. In his discussion of the Parsis, he also man-
ages to recognize the basic differences of religion among the Banians (i.e. the
Hindus), the Muslims, and the Parsis (Lord 1630: [D]1):
Having declaredthe Religion, Rites, Customes & Ceremonies,of a people living in the
East Indies called the Banians, a Sect not throughlypublishtby any heretofore,whilst
my observationwas bestowed in such Inquiry,I observed in the towne of Surrattthe
place where I resided,anotherSect called the Persees, who because I did discernethem
to differ both fro the Moore & Banian in the course of their living, & in the forme of
theirReligion.
Even in Lord's rather garbled account, which was based on his visual obser-
vations and conversations with native interpreters (whose grasp of English was
evidently limited),the basic outline of the set of beliefs andpracticesthatcame
to be known as Hinduism is clearly visible. Since this is one of the earliest
known extendedEuropeandescriptionsof Hinduism,it seems fair to claim that
Hinduism,if it was in fact constructedby the Europeans,can be tracedback to
the very earliest Europeanaccounts. The fact that virtually all Europeanac-
counts-whatever the language or period in which they were written, and
whetheror not they are likely to have mutuallyinfluencedeach other-follow
this same generaloutline suggests thatthe Europeanwriterswere in fact "con-
structing"Hinduismdirectlyon the basis of what they observedand what they
were told by theirnativeinformants.These informantswere in turnsimply sum-
marizinga constructionof Hinduismthat alreadyexisted in their own collec-
tive consciousness. This does not mean that Hinduismwas unchangedduring
this period, nor that the Europeanand colonial presence did not foster impor-
tant changes in the way Hinduism was conceptualizedby the Hindus them-
selves, but it does clearly show thatthe idea thatHinduismwas constructedor
inventedby nineteenthcenturyEuropeansis mistaken.

HINDU IDENTITY BEFORE 1800

A large partof the claim thatHinduismdid not exist before it was inventedby
the Britishin the early 1800s dependson the belief thatbeforethis datethe Hin-
dus themselves lacked a self-conscious religious identity,as opposed to a dif-
fuse ethno-geographicalidentity.The textual evidence againstthis claim is so
overwhelmingthatI am franklyat a loss to explainwhy so many worthyschol-
ars apparentlyaccept it. I have only two tentativeideas to offer as at least par-
tial explanations.
First, many liberal and left scholars have been reluctantto accept the idea
that the often antagonisticmodem Hindu and Muslim identities,both individ-
ual and communitarian,aroseout of political andreligious conflicts duringthe
historicalperiodsof the Delhi Sultanate,the MughalEmpire,and the regional
Sultanates.These scholarstendto attributethe invention,construction,or imag-
ining of the modernform of all majorIndianinstitutionsto the influence of the
colonial statein particular,and the nineteenth-centuryimaginaryin general.As
the political scientistPaul Brass once said to me, only half joking: "Everything
was invented in the nineteenthcentury."When it comes to communalismper
se, as opposed to the wider concept of Hindu identity, most liberal and left
scholars-from Bipan Chandra(1984) to GyanendraPandey (1990) to Veena
Das (1990) and beyond-insist that it was originallya productof the colonial
period.The argumentsof scholarslike ChristopherBayly and SheldonPollock
that the roots of communalismlie deeperin the past have often met with open
hostility.31
31 See
Pandey's (1990:15) hostile reaction to Bayly's 1985 article about the "pre-history"of
communalismand B. Chattopadhyaya's(1998:98-115) extendedcriticismof Pollock's 1993 arti-
cle. "RamayanaandPoliticalImaginationin India."Chattopadhyaya'smost palpablehit is thatPol-
 The second possible reason why many scholarshave not acknowledgedthe
existence of a conscious Hindureligious identitybefore 1800 is perhapsmore
interesting.My argumenthere startsfrom the fact that Sanskritliteraturewrit-
ten before this date systematically ignores the Muslim presence. Muslims,
when they appear,are mostly describedsimply as "foreigners."In his fascinat-
ing new book on referencesto Muslimsin Sanskritsources,mostly inscriptions,
B. Chattopadhyaya(1998:92-97) lists about seventy-five Sanskritreferences
to Muslims. Among the terms used are tdjika, turuska, mleccha, parasfka, ya-
vana, hammira, and sdka. The term musalamdna is listed only once, from an
inscriptionof 1264 C.E.Althoughsome of these termsareethnicin origin,most
had acquiredthe broadgeneric meaningof "foreigner"or "westAsian"by the
time they were appliedto Muslims.
In both Sanskritand vernacularliterature,Muslims were often portrayedin
the implicit, coded form of demons, as evidently happenedin severalmedieval
renderingsof the Ramayana.Most importantin NorthIndiawas, of course,Tu-
lasi-das's late sixteenthcenturyHindi Ram-carit-manas,which soon achieved
the status of a foundationaltext for the Hinduismof a broadrange of castes,
with the notableexceptionof the mostly low-caste followers of the KabirPanth
and related nirgunl sects. Kampan'sTamil Iramavataramplays a somewhat
analogousrole in the South. Early Europeansand Christianswere sometimes
similarlycoded as demons.
As far as I know, no premodernSanskrittext includes anythingapproaching
a systematic discussion of Muslim beliefs and practices. Similarly,the terms
"Hindu"and "Hindudharma"were never admittedto the premodernSanskrit
lexicon. The roughly equivalent term "sanatana-dharma" can, it is true, be
traced back to the Bhagavad-gita and the Puranas,but, as Wilhelm Halbfass
and other scholarshave argued,its precise meaninghas always been ambigu-
ous.32
Why exactly Hindu Sanskritliteraturewrittenbefore 1800 treatedforeign-
ers and foreignreligions, even indigenousBuddhism,in this Olympianfashion
is not easy to understand,and would make an excellent topic for a separatees-
say, but here it is sufficient to note that this systematicignoringof non-Hindu
culturaltraditions,whateverits cause, was a deeply embeddedtraitof premod-
ern Hinduism.Halbfass's (1988:187) ratherharshjudgment in this regardis,
I think, inescapable:

lock has made a very selective readingof the evidence when he associates the reactionby Hindus
against Islam and Turko-Afghanor Mughal conquest exclusively with the rise of the devotional
cult to Rama.The Hindusundoubtedlyconceptualizedand mythologicallyrepresentedthe conflict
between Hindu kings and Turko-Afghansor Mughals with figures other thanjust Rama and Ra-
vana, but this, I think,only puts an interestingqualificationto Pollock's basic argument.
32 See the discussion andreferencesin Halbfass 1988:310-48. I
suspectthatthe use of the term
"eternaldharma"may have been used in partto distinguishBrahmanicaland Hindureligion from
the more historical religions of Buddhism and Jainism, but Buddhist scholars have told me that
Buddhismitself is sometimes called the "eternaldharma."
TheIndocentrism developedin "orthodox" Hinduthoughttranscendsby farwhatis or-
dinarilycalled"ethnocentrism."Itis notsimplyanunquestionedperspective orbias,but
a sophisticatedtheoreticalstructureof self-universalization
and self-isolation.Seen
fromwithinthiscomplex,highlydifferentiated themlecchasarenothingbut
structure,
a faintanddistantphenomenon at thehorizonof theindigenoustradition.
Theydo not
possess an "otherness"againstwhich one's own identitycould be asserted,or in which
it couldbe reflected.Theyareneithertargetsof possibleconversion,nor sourcesof
potentialinspiration.
What I am suggesting here is that many modern scholars,especially those
who workprincipallywith Sanskritsources,may have unconsciouslyabsorbed
some of the self-imposedculturalisolationof premodernSanskritliteratureand
then concluded that there was no Hindu awarenessof the Muslim Other.As a
consequence, they may also have assumed that the Hindus had no clear con-
trastiveawarenessof their own religious identity.
Whateverthe reason for the scholarly acceptanceof the idea that there was
no religious Hindu self-identitybefore 1800, the evidence againstthis view in
vernacularHindu literatureis clear and abundant.The bulk of this evidence
takesthe formof texts composedby the popularreligiouspoet-singersof North
India,most of them membersof non-Brahmincastes. This literaturedoes pre-
cisely what Sanskritliteraturerefuses to do: it establishes a Hindu religious
identitythrougha process of mutualself-definitionwith a contrastingMuslim
Other.In practice,therecan be no Hinduidentityunless this is defined by con-
trast against such an Other.Withoutthe Muslim (or some other non-Hindu),
Hinduscan only be Vaishnavas,Saivas, Smartasor the like. The presenceof the
Otheris a necessaryprerequisitefor an active recognitionof what the different
Hindu sects and schools hold in common.
To illustratethis process of mutualself definition, I will use passages from
the nirgun- poet Kabir,the RamanandiAnantadas,the VarakariEkanath,and
the Krishnadevotee Vidyapati.Many more such passages can easily be cited,
especially from the poets of the nirguni sant traditionsuch as Raidas, Nanak,
Dadu, Rajjab,and PalatuDas, to name just a few. More orthodoxvernacular
poets such as Tulasi Das generally follow the evasive Sanskriticstrategiesof
representingMuslims as ill-defined mlecchas or coded demons, but even in
these writings such coded Muslims are often a palpable, and even necessary,
presence.
Ekanathwas a Brahmin,a scholarand author,who spent most of his life at
Paithanin Maharashtra.He was borntherein 1533 and died therein 1599. Al-
thoughhe knew Sanskritwell, most of his numerouscompositionswere writ-
ten in Marathi.Among them is a humorouspoetic dialogue between a Hindu
Brahmin and a Muslim, the Hindu-turka-samvada.33 The term "Turk"(turka),
like "Hindu,"can be used in a ethno-geographicalsense, but here-as in the

33 The text has been translatedand commentedon in English by Eleanor Zelliot (1982). The
passages cited are from this translation.